OV - Compare Loops

Loops

▶mmq.cpp: 1138 - 34.48 %

ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default		Run icx_default							Run gcc_default		Run icx_5							Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions	/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h: 235-235 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1138-1151 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2035-2040 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2060-2074 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2093-2101 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2146-2152 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2488-2488						Loop Source Regions		Loop Source Regions	/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h: 235-235 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1138-1151 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2035-2040 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2060-2060 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2096-2096 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2146-2153 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2176-2176 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2488-2488						Loop Source Regions		Loop Source Regions
		546	0.04	0.02	0.32	0	0	567.61			510	0.02	0.01	0.13	0	0	1287.28
		557	1.89	1.67	34.04	0	0	679.68

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 2 analyzed binary loops (libggml-cpu.so - 546, libggml-cpu.so - 557)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis	Count	Analysis						Count	Analysis	Count	Analysis						Count	Analysis	Count	Analysis	Count
		Control Flow Issues
		Vectorization Roadblocks
		Presence of more than 4 paths						1

▶simd-mappings.h: 130 - 9.01 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default		Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1140-1140 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9402-9402 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13848-13848 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14192-14192 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14946-14946
										487	0.22	0.12	2.48	100	100	1302.84
										486	0.20	0.11	2.34	100	100	1162.49
										489	0.18	0.09	1.89	100	100	1016.41
										488	0.22	0.11	2.30	100	100	829.14

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 4 analyzed binary loops (libggml-cpu.so - 487, libggml-cpu.so - 486, libggml-cpu.so - 489, libggml-cpu.so - 488)
Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis						Count
										Data Access Issues
										Presence of constant non-unit stride data access						1
										Presence of indirect access						1
										Vectorization Roadblocks
										Presence of constant non-unit stride data access						1
										Presence of indirect access						1

▶mmq.cpp: 1140 - 8.86 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default		Run icx_5		Run aocc_10							Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1140-1142 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1150-1151						Loop Source Regions
								741	0.18	0.11	2.30	100	100	1293.52
								742	0.20	0.13	2.70	100	100	1331.18
								740	0.18	0.09	1.97	100	100	1010.63
								739	0.17	0.09	1.88	100	100	1030.85

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 4 analyzed binary loops (libggml-cpu.so - 741, libggml-cpu.so - 742, libggml-cpu.so - 740, libggml-cpu.so - 739)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis						Count	Analysis	Count
								Data Access Issues
								Presence of constant non-unit stride data access						1
								Presence of indirect access						1
								Vectorization Roadblocks
								Presence of constant non-unit stride data access						1
								Presence of indirect access						1

▶binary-ops.cpp: 18 - 2.77 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 18-18 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 18-18 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 18-18 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 18-18 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 18-18 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 18-18 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32
914	0.05	0.02	0.44	0	6.25	27.06	812	0.05	0.02	0.47	0	6.25	22.17	716	0.05	0.02	0.46	100	50	21.56	768	0.05	0.02	0.45	0	6.25	23.19	905	0.06	0.02	0.46	0	6.25	27.74	705	0.07	0.02	0.49	0	6.25	23.04

Sum on 1 analyzed binary loop (libggml-cpu.so - 914)							Sum on 1 analyzed binary loop (libggml-cpu.so - 812)							Sum on 1 analyzed binary loop (libggml-cpu.so - 716)							Sum on 1 analyzed binary loop (libggml-cpu.so - 768)							Sum on 1 analyzed binary loop (libggml-cpu.so - 905)							Sum on 1 analyzed binary loop (libggml-cpu.so - 705)
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count
Loop Computation Issues							Loop Computation Issues							Loop Computation Issues							Loop Computation Issues							Loop Computation Issues							Loop Computation Issues
Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1
Data Access Issues							Data Access Issues							Data Access Issues							Data Access Issues							Data Access Issues							Data Access Issues
Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						0	Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						1
More than 10% of the vector loads instructions are unaligned							More than 10% of the vector loads instructions are unaligned							More than 10% of the vector loads instructions are unaligned						1	More than 10% of the vector loads instructions are unaligned							More than 10% of the vector loads instructions are unaligned							More than 10% of the vector loads instructions are unaligned						0
Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks
Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						1

▶vec.h: 491 - 2.74 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 491-497						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 491-497						Loop Source Regions	/cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 935-935 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 11216-11216 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14192-14192 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 15160-15160 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 491-491						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 491-497						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 491-497						Loop Source Regions
2437	0.05	0.02	0.47	0	0	298.17	2188	0.07	0.03	0.52	0	0	213.55	1663	0.07	0.04	0.73	100	70	122.94	2199	0.06	0.03	0.54	0	0	400.46	1790	0.06	0.02	0.49	0	0	399.83

Sum on 1 analyzed binary loop (libggml-cpu.so - 2437)							Sum on 1 analyzed binary loop (libggml-cpu.so - 2188)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1663)							Sum on 1 analyzed binary loop (libggml-cpu.so - 2199)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1790)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis	Count
Control Flow Issues							Control Flow Issues							Control Flow Issues							Control Flow Issues							Control Flow Issues
Presence of 2 to 4 paths						1	Presence of 2 to 4 paths							Presence of 2 to 4 paths							Presence of 2 to 4 paths							Presence of 2 to 4 paths						1
Data Access Issues							Data Access Issues							Data Access Issues							Data Access Issues							Data Access Issues
Presence of constant non-unit stride data access							Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access							Presence of constant non-unit stride data access
Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks							Vectorization Roadblocks
Presence of 2 to 4 paths						1	Presence of 2 to 4 paths							Presence of 2 to 4 paths						0	Presence of 2 to 4 paths							Presence of 2 to 4 paths						1
Presence of constant non-unit stride data access						0	Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						0

▶<unknown>: 0 - 2.60 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions							Loop Source Regions							Loop Source Regions							Loop Source Regions							Loop Source Regions							Loop Source Regions
2738	0.01	0.00	0.00	0	0	0	2299	0.01	0.00	0.00	0	0	0	2750	0.01	0.00	0.00	0	0	0	2408	0.01	0.00	0.00	0	0	0	944	0.01	0.00	0.00	0	0	0	2491	0.01	0.00	0.00	0	0	0
2485	0.01	0.00	0.00	0	0	0	2402	0.01	0.00	0.00	0	0	0	1331	0.00	0.00	0.00	0	0	0	2521	0.00	0.00	0.00	0	0	0	2135	0.01	0.00	0.00	0	0	0	1220	0.01	0.00	0.00	0	0	0
2628	0.01	0.00	0.00	0	0	0	2414	0.01	0.00	0.00	0	0	0	2807	0.01	0.00	0.00	0	0	0	2302	0.00	0.00	0.00	0	0	0	2132	0.01	0.00	0.00	0	0	0	1337	0.01	0.00	0.00	0	0	0
2640	0.02	0.00	0.00	0	0	0	2405	0.01	0.00	0.00	0	0	0	3041	0.01	0.00	0.00	0	0	0	2307	0.01	0.00	0.00	0	0	0	2264	0.01	0.00	0.00	0	0	0	2563	0.01	0.00	0.00	0	0	0
2450	0.01	0.00	0.00	0	0	0	2308	0.01	0.00	0.00	0	0	0	3021	0.00	0.00	0.00	0	0	0	2311	0.00	0.00	0.00	0	0	0	2273	0.02	0.00	0.00	0	0	0	2675	0.01	0.00	0.00	0	0	0
2842	0.01	0.00	0.00	0	0	0	2301	0.02	0.00	0.00	0	0	0	2914	0.01	0.00	0.00	0	0	0	1932	0.02	0.01	0.15	100	50	6.48	2462	0.01	0.00	0.00	0	0	0	2808	0.01	0.00	0.00	0	0	0
2638	0.01	0.00	0.00	0	0	0	2164	0.01	0.00	0.00	0	0	0	2909	0.00	0.00	0.00	0	0	0	360	0.00	0.00	0.00	0	0	NA	2272	0.02	0.00	0.00	0	0	0	2417	0.01	0.00	0.00	0	0	0
2741	0.01	0.00	0.00	0	0	0	2304	0.02	0.00	0.00	0	0	0	2810	0.01	0.00	0.00	0	0	0	514	0.02	0.00	0.07	0	0	401.68	2370	0.01	0.00	0.00	0	0	0	2620	0.01	0.00	0.00	0	0	0
695	0.01	0.00	0.01	0	0	1644.21	1972	0.03	0.01	0.14	100	50	12.59	3022	0.01	0.00	0.00	0	0	0	359	0.01	0.00	0.00	0	0	0	1613	0.02	0.00	0.06	0	0	336.63	2568	0.01	0.00	0.00	0	0	0
107	0.00	0.00	0.00	0	0	NA	365	0.00	0.00	0.00	0	0	NA	512	0.01	0.00	0.03	0	0	1414.21	2317	0.00	0.00	0.00	0	0	NA	422	0.00	0.00	0.00	0	0	NA	2677	0.02	0.00	0.00	0	0	0
2600	0.00	0.00	0.00	0	0	NA	2302	0.00	0.00	0.00	0	0	NA	519	0.03	0.00	0.10	0	0	326.95	1783	0.01	0.00	0.01	0	0	0	1792	0.00	0.00	0.01	0	0	405.74	79	0.00	0.00	0.00	0	0	NA
1759	0.02	0.00	0.07	0	0	137.09	410	0.00	0.00	0.00	0	0	NA	520	0.02	0.00	0.08	0	0	1569.88	63	0.01	0.00	0.01	0	0	1.35	1546	0.00	0.00	0.00	0	0	0	585	0.00	0.00	0.00	0	0	NA
701	0.00	0.00	0.00	0	0	453.58	2301	0.00	0.00	0.00	0	0	NA	1387	0.01	0.00	0.00	0	0	0	1659	0.01	0.00	0.03	0	0	327.01	1549	0.01	0.00	0.01	0	0	0	1382	0.01	0.00	0.00	0	0	0
108	0.01	0.00	0.01	0	0	16.8	554	0.00	0.00	0.00	0	0	NA	91	0.00	0.00	0.00	0	0	NA	522	0.00	0.00	0.00	0	0	NA	1773	0.01	0.00	0.02	0	0	98.29	581	0.01	0.00	0.00	0	0	2274.74
377	0.01	0.00	0.00	0	0	0	1572	0.01	0.00	0.03	0	0	267.89	1678	0.01	0.00	0.02	0	0	71.82	521	0.00	0.00	0.00	0	0	0	68	0.00	0.00	0.00	0	0	NA	1651	0.01	0.00	0.01	0	0	129.95
2105	0.01	0.00	0.05	0	0	854.14	403	0.00	0.00	0.00	0	0	0	513	0.01	0.00	0.01	0	0	2862.22	62	0.00	0.00	0.00	0	0	NA	736	0.01	0.00	0.03	0	0	1475.96	578	0.01	0.00	0.02	0	0	1649.18
2441	0.00	0.00	0.00	0	0	415.78	1811	0.02	0.00	0.07	0	0	285.78	1665	0.00	0.00	0.01	0	0	170.51	2185	0.01	0.00	0.04	0	0	69.79	388	0.01	0.00	0.00	0	0	0	1639	0.01	0.00	0.02	0	0	3391.37
705	0.01	0.00	0.05	0	0	2132.44	1720	0.01	0.00	0.01	0	0	0	525	0.00	0.00	0.00	0	0	NA	2202	0.01	0.00	0.05	0	0	20.19	732	0.02	0.00	0.05	0	0	84.66	59	0.01	0.00	0.01	0	0	0
407	0.01	0.00	0.00	0	0	0	122	0.01	0.00	0.01	0	0	1.18	71	0.01	0.00	0.01	0	0	0	1900	0.01	0.00	0.05	0	0	390.06	738	0.01	0.00	0.01	0	0	1300.11	579	0.01	0.00	0.02	0	0	2166.02
1969	0.01	0.00	0.01	0	0	0	552	0.02	0.00	0.06	0	0	2031.78	524	0.00	0.00	0.00	0	0	142.52	511	0.02	0.00	0.05	0	0	2523.7	737	0.01	0.00	0.01	0	0	576.38	577	0.01	0.00	0.02	0	0	1153.16
2410	0.01	0.00	0.02	0	0	105.93	555	0.00	0.00	0.00	0	0	0	990	0.01	0.00	0.05	0	0	398.71								731	0.01	0.00	0.02	0	0	255.1	588	0.01	0.00	0.03	0	0	587.64
704	0.01	0.00	0.04	0	0	441.95	556	0.02	0.00	0.10	0	0	281.07	331	0.01	0.00	0.01	0	0	0								735	0.02	0.00	0.02	0	0	2464.79	971	0.01	0.00	0.04	0	0	991.03
317	0.01	0.00	0.00	0	0	0	2190	0.02	0.00	0.06	0	0	15.24	1321	0.01	0.00	0.09	0	0	136.82								733	0.01	0.00	0.03	0	0	807.15	587	0.01	0.00	0.04	0	0	2599.06
838	0.01	0.00	0.00	0	0	0	2175	0.02	0.00	0.04	0	0	36.25	591	0.00	0.00	0.00	0	0	NA								720	0.01	0.00	0.04	0	0	2650.6	1319	0.02	0.00	0.09	0	0	118.01
826	0.01	0.00	0.00	0	0	0								514	0.01	0.00	0.03	0	0	1697.81								1465	0.01	0.00	0.07	0	0	137.38	580	0.02	0.00	0.03	0	0	1078.54
														580	0.01	0.00	0.01	0	0	973.87								69	0.01	0.00	0.00	0	0	37.92	684	0.01	0.00	0.00	0	0	0
														1757	0.00	0.00	0.00	0	0	NA								719	0.02	0.00	0.04	0	0	420.37
														511	0.01	0.00	0.02	0	0	1432.3								734	0.01	0.00	0.02	0	0	1784.04
														474	0.01	0.00	0.00	0	0	0								710	0.01	0.00	0.01	0	0	910.08
																												1886	0.00	0.00	0.00	0	0	NA
																												310	0.00	0.00	0.00	0	0	NA
																												815	0.01	0.00	0.00	0	0	0
																												962	0.01	0.00	0.00	0	0	0

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count

▶mmq.cpp: 303 - 2.48 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default		Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 303-330 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 651-663 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 816-818 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 825-825 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1390-1392						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 303-330 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 651-663 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 816-818 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 825-825 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1390-1392						Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 303-330 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 651-663 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 816-818 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 825-825 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1390-1392						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 303-330 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 651-663 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 816-818 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 825-825 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1390-1392						Loop Source Regions
384	0.05	0.03	0.63	90.91	38.76	0	638	0.06	0.03	0.58	90.91	38.76	0			612	0.06	0.03	0.64	90.91	38.76	0	395	0.06	0.03	0.63	90.91	38.76	0

Sum on 1 analyzed binary loop (libggml-cpu.so - 384)							Sum on 1 analyzed binary loop (libggml-cpu.so - 638)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 612)							Sum on 1 analyzed binary loop (libggml-cpu.so - 395)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis	Count	Analysis						Count	Analysis						Count	Analysis	Count
Loop Computation Issues							Loop Computation Issues									Loop Computation Issues							Loop Computation Issues
Presence of a large number of scalar integer instructions						1	Presence of a large number of scalar integer instructions						1			Presence of a large number of scalar integer instructions						1	Presence of a large number of scalar integer instructions						1
Data Access Issues							Data Access Issues									Data Access Issues							Data Access Issues
Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access						1			Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access						1
Presence of special instructions executing on a single port						1	Presence of special instructions executing on a single port						1			Presence of special instructions executing on a single port						1	Presence of special instructions executing on a single port						1
More than 20% of the loads are accessing the stack						1	More than 20% of the loads are accessing the stack						1			More than 20% of the loads are accessing the stack						1	More than 20% of the loads are accessing the stack						1
Vectorization Roadblocks							Vectorization Roadblocks									Vectorization Roadblocks							Vectorization Roadblocks
Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access						1			Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access						1
Inefficient Vectorization							Inefficient Vectorization									Inefficient Vectorization							Inefficient Vectorization
Presence of special instructions executing on a single port						1	Presence of special instructions executing on a single port						1			Presence of special instructions executing on a single port						1	Presence of special instructions executing on a single port						1

▶ops.cpp: 4325 - 1.92 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 4325-4326						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 4325-4326						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 4325-4326						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 4325-4326						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 4325-4326						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 4325-4326
1758	0.04	0.01	0.30	0	7.81	73.81	1577	0.04	0.02	0.34	0	7.81	60.44	1322	0.05	0.02	0.33	0	7.81	59.58	1664	0.04	0.02	0.34	0	7.81	60.63	1466	0.04	0.01	0.30	0	7.81	74.85	1320	0.04	0.02	0.33	0	7.81	65.2

Sum on 1 analyzed binary loop (libggml-cpu.so - 1758)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1577)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1322)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1664)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count
Loop Computation Issues							Loop Computation Issues							Loop Computation Issues							Loop Computation Issues
Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1

▶vec.cpp: 311 - 1.83 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 311-316						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 311-316						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 311-311 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14192-14192 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 15160-15160						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 311-316						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 311-316						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 311-311 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14192-14192 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 15160-15160
1232	0.04	0.02	0.32	0	0	404.79	1119	0.04	0.01	0.23	0	0	448.29	959	0.06	0.02	0.43	100	66.67	136.65	1164	0.03	0.01	0.25	0	0	519.78	1152	0.04	0.02	0.33	0	0	334.09	951	0.04	0.01	0.26	100	66.67	481.87

Sum on 1 analyzed binary loop (libggml-cpu.so - 1232)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							Sum on 1 analyzed binary loop (libggml-cpu.so - 959)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1164)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count
Control Flow Issues														Control Flow Issues							Control Flow Issues
Presence of 2 to 4 paths														Presence of 2 to 4 paths							Presence of 2 to 4 paths						1
Data Access Issues														Data Access Issues							Data Access Issues
Presence of constant non-unit stride data access														Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access
Vectorization Roadblocks														Vectorization Roadblocks							Vectorization Roadblocks
Presence of 2 to 4 paths														Presence of 2 to 4 paths						0	Presence of 2 to 4 paths						1
Presence of constant non-unit stride data access														Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access						0

▶ggml-cpu.c: 3204 - 1.53 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c: 3204-3207						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c: 3204-3207						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c: 3204-3204 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 935-935 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9376-9376 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 11216-11216						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c: 3204-3207						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c: 3204-3207						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c: 3204-3204 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 935-935 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9376-9376 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 11216-11216
5	0.04	0.01	0.31	100	100	0	6	0.04	0.01	0.25	100	100	0	5	0.03	0.01	0.26	100	83.33	0	6	0.03	0.01	0.24	100	100	0	3	0.04	0.01	0.24	100	100	1.38	3	0.04	0.01	0.23	100	83.33	0

Sum on 1 analyzed binary loop (libggml-cpu.so - 5)							Sum on 1 analyzed binary loop (libggml-cpu.so - 6)							Sum on 1 analyzed binary loop (libggml-cpu.so - 5)							Sum on 1 analyzed binary loop (libggml-cpu.so - 6)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count

▶quants.c: 298 - 1.50 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 298-347 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 353-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 346-346 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 355-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 389-404						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 298-347 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 353-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 346-346 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 355-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 389-404						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 298-298 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 304-304 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 319-321 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx2intrin.h: 79-79 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx2intrin.h: 86-86 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx2intrin.h: 1046-1046 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/xmmintrin.h: 184-184 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/xmmintrin.h: 240-240 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/xmmintrin.h: 793-793 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 354-354 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 389-404 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/pmmintrin.h: 71-71 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 186-186 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 296-296 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 320-320 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 475-475 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 526-526 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 905-905 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 935-935 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 1066-1066 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 1329-1329 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 1472-1472						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 298-347 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 353-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 346-346 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 355-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 389-404						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 298-347 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 353-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 346-346 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 355-355 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 389-404						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 298-298 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 304-304 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c: 319-321 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx2intrin.h: 79-79 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx2intrin.h: 86-86 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx2intrin.h: 1046-1046 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/xmmintrin.h: 184-184 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/xmmintrin.h: 240-240 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/xmmintrin.h: 793-793 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 354-354 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/./ggml-impl.h: 389-404 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/pmmintrin.h: 71-71 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 186-186 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 296-296 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 320-320 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 475-475 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 526-526 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 905-905 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 935-935 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 1066-1066 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avxintrin.h: 1329-1329
3307	1.13	0.01	0.25	58.33	28.75	463.99	2924	1.13	0.01	0.24	60.7	29.66	459.46	2214	1.14	0.01	0.26	60	28.75	430.09	2972	1.09	0.01	0.23	60.7	29.66	479.66	2455	1.10	0.01	0.24	60.7	29.66	469.17	2188	1.16	0.01	0.27	59.65	29.28	416.86

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							Sum on 1 analyzed binary loop (libggml-cpu.so - 2214)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count
														Loop Computation Issues
														Presence of expensive FP instructions						1
														Control Flow Issues
														Presence of more than 4 paths						1
														Data Access Issues
														More than 10% of the vector loads instructions are unaligned						1
														Presence of special instructions executing on a single port						1
														Vectorization Roadblocks
														Presence of more than 4 paths						1
														Inefficient Vectorization
														Presence of special instructions executing on a single port						1

▶vec.h: 1084 - 1.28 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default		Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 372-373 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1084-1101 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1113-1116						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 372-373 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1084-1101 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1113-1116						Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 372-373 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1084-1101 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1113-1116						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 372-373 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1084-1101 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1115-1115						Loop Source Regions
1236	0.04	0.01	0.31	98	98.13	1331.34	1126	0.04	0.01	0.30	100	100	1339.05			1171	0.04	0.02	0.32	100	100	1281.77	1158	0.04	0.02	0.35	98	98.13	1144.58

Sum on 1 analyzed binary loop (libggml-cpu.so - 1236)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1126)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 1171)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1158)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis	Count	Analysis						Count	Analysis						Count	Analysis	Count
Loop Computation Issues							Loop Computation Issues									Loop Computation Issues							Loop Computation Issues
Presence of expensive FP instructions						1	Presence of expensive FP instructions						1			Presence of expensive FP instructions						1	Presence of expensive FP instructions						1
Control Flow Issues							Control Flow Issues									Control Flow Issues							Control Flow Issues
Presence of 2 to 4 paths						1	Presence of 2 to 4 paths						1			Presence of 2 to 4 paths						1	Presence of 2 to 4 paths						1
Data Access Issues							Data Access Issues									Data Access Issues							Data Access Issues
Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access									Presence of constant non-unit stride data access							Presence of constant non-unit stride data access						1
Presence of special instructions executing on a single port						1	Presence of special instructions executing on a single port									Presence of special instructions executing on a single port							Presence of special instructions executing on a single port						1
Vectorization Roadblocks							Vectorization Roadblocks									Vectorization Roadblocks							Vectorization Roadblocks
Presence of 2 to 4 paths						1	Presence of 2 to 4 paths						1			Presence of 2 to 4 paths						1	Presence of 2 to 4 paths						1
Presence of constant non-unit stride data access						1	Presence of constant non-unit stride data access						0			Presence of constant non-unit stride data access						0	Presence of constant non-unit stride data access						1
Inefficient Vectorization							Inefficient Vectorization									Inefficient Vectorization							Inefficient Vectorization
Presence of special instructions executing on a single port						1	Presence of special instructions executing on a single port						0			Presence of special instructions executing on a single port						0	Presence of special instructions executing on a single port						1
Use of masked instructions						1	Use of masked instructions						1			Use of masked instructions						1	Use of masked instructions						1

▶ops.cpp: 6210 - 1.04 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default							Run icx_5		Run aocc_10							Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6210-6211 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6220-6230 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6238-6245						Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6210-6211 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6220-6230 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6238-6245						Loop Source Regions
				996	0.05	0.03	0.55	19.61	9.8	404.05			1617	0.05	0.02	0.49	2.27	7.95	415.12

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 996)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 1617)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis	Count	Analysis	Count	Analysis						Count	Analysis	Count	Analysis						Count	Analysis	Count
				Loop Computation Issues									Loop Computation Issues
				Presence of expensive FP instructions						1			Presence of expensive FP instructions						1
				Presence of a large number of scalar integer instructions						1			Presence of a large number of scalar integer instructions						1
				Control Flow Issues									Control Flow Issues
				Presence of calls						1			Presence of calls						1
				Presence of more than 4 paths						1			Presence of more than 4 paths						1
				Data Access Issues									Data Access Issues
				Presence of special instructions executing on a single port						1			Presence of special instructions executing on a single port						0
				More than 20% of the loads are accessing the stack						1			More than 20% of the loads are accessing the stack						1
				Vectorization Roadblocks									Vectorization Roadblocks
				Presence of calls						1			Presence of calls						1
				Presence of more than 4 paths						1			Presence of more than 4 paths						1
				Inefficient Vectorization									Inefficient Vectorization
				Presence of special instructions executing on a single port						1			Presence of special instructions executing on a single port						0
				Use of masked instructions						0			Use of masked instructions						1

▶ops.cpp: 6220 - 0.86 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default		Run icx_5							Run aocc_10		Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6220-6220 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6229-6230 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6238-6245						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6220-6222 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6229-6230 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6238-6245						Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6220-6222 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6229-6230 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6238-6245						Loop Source Regions		Loop Source Regions
2113	0.05	0.02	0.35	1.96	6.62	410.57	1836	0.04	0.01	0.25	15.56	9.17	396.71			1895	0.04	0.01	0.26	15.56	9.17	378.72

Sum on 1 analyzed binary loop (libggml-cpu.so - 2113)							Sum on 1 analyzed binary loop (libggml-cpu.so - 1836)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 1895)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis	Count	Analysis						Count	Analysis	Count	Analysis	Count
Loop Computation Issues							Loop Computation Issues									Loop Computation Issues
Presence of expensive FP instructions						1	Presence of expensive FP instructions						1			Presence of expensive FP instructions						1
Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1	Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1			Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1
Control Flow Issues							Control Flow Issues									Control Flow Issues
Presence of calls						1	Presence of calls						1			Presence of calls						1
Data Access Issues							Data Access Issues									Data Access Issues
More than 20% of the loads are accessing the stack						1	More than 20% of the loads are accessing the stack						1			More than 20% of the loads are accessing the stack						1
Vectorization Roadblocks							Vectorization Roadblocks									Vectorization Roadblocks
Presence of calls						1	Presence of calls						1			Presence of calls						1

▶ops.cpp: 8885 - 0.82 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8885-8886						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8885-8886						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8885-8886						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8885-8886						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8885-8886						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8885-8886
2419	0.02	0.01	0.12	0	6.25	0.79	2179	0.03	0.01	0.13	0	6.25	0.71	1675	0.03	0.01	0.13	0	6.25	1.19	2191	0.03	0.01	0.17	0	6.25	0.73	1777	0.02	0.01	0.14	0	6.25	1.02	1648	0.04	0.01	0.14	0	6.25	0.75

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count

▶vec.cpp: 372 - 0.74 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default							Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 372-372 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9376-9376 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9402-9402 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 10660-10660 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13204-13204 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13607-13607 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13728-13728 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13788-13788 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13848-13848 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13908-13908 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14113-14113 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14192-14192 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14544-14544 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 15536-15536 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1101-1101						Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.cpp: 372-372 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9376-9376 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9402-9402 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 10660-10660 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13204-13204 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13607-13607 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13728-13728 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13788-13788 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13848-13848 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 13908-13908 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14113-14113 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14192-14192 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 14544-14544 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 15536-15536 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/vec.h: 1101-1101
				963	0.04	0.02	0.34	100	100	1215.79					953	0.06	0.02	0.41	100	100	954.92

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 963)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 953)
Analysis	Count	Analysis	Count	Analysis						Count	Analysis	Count	Analysis	Count	Analysis						Count
				Loop Computation Issues											Loop Computation Issues
				Presence of expensive FP instructions						1					Presence of expensive FP instructions						1
				Control Flow Issues											Control Flow Issues
				Presence of 2 to 4 paths						1					Presence of 2 to 4 paths						1
				Data Access Issues											Data Access Issues
				Presence of constant non-unit stride data access						0					Presence of constant non-unit stride data access						1
				Presence of special instructions executing on a single port						1					Presence of special instructions executing on a single port						1
				Vectorization Roadblocks											Vectorization Roadblocks
				Presence of 2 to 4 paths						1					Presence of 2 to 4 paths						1
				Presence of constant non-unit stride data access						0					Presence of constant non-unit stride data access						1
				Inefficient Vectorization											Inefficient Vectorization
				Presence of special instructions executing on a single port						1					Presence of special instructions executing on a single port						1
				Use of masked instructions						1					Use of masked instructions						1

▶mmq.cpp: 822 - 0.74 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default							Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions	/cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512vnniintrin.h: 41-41 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 822-825 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1390-1392 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9491-9491						Loop Source Regions		Loop Source Regions		Loop Source Regions	/cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512vnniintrin.h: 41-41 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 822-825 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 1390-1392 /cluster/comp/gcc/14.2.0/lib/gcc/x86_64-pc-linux-gnu/14.2.0/include/avx512fintrin.h: 9491-9491
				455	0.05	0.02	0.36	78.57	69.64	0					409	0.04	0.02	0.38	29.73	31.25	0

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 455)							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 409)
Analysis	Count	Analysis	Count	Analysis						Count	Analysis	Count	Analysis	Count	Analysis						Count
				Loop Computation Issues											Loop Computation Issues
				Presence of a large number of scalar integer instructions						1					Presence of a large number of scalar integer instructions						1
				Control Flow Issues											Control Flow Issues
				Presence of calls						1					Presence of calls						1
				Data Access Issues											Data Access Issues
				Presence of constant non-unit stride data access						1					Presence of constant non-unit stride data access						1
				Vectorization Roadblocks											Vectorization Roadblocks
				Presence of calls						1					Presence of calls						1
				Presence of constant non-unit stride data access						1					Presence of constant non-unit stride data access						1

▶binary-ops.cpp: 10 - 0.69 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default							Run icx_default							Run gcc_default							Run icx_5							Run aocc_10							Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 10-10 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 10-10 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 10-10 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 10-10 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 10-10 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32						Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 10-10 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp: 31-32
816	0.02	0.01	0.12	0	6.25	83.68	698	0.02	0.01	0.12	0	6.25	97.17	634	0.01	0.01	0.11	100	50	76.68	632	0.02	0.01	0.11	0	6.25	101.67	837	0.02	0.01	0.12	0	6.25	89.02	633	0.02	0.01	0.12	0	6.25	88.52

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count	Analysis						Count

▶ops.cpp: 6238 - 0.43 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default		Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 6238-6245
										965	0.05	0.02	0.43	6.25	8.2	287.26

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		Sum on 1 analyzed binary loop (libggml-cpu.so - 965)
Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis						Count
										Loop Computation Issues
										Presence of expensive FP instructions						1
										Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA						1
										Presence of a large number of scalar integer instructions						1
										Control Flow Issues
										Presence of calls						1
										Data Access Issues
										More than 20% of the loads are accessing the stack						1
										Vectorization Roadblocks
										Presence of calls						1
										Presence of more than 4 paths						1

▶mmq.cpp: 2096 - 0.26 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default		Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2096-2101
										589	0.05	0.01	0.26	25	18.13	2805.05

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis						Count

▶ops.cpp: 8817 - 0.25 %

ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default		Run icx_default		Run gcc_default							Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions		Loop Source Regions		Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h: 130-130 /beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/ops.cpp: 8817-8819						Loop Source Regions		Loop Source Regions		Loop Source Regions
				1673	0.03	0.01	0.25	0	9.38	94.18

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis	Count	Analysis	Count	Analysis						Count	Analysis	Count	Analysis	Count	Analysis	Count

▶mmq.cpp: 2068 - 0.23 %

ASM Loop ID	Max Time Over Threads (s)	Time w.r.t. Wall Time (s)	Cov (%)	Vect. Ratio (%)	Vector Length Use (%)	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s	ASM Loop ID	GFLOP/s
Run orig_default							Run icx_default		Run gcc_default		Run icx_5		Run aocc_10		Run gcc_6
Loop Source Regions	/beegfs/hackathon/users/eoseret/qaas_runs_test/isix06.benchmarkcenter.megware.com/176-400-1862/llama.cpp/build/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp: 2068-2078						Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions		Loop Source Regions
710	0.04	0.01	0.23	0	0	28.03

No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.							No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.		No Optimizer analysis found for any assembly loop. More loops can be analyzed using option --optimizer-loop-count.
Analysis						Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count	Analysis	Count

Report Configuration

Loops

▶mmq.cpp: 1138 - 34.48 %

▶simd-mappings.h: 130 - 9.01 %

▶mmq.cpp: 1140 - 8.86 %

▶binary-ops.cpp: 18 - 2.77 %

▶vec.h: 491 - 2.74 %

▶<unknown>: 0 - 2.60 %

▶mmq.cpp: 303 - 2.48 %

▶ops.cpp: 4325 - 1.92 %

▶vec.cpp: 311 - 1.83 %

▶ggml-cpu.c: 3204 - 1.53 %

▶quants.c: 298 - 1.50 %

▶vec.h: 1084 - 1.28 %

▶ops.cpp: 6210 - 1.04 %

▶ops.cpp: 6220 - 0.86 %

▶ops.cpp: 8885 - 0.82 %

▶vec.cpp: 372 - 0.74 %

▶mmq.cpp: 822 - 0.74 %

▶binary-ops.cpp: 10 - 0.69 %

▶ops.cpp: 6238 - 0.43 %

▶mmq.cpp: 2096 - 0.26 %

▶ops.cpp: 8817 - 0.25 %

▶mmq.cpp: 2068 - 0.23 %