| Loop Id: 2 | Module: kmeans-gcc-O3-funroll | Source: main.cpp:60-67 | Coverage: 5.89% |
|---|
| Loop Id: 2 | Module: kmeans-gcc-O3-funroll | Source: main.cpp:60-67 | Coverage: 5.89% |
|---|
0x2a40 MOV %R8D,%ECX |
0x2a43 VMOVSD (%RDI),%XMM3 |
0x2a47 VMOVSD 0x8(%RDI),%XMM2 |
0x2a4c MOV %R9,%RBX |
0x2a4f XOR %EDX,%EDX |
0x2a51 VMOVAPD %XMM5,%XMM0 |
0x2a55 AND $0x3,%ECX |
0x2a58 JE 2ae9 |
0x2a5e CMP $0x1,%ECX |
0x2a61 JE 2aba |
0x2a63 CMP $0x2,%ECX |
0x2a66 JE 2a95 |
0x2a68 VSUBSD 0x8(%R9),%XMM2,%XMM4 |
0x2a6e VSUBSD (%R9),%XMM3,%XMM1 |
0x2a73 VMULSD %XMM4,%XMM4,%XMM6 |
0x2a77 VFMADD132SD %XMM1,%XMM6,%XMM1 |
0x2a7c VCOMISD %XMM1,%XMM5 |
0x2a80 JBE 2a8c |
0x2a82 MOVL $0,(%RSI) |
0x2a88 VMOVAPD %XMM1,%XMM0 |
0x2a8c MOV $0x1,%EDX |
0x2a91 LEA 0x10(%R9),%RBX |
0x2a95 VSUBSD 0x8(%RBX),%XMM2,%XMM8 |
0x2a9a VSUBSD (%RBX),%XMM3,%XMM7 |
0x2a9e VMULSD %XMM8,%XMM8,%XMM9 |
0x2aa3 VFMADD132SD %XMM7,%XMM9,%XMM7 |
0x2aa8 VCOMISD %XMM7,%XMM0 |
0x2aac JBE 2ab4 |
0x2aae MOV %EDX,(%RSI) |
0x2ab0 VMOVAPD %XMM7,%XMM0 |
0x2ab4 INC %EDX |
0x2ab6 ADD $0x10,%RBX |
0x2aba VSUBSD 0x8(%RBX),%XMM2,%XMM11 |
0x2abf VSUBSD (%RBX),%XMM3,%XMM10 |
0x2ac3 VMULSD %XMM11,%XMM11,%XMM12 |
0x2ac8 VFMADD132SD %XMM10,%XMM12,%XMM10 |
0x2acd VCOMISD %XMM10,%XMM0 |
0x2ad2 JBE 2ada |
0x2ad4 MOV %EDX,(%RSI) |
0x2ad6 VMOVAPD %XMM10,%XMM0 |
0x2ada INC %EDX |
0x2adc ADD $0x10,%RBX |
0x2ae0 CMP %EDX,%R8D |
0x2ae3 JE 2b8b |
(1) 0x2ae9 VSUBSD 0x8(%RBX),%XMM2,%XMM14 |
(1) 0x2aee VSUBSD (%RBX),%XMM3,%XMM13 |
(1) 0x2af2 VMULSD %XMM14,%XMM14,%XMM15 |
(1) 0x2af7 VFMADD132SD %XMM13,%XMM15,%XMM13 |
(1) 0x2afc VCOMISD %XMM13,%XMM0 |
(1) 0x2b01 JBE 2b09 |
(1) 0x2b03 MOV %EDX,(%RSI) |
(1) 0x2b05 VMOVAPD %XMM13,%XMM0 |
(1) 0x2b09 VSUBSD 0x18(%RBX),%XMM2,%XMM4 |
(1) 0x2b0e VSUBSD 0x10(%RBX),%XMM3,%XMM1 |
(1) 0x2b13 INC %EDX |
(1) 0x2b15 LEA 0x10(%RBX),%R11 |
(1) 0x2b19 VMULSD %XMM4,%XMM4,%XMM6 |
(1) 0x2b1d VFMADD132SD %XMM1,%XMM6,%XMM1 |
(1) 0x2b22 VCOMISD %XMM1,%XMM0 |
(1) 0x2b26 JBE 2b2e |
(1) 0x2b28 MOV %EDX,(%RSI) |
(1) 0x2b2a VMOVAPD %XMM1,%XMM0 |
(1) 0x2b2e VSUBSD 0x18(%R11),%XMM2,%XMM8 |
(1) 0x2b34 VSUBSD 0x10(%R11),%XMM3,%XMM7 |
(1) 0x2b3a LEA 0x1(%RDX),%EAX |
(1) 0x2b3d VMULSD %XMM8,%XMM8,%XMM9 |
(1) 0x2b42 VFMADD132SD %XMM7,%XMM9,%XMM7 |
(1) 0x2b47 VCOMISD %XMM7,%XMM0 |
(1) 0x2b4b JBE 2b53 |
(1) 0x2b4d MOV %EAX,(%RSI) |
(1) 0x2b4f VMOVAPD %XMM7,%XMM0 |
(1) 0x2b53 VSUBSD 0x28(%R11),%XMM2,%XMM11 |
(1) 0x2b59 VSUBSD 0x20(%R11),%XMM3,%XMM10 |
(1) 0x2b5f LEA 0x2(%RDX),%R12D |
(1) 0x2b63 VMULSD %XMM11,%XMM11,%XMM12 |
(1) 0x2b68 VFMADD132SD %XMM10,%XMM12,%XMM10 |
(1) 0x2b6d VCOMISD %XMM10,%XMM0 |
(1) 0x2b72 JBE 2b7b |
(1) 0x2b74 MOV %R12D,(%RSI) |
(1) 0x2b77 VMOVAPD %XMM10,%XMM0 |
(1) 0x2b7b ADD $0x3,%EDX |
(1) 0x2b7e LEA 0x30(%R11),%RBX |
(1) 0x2b82 CMP %EDX,%R8D |
(1) 0x2b85 JNE 2ae9 |
0x2b8b ADD $0x10,%RDI |
0x2b8f ADD $0x4,%RSI |
0x2b93 CMP %RDI,%R10 |
0x2b96 JNE 2a40 |
/home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp: 60 - 67 |
-------------------------------------------------------------------------------- |
60: double optimal_dist = DBL_MAX; |
61: for (int j = 0; j < k; ++j) { |
62: double dist = |
63: (points[i].x - centroids[j].x) * (points[i].x - centroids[j].x) + |
64: (points[i].y - centroids[j].y) * (points[i].y - centroids[j].y); |
65: if (dist < optimal_dist) { |
66: optimal_dist = dist; |
67: assignment[i] = j; |
| Coverage (%) | Name | Source Location | Module |
|---|---|---|---|
| ►98.38+ | GOMP_parallel | parallel.c:140 | libgomp.so.1.0.0 |
| ○ | k_means(int, point_t*, point_t[...] | main.cpp:73 | kmeans-gcc-O3-funroll |
| ○ | main | main.cpp:20 | kmeans-gcc-O3-funroll |
| ○ | __libc_init_first | libc.so.6 | |
| ○ | __libc_start_main | libc.so.6 | |
| ○ | _start | basic_string.h:809 | kmeans-gcc-O3-funroll |
| Coverage (%) | Name | Source Location | Module |
|---|---|---|---|
| ►48.98+ | gomp_thread_start | team.c:130 | libgomp.so.1.0.0 |
| ○ | pthread_condattr_setpshared | libc.so.6 | |
| ○ | __clone | libc.so.6 | |
| ►48.55+ | GOMP_parallel | parallel.c:140 | libgomp.so.1.0.0 |
| ○ | k_means(int, point_t*, point_t[...] | main.cpp:73 | kmeans-gcc-O3-funroll |
| ○ | main | main.cpp:20 | kmeans-gcc-O3-funroll |
| ○ | __libc_init_first | libc.so.6 | |
| ○ | __libc_start_main | libc.so.6 | |
| ○ | _start | basic_string.h:809 | kmeans-gcc-O3-funroll |
| Coverage (%) | Name | Source Location | Module |
|---|---|---|---|
| ►69.80+ | gomp_thread_start | team.c:130 | libgomp.so.1.0.0 |
| ○ | pthread_condattr_setpshared | libc.so.6 | |
| ○ | __clone | libc.so.6 | |
| ►22.13+ | GOMP_parallel | parallel.c:140 | libgomp.so.1.0.0 |
| ○ | k_means(int, point_t*, point_t[...] | main.cpp:73 | kmeans-gcc-O3-funroll |
| ○ | main | main.cpp:20 | kmeans-gcc-O3-funroll |
| ○ | __libc_init_first | libc.so.6 | |
| ○ | __libc_start_main | libc.so.6 | |
| ○ | _start | basic_string.h:809 | kmeans-gcc-O3-funroll |
| Coverage (%) | Name | Source Location | Module |
|---|---|---|---|
| ►87.00+ | gomp_thread_start | team.c:130 | libgomp.so.1.0.0 |
| ○ | pthread_condattr_setpshared | libc.so.6 | |
| ○ | __clone | libc.so.6 | |
| ►12.13+ | GOMP_parallel | parallel.c:140 | libgomp.so.1.0.0 |
| ○ | k_means(int, point_t*, point_t[...] | main.cpp:73 | kmeans-gcc-O3-funroll |
| ○ | main | main.cpp:20 | kmeans-gcc-O3-funroll |
| ○ | __libc_init_first | libc.so.6 | |
| ○ | __libc_start_main | libc.so.6 | |
| ○ | _start | basic_string.h:809 | kmeans-gcc-O3-funroll |
| Coverage (%) | Name | Source Location | Module |
|---|---|---|---|
| ►93.98+ | gomp_thread_start | team.c:130 | libgomp.so.1.0.0 |
| ○ | pthread_condattr_setpshared | libc.so.6 | |
| ○ | __clone | libc.so.6 | |
| ►5.77+ | GOMP_parallel | parallel.c:140 | libgomp.so.1.0.0 |
| ○ | k_means(int, point_t*, point_t[...] | main.cpp:73 | kmeans-gcc-O3-funroll |
| ○ | main | main.cpp:20 | kmeans-gcc-O3-funroll |
| ○ | __libc_init_first | libc.so.6 | |
| ○ | __libc_start_main | libc.so.6 | |
| ○ | _start | basic_string.h:809 | kmeans-gcc-O3-funroll |
| Coverage (%) | Name | Source Location | Module |
|---|---|---|---|
| ►95.94+ | gomp_thread_start | team.c:130 | libgomp.so.1.0.0 |
| ○ | pthread_condattr_setpshared | libc.so.6 | |
| ○ | __clone | libc.so.6 | |
| ►3.96+ | GOMP_parallel | parallel.c:140 | libgomp.so.1.0.0 |
| ○ | k_means(int, point_t*, point_t[...] | main.cpp:73 | kmeans-gcc-O3-funroll |
| ○ | main | main.cpp:20 | kmeans-gcc-O3-funroll |
| ○ | __libc_init_first | libc.so.6 | |
| ○ | __libc_start_main | libc.so.6 | |
| ○ | _start | basic_string.h:809 | kmeans-gcc-O3-funroll |
| min | med | avg | max |
|---|---|---|---|
| Percentile Index | 10 | 20 | 30 | 40 | 50 | 60 | 70 | 80 | 90 | 100 |
|---|---|---|---|---|---|---|---|---|---|---|
| Value |
| min | med | avg | max |
|---|---|---|---|
| Percentile Index | 10 | 20 | 30 | 40 | 50 | 60 | 70 | 80 | 90 | 100 |
|---|---|---|---|---|---|---|---|---|---|---|
| Value |
| Path / |
| Metric | Value |
|---|---|
| CQA speedup if no scalar integer | 1.47 |
| CQA speedup if FP arith vectorized | 2.30 |
| CQA speedup if fully vectorized | 11.82 |
| CQA speedup if no inter-iteration dependency | NA |
| CQA speedup if next bottleneck killed | 1.34 |
| Bottlenecks | micro-operation queue, |
| Function | k_means(int, point_t*, point_t*, int*, point_t*, int, int) [clone ._omp_fn.0] |
| Source | main.cpp:60-67 |
| Source loop unroll info | NA |
| Source loop unroll confidence level | NA |
| Unroll/vectorization loop type | NA |
| Unroll factor | NA |
| CQA cycles | 11.75 |
| CQA cycles if no scalar integer | 8.00 |
| CQA cycles if FP arith vectorized | 5.11 |
| CQA cycles if fully vectorized | 0.99 |
| Front-end cycles | 11.75 |
| P0 cycles | 8.75 |
| P1 cycles | 8.75 |
| P2 cycles | 4.00 |
| P3 cycles | 4.00 |
| P4 cycles | 3.00 |
| P5 cycles | 8.75 |
| P6 cycles | 8.75 |
| P7 cycles | 3.00 |
| DIV/SQRT cycles | 0.00 |
| Inter-iter dependencies cycles | NA |
| FE+BE cycles (UFS) | 12.02 |
| Stall cycles (UFS) | 0.00 |
| Nb insns | 48.00 |
| Nb uops | 47.00 |
| Nb loads | 8.00 |
| Nb stores | 3.00 |
| Nb stack references | 0.00 |
| FLOP/cycle | 1.28 |
| Nb FLOP add-sub | 6.00 |
| Nb FLOP mul | 3.00 |
| Nb FLOP fma | 3.00 |
| Nb FLOP div | 0.00 |
| Nb FLOP rcp | 0.00 |
| Nb FLOP sqrt | 0.00 |
| Nb FLOP rsqrt | 0.00 |
| Bytes/cycle | 6.47 |
| Bytes prefetched | 0.00 |
| Bytes loaded | 64.00 |
| Bytes stored | 12.00 |
| Stride 0 | NA |
| Stride 1 | NA |
| Stride n | NA |
| Stride unknown | NA |
| Stride indirect | NA |
| Vectorization ratio all | 12.12 |
| Vectorization ratio load | 0.00 |
| Vectorization ratio store | 0.00 |
| Vectorization ratio mul | 0.00 |
| Vectorization ratio add_sub | 0.00 |
| Vectorization ratio fma | 0.00 |
| Vectorization ratio div_sqrt | NA |
| Vectorization ratio other | 28.57 |
| Vector-efficiency ratio all | 11.74 |
| Vector-efficiency ratio load | 12.50 |
| Vector-efficiency ratio store | 6.25 |
| Vector-efficiency ratio mul | 12.50 |
| Vector-efficiency ratio add_sub | 10.94 |
| Vector-efficiency ratio fma | 12.50 |
| Vector-efficiency ratio div_sqrt | NA |
| Vector-efficiency ratio other | 12.95 |
| Metric | Value |
|---|---|
| CQA speedup if no scalar integer | 1.47 |
| CQA speedup if FP arith vectorized | 2.30 |
| CQA speedup if fully vectorized | 11.82 |
| CQA speedup if no inter-iteration dependency | NA |
| CQA speedup if next bottleneck killed | 1.34 |
| Bottlenecks | micro-operation queue, |
| Function | k_means(int, point_t*, point_t*, int*, point_t*, int, int) [clone ._omp_fn.0] |
| Source | main.cpp:60-67 |
| Source loop unroll info | NA |
| Source loop unroll confidence level | NA |
| Unroll/vectorization loop type | NA |
| Unroll factor | NA |
| CQA cycles | 11.75 |
| CQA cycles if no scalar integer | 8.00 |
| CQA cycles if FP arith vectorized | 5.11 |
| CQA cycles if fully vectorized | 0.99 |
| Front-end cycles | 11.75 |
| P0 cycles | 8.75 |
| P1 cycles | 8.75 |
| P2 cycles | 4.00 |
| P3 cycles | 4.00 |
| P4 cycles | 3.00 |
| P5 cycles | 8.75 |
| P6 cycles | 8.75 |
| P7 cycles | 3.00 |
| DIV/SQRT cycles | 0.00 |
| Inter-iter dependencies cycles | NA |
| FE+BE cycles (UFS) | 12.02 |
| Stall cycles (UFS) | 0.00 |
| Nb insns | 48.00 |
| Nb uops | 47.00 |
| Nb loads | 8.00 |
| Nb stores | 3.00 |
| Nb stack references | 0.00 |
| FLOP/cycle | 1.28 |
| Nb FLOP add-sub | 6.00 |
| Nb FLOP mul | 3.00 |
| Nb FLOP fma | 3.00 |
| Nb FLOP div | 0.00 |
| Nb FLOP rcp | 0.00 |
| Nb FLOP sqrt | 0.00 |
| Nb FLOP rsqrt | 0.00 |
| Bytes/cycle | 6.47 |
| Bytes prefetched | 0.00 |
| Bytes loaded | 64.00 |
| Bytes stored | 12.00 |
| Stride 0 | NA |
| Stride 1 | NA |
| Stride n | NA |
| Stride unknown | NA |
| Stride indirect | NA |
| Vectorization ratio all | 12.12 |
| Vectorization ratio load | 0.00 |
| Vectorization ratio store | 0.00 |
| Vectorization ratio mul | 0.00 |
| Vectorization ratio add_sub | 0.00 |
| Vectorization ratio fma | 0.00 |
| Vectorization ratio div_sqrt | NA |
| Vectorization ratio other | 28.57 |
| Vector-efficiency ratio all | 11.74 |
| Vector-efficiency ratio load | 12.50 |
| Vector-efficiency ratio store | 6.25 |
| Vector-efficiency ratio mul | 12.50 |
| Vector-efficiency ratio add_sub | 10.94 |
| Vector-efficiency ratio fma | 12.50 |
| Vector-efficiency ratio div_sqrt | NA |
| Vector-efficiency ratio other | 12.95 |
| Path / |
| Function | k_means(int, point_t*, point_t*, int*, point_t*, int, int) [clone ._omp_fn.0] |
| Source file and lines | main.cpp:60-67 |
| Module | kmeans-gcc-O3-funroll |
| nb instructions | 48 |
| nb uops | 47 |
| loop length | 186 |
| used x86 registers | 8 |
| used mmx registers | 0 |
| used xmm registers | 13 |
| used ymm registers | 0 |
| used zmm registers | 0 |
| nb stack references | 0 |
| ADD-SUB / MUL ratio | 2.00 |
| micro-operation queue | 11.75 cycles |
| front end | 11.75 cycles |
| P0 | P1 | P2 | P3 | P4 | P5 | P6 | P7 | |
|---|---|---|---|---|---|---|---|---|
| uops | 8.75 | 8.75 | 4.00 | 4.00 | 3.00 | 8.75 | 8.75 | 3.00 |
| cycles | 8.75 | 8.75 | 4.00 | 4.00 | 3.00 | 8.75 | 8.75 | 3.00 |
| Cycles executing div or sqrt instructions | NA |
| FE+BE cycles | 12.02 |
| Stall cycles | 0.00 |
| Front-end | 11.75 |
| Dispatch | 8.75 |
| Overall L1 | 11.75 |
| all | 0% |
| load | NA (no load vectorizable/vectorized instructions) |
| store | 0% |
| mul | NA (no mul vectorizable/vectorized instructions) |
| add-sub | 0% |
| fma | NA (no fma vectorizable/vectorized instructions) |
| other | 0% |
| all | 19% |
| load | 0% |
| store | NA (no store vectorizable/vectorized instructions) |
| mul | 0% |
| add-sub | 0% |
| fma | 0% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 57% |
| all | 12% |
| load | 0% |
| store | 0% |
| mul | 0% |
| add-sub | 0% |
| fma | 0% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 28% |
| all | 6% |
| load | NA (no load vectorizable/vectorized instructions) |
| store | 6% |
| mul | NA (no mul vectorizable/vectorized instructions) |
| add-sub | 6% |
| fma | NA (no fma vectorizable/vectorized instructions) |
| other | 6% |
| all | 14% |
| load | 12% |
| store | NA (no store vectorizable/vectorized instructions) |
| mul | 12% |
| add-sub | 12% |
| fma | 12% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 19% |
| all | 11% |
| load | 12% |
| store | 6% |
| mul | 12% |
| add-sub | 10% |
| fma | 12% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 12% |
| Instruction | Nb FU | P0 | P1 | P2 | P3 | P4 | P5 | P6 | P7 | Latency | Recip. throughput | Vectorization |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MOV %R8D,%ECX | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | scal (6.3%) |
| VMOVSD (%RDI),%XMM3 | 1 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4-5 | 0.50 | scal (12.5%) |
| VMOVSD 0x8(%RDI),%XMM2 | 1 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4-5 | 0.50 | scal (12.5%) |
| MOV %R9,%RBX | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | N/A |
| XOR %EDX,%EDX | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | scal (6.3%) |
| VMOVAPD %XMM5,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| AND $0x3,%ECX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2ae9 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x129> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| CMP $0x1,%ECX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2aba <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xfa> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| CMP $0x2,%ECX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2a95 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xd5> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| VSUBSD 0x8(%R9),%XMM2,%XMM4 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VSUBSD (%R9),%XMM3,%XMM1 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VMULSD %XMM4,%XMM4,%XMM6 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VFMADD132SD %XMM1,%XMM6,%XMM1 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VCOMISD %XMM1,%XMM5 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 1 | scal (12.5%) |
| JBE 2a8c <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xcc> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| MOVL $0,(%RSI) | 1 | 0 | 0 | 0.33 | 0.33 | 1 | 0 | 0 | 0.33 | 2 | 1 | scal (6.3%) |
| VMOVAPD %XMM1,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| MOV $0x1,%EDX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| LEA 0x10(%R9),%RBX | 1 | 0 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 1 | 0.50 | N/A |
| VSUBSD 0x8(%RBX),%XMM2,%XMM8 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VSUBSD (%RBX),%XMM3,%XMM7 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VMULSD %XMM8,%XMM8,%XMM9 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VFMADD132SD %XMM7,%XMM9,%XMM7 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VCOMISD %XMM7,%XMM0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 1 | scal (12.5%) |
| JBE 2ab4 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xf4> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| MOV %EDX,(%RSI) | 1 | 0 | 0 | 0.33 | 0.33 | 1 | 0 | 0 | 0.33 | 3 | 1 | scal (6.3%) |
| VMOVAPD %XMM7,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| INC %EDX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| ADD $0x10,%RBX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| VSUBSD 0x8(%RBX),%XMM2,%XMM11 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VSUBSD (%RBX),%XMM3,%XMM10 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VMULSD %XMM11,%XMM11,%XMM12 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VFMADD132SD %XMM10,%XMM12,%XMM10 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VCOMISD %XMM10,%XMM0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 1 | scal (12.5%) |
| JBE 2ada <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x11a> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| MOV %EDX,(%RSI) | 1 | 0 | 0 | 0.33 | 0.33 | 1 | 0 | 0 | 0.33 | 3 | 1 | scal (6.3%) |
| VMOVAPD %XMM10,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| INC %EDX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| ADD $0x10,%RBX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| CMP %EDX,%R8D | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2b8b <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x1cb> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| ADD $0x10,%RDI | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| ADD $0x4,%RSI | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| CMP %RDI,%R10 | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| JNE 2a40 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x80> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| Function | k_means(int, point_t*, point_t*, int*, point_t*, int, int) [clone ._omp_fn.0] |
| Source file and lines | main.cpp:60-67 |
| Module | kmeans-gcc-O3-funroll |
| nb instructions | 48 |
| nb uops | 47 |
| loop length | 186 |
| used x86 registers | 8 |
| used mmx registers | 0 |
| used xmm registers | 13 |
| used ymm registers | 0 |
| used zmm registers | 0 |
| nb stack references | 0 |
| ADD-SUB / MUL ratio | 2.00 |
| micro-operation queue | 11.75 cycles |
| front end | 11.75 cycles |
| P0 | P1 | P2 | P3 | P4 | P5 | P6 | P7 | |
|---|---|---|---|---|---|---|---|---|
| uops | 8.75 | 8.75 | 4.00 | 4.00 | 3.00 | 8.75 | 8.75 | 3.00 |
| cycles | 8.75 | 8.75 | 4.00 | 4.00 | 3.00 | 8.75 | 8.75 | 3.00 |
| Cycles executing div or sqrt instructions | NA |
| FE+BE cycles | 12.02 |
| Stall cycles | 0.00 |
| Front-end | 11.75 |
| Dispatch | 8.75 |
| Overall L1 | 11.75 |
| all | 0% |
| load | NA (no load vectorizable/vectorized instructions) |
| store | 0% |
| mul | NA (no mul vectorizable/vectorized instructions) |
| add-sub | 0% |
| fma | NA (no fma vectorizable/vectorized instructions) |
| other | 0% |
| all | 19% |
| load | 0% |
| store | NA (no store vectorizable/vectorized instructions) |
| mul | 0% |
| add-sub | 0% |
| fma | 0% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 57% |
| all | 12% |
| load | 0% |
| store | 0% |
| mul | 0% |
| add-sub | 0% |
| fma | 0% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 28% |
| all | 6% |
| load | NA (no load vectorizable/vectorized instructions) |
| store | 6% |
| mul | NA (no mul vectorizable/vectorized instructions) |
| add-sub | 6% |
| fma | NA (no fma vectorizable/vectorized instructions) |
| other | 6% |
| all | 14% |
| load | 12% |
| store | NA (no store vectorizable/vectorized instructions) |
| mul | 12% |
| add-sub | 12% |
| fma | 12% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 19% |
| all | 11% |
| load | 12% |
| store | 6% |
| mul | 12% |
| add-sub | 10% |
| fma | 12% |
| div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
| other | 12% |
| Instruction | Nb FU | P0 | P1 | P2 | P3 | P4 | P5 | P6 | P7 | Latency | Recip. throughput | Vectorization |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MOV %R8D,%ECX | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | scal (6.3%) |
| VMOVSD (%RDI),%XMM3 | 1 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4-5 | 0.50 | scal (12.5%) |
| VMOVSD 0x8(%RDI),%XMM2 | 1 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4-5 | 0.50 | scal (12.5%) |
| MOV %R9,%RBX | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | N/A |
| XOR %EDX,%EDX | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | scal (6.3%) |
| VMOVAPD %XMM5,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| AND $0x3,%ECX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2ae9 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x129> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| CMP $0x1,%ECX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2aba <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xfa> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| CMP $0x2,%ECX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2a95 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xd5> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| VSUBSD 0x8(%R9),%XMM2,%XMM4 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VSUBSD (%R9),%XMM3,%XMM1 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VMULSD %XMM4,%XMM4,%XMM6 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VFMADD132SD %XMM1,%XMM6,%XMM1 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VCOMISD %XMM1,%XMM5 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 1 | scal (12.5%) |
| JBE 2a8c <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xcc> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| MOVL $0,(%RSI) | 1 | 0 | 0 | 0.33 | 0.33 | 1 | 0 | 0 | 0.33 | 2 | 1 | scal (6.3%) |
| VMOVAPD %XMM1,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| MOV $0x1,%EDX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| LEA 0x10(%R9),%RBX | 1 | 0 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 1 | 0.50 | N/A |
| VSUBSD 0x8(%RBX),%XMM2,%XMM8 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VSUBSD (%RBX),%XMM3,%XMM7 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VMULSD %XMM8,%XMM8,%XMM9 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VFMADD132SD %XMM7,%XMM9,%XMM7 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VCOMISD %XMM7,%XMM0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 1 | scal (12.5%) |
| JBE 2ab4 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0xf4> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| MOV %EDX,(%RSI) | 1 | 0 | 0 | 0.33 | 0.33 | 1 | 0 | 0 | 0.33 | 3 | 1 | scal (6.3%) |
| VMOVAPD %XMM7,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| INC %EDX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| ADD $0x10,%RBX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| VSUBSD 0x8(%RBX),%XMM2,%XMM11 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VSUBSD (%RBX),%XMM3,%XMM10 | 1 | 0.50 | 0.50 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VMULSD %XMM11,%XMM11,%XMM12 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VFMADD132SD %XMM10,%XMM12,%XMM10 | 1 | 0.50 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (12.5%) |
| VCOMISD %XMM10,%XMM0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 1 | scal (12.5%) |
| JBE 2ada <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x11a> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| MOV %EDX,(%RSI) | 1 | 0 | 0 | 0.33 | 0.33 | 1 | 0 | 0 | 0.33 | 3 | 1 | scal (6.3%) |
| VMOVAPD %XMM10,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.25 | vect (25.0%) |
| INC %EDX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| ADD $0x10,%RBX | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| CMP %EDX,%R8D | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | scal (6.3%) |
| JE 2b8b <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x1cb> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| ADD $0x10,%RDI | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| ADD $0x4,%RSI | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| CMP %RDI,%R10 | 1 | 0.25 | 0.25 | 0 | 0 | 0 | 0.25 | 0.25 | 0 | 1 | 0.25 | N/A |
| JNE 2a40 <_Z7k_meansiP7point_tS0_PiS0_ii._omp_fn.0+0x80> | 1 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0.50-1 | N/A |
| Run run_1_thread | Number processes: 1Number nodes: 1Number processes per node: 1Run Command: <executable> input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 1 |
|---|---|
| Run run_2_threads | Number processes: 1Number nodes: 1Number processes per node: 1Run Command: <executable> input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 2 |
| Run run_4_threads | Number processes: 1Number nodes: 1Number processes per node: 1Run Command: <executable> input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 4 |
| Run run_8_threads | Number processes: 1Number nodes: 1Number processes per node: 1Run Command: <executable> input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 8 |
| Run run_16_threads | Number processes: 1Number nodes: 1Number processes per node: 1Run Command: <executable> input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 16 |
| Run run_26_threads | Number processes: 1Number nodes: 1Number processes per node: 1Run Command: <executable> input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 26 |
| (run_1_thread) Efficiency | (run_1_thread) Potential Speed-Up (%) | (run_2_threads) Efficiency | (run_2_threads) Potential Speed-Up (%) | (run_4_threads) Efficiency | (run_4_threads) Potential Speed-Up (%) | (run_8_threads) Efficiency | (run_8_threads) Potential Speed-Up (%) | (run_16_threads) Efficiency | (run_16_threads) Potential Speed-Up (%) | (run_26_threads) Efficiency | (run_26_threads) Potential Speed-Up (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 0 | 0.97 | 0.2 | 0.89 | 0.63 | 0.69 | 1.83 | 0.51 | 2.78 | 0.39 | 3.45 |
| Run | Number of threads | Efficiency (ideal is 1) | Speedup | Ideal Speedup | Time (s) | Coverage (%) |
|---|---|---|---|---|---|---|
| run_1_thread | 1 | 1 | 1 | 1 | 9.5850009918213 | 5.8888592720032 |
| run_2_threads | 2 | 0.97 | 1.93 | 2 | 4.6899995803833 | 5.7172589302063 |
| run_4_threads | 4 | 0.89 | 3.55 | 4 | 2.5099997520447 | 5.5329370498657 |
| run_8_threads | 8 | 0.69 | 5.5 | 8 | 1.3250000476837 | 5.8356084823608 |
| run_16_threads | 16 | 0.51 | 8.22 | 16 | 0.67000007629395 | 5.7235517501831 |
| run_26_threads | 26 | 0.39 | 10.14 | 26 | 0.46500009298325 | 5.6526303291321 |
