derek/TheBandwidthBenchmark
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #4 from Sholde/icx

Browse Source 
Add icx target
This commit is contained in:
Jan Eitzinger
2022-12-12 11:37:49 +01:00
committed by GitHub
parent 91896f1cc0 c894815ae7
commit 0bb56ef09e
3 changed files with 30 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
README.md
View File

@@ -34,7 +34,7 @@ As added benefit the code is a blueprint for a minimal benchmarking application
1. Configure the toolchain and additional options in `config.mk`:
```
# Supported: GCC, CLANG, ICC
# Supported: GCC, CLANG, ICC, ICX
TAG ?= GCC
ENABLE_OPENMP ?= false
ENABLE_LIKWID ?= false
@@ -49,7 +49,7 @@ OPTIONS += -DARRAY_ALIGNMENT=64
The verbosity options enable detailed output about affinity settings, allocation sizes and timer resolution.
*Notice:* OpenMP involves significant overhead through barrier cost, especially on systems with many memory domains. The default problem size is set to almost 4GB to have enough work vs overhead. If you suspect that the result should be better you may try to further increase the problem size. To compare to original stream results on X86 systems you have to ensure that streaming store instructions are used. For the ICC toolchain this is now the default (Option `-qopt-streaming-stores=always`).
*Notice:* OpenMP involves significant overhead through barrier cost, especially on systems with many memory domains. The default problem size is set to almost 4GB to have enough work vs overhead. If you suspect that the result should be better you may try to further increase the problem size. To compare to original stream results on X86 systems you have to ensure that streaming store instructions are used. For the ICC toolchain this is now the default (Option `-qopt-streaming-stores=always`).
2. Build with:
```
@@ -149,13 +149,13 @@ To extract the results and output in a plottable format execute:
The script will pick up all result files in the directory specified and create a column format output file.
In this case:
```
#nt Init Sum Copy Update Triad Daxpy STriad SDaxpy
1 4109 11900 5637 8025 7407 9874 8981 11288
2 8057 22696 11011 15174 14821 18786 17599 21475
4 15602 39327 21020 28197 27287 33633 31939 37146
6 22592 45877 29618 37155 36664 40259 39911 41546
8 28641 46878 35763 40111 40106 41293 41022 41950
10 33151 46741 38187 40269 39960 40922 40567 41606
#nt Init Sum Copy Update Triad Daxpy STriad SDaxpy
1 4109 11900 5637 8025 7407 9874 8981 11288
2 8057 22696 11011 15174 14821 18786 17599 21475
4 15602 39327 21020 28197 27287 33633 31939 37146
6 22592 45877 29618 37155 36664 40259 39911 41546
8 28641 46878 35763 40111 40106 41293 41022 41950
10 33151 46741 38187 40269 39960 40922 40567 41606
```
Please be aware the the single core memory bandwidth as well as the scaling behavior depends on the frequency settings.

2
config.mk
View File

@@ -1,4 +1,4 @@
# Supported: GCC, CLANG, ICC
# Supported: GCC, CLANG, ICC, ICX
TAG ?= ICC
ENABLE_OPENMP ?= true
ENABLE_LIKWID ?= false

20
include_ICX.mk Normal file
View File

@@ -0,0 +1,20 @@
CC = icx
GCC = gcc
LINKER = $(CC)
ifeq ($(ENABLE_OPENMP),true)
OPENMP = -qopenmp
endif
ifeq ($(ENABLE_LTO),true)
FAST_WORKAROUND = -ipo -O3 -static -fp-model=fast
else
FAST_WORKAROUND = -O3 -static -fp-model=fast
endif
VERSION = --version
CFLAGS = $(FAST_WORKAROUND) -xHost -qopt-streaming-stores=always -std=c99 -ffreestanding $(OPENMP)
LFLAGS = $(OPENMP)
DEFINES = -D_GNU_SOURCE
INCLUDES =
LIBS =