llama tr

AMD Ryzen Threadripper 3990X 64-Core testing with a Gigabyte TRX40 AORUS PRO WIFI (F6 BIOS) and AMD Radeon RX 5700 8GB on Ubuntu 23.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2401109-PTS-LLAMATR408
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results
Show Result Confidence Charts

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
a
January 11
  36 Minutes
b
January 11
  14 Minutes
c
January 11
  12 Minutes
f
January 11
  13 Minutes
Invert Hiding All Results Option
  19 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


llama trOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen Threadripper 3990X 64-Core @ 2.90GHz (64 Cores / 128 Threads)Gigabyte TRX40 AORUS PRO WIFI (F6 BIOS)AMD Starship/Matisse128GBSamsung SSD 970 EVO Plus 500GBAMD Radeon RX 5700 8GB (1750/875MHz)AMD Navi 10 HDMI AudioDELL P2415QIntel I211 + Intel Wi-Fi 6 AX200Ubuntu 23.106.5.0-14-generic (x86_64)GNOME Shell 45.0X Server + Wayland4.6 Mesa 23.2.1-1ubuntu3 (LLVM 15.0.7 DRM 3.54)GCC 13.2.0ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilerFile-SystemScreen ResolutionLlama Tr BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x830107a- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_rstack_overflow: Mitigation of safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

abcfResult OverviewPhoronix Test Suite100%101%103%104%106%Llama.cppLlama.cppCacheBenchCacheBenchLlama.cppCacheBenchllama-2-7b.Q4_0.ggufllama-2-13b.Q4_0.ggufR.M.WReadllama-2-70b-chat.Q5_0.ggufWrite

llama trllama-cpp: llama-2-70b-chat.Q5_0.ggufcachebench: Readcachebench: Writecachebench: Read / Modify / Writellama-cpp: llama-2-13b.Q4_0.ggufllama-cpp: llama-2-7b.Q4_0.ggufabcf1.4510914.54234463817.965300115970.7977648.2114.231.4510857.24316364184.073674117095.8227488.0114.231.4510863.41747363783.849327116205.996158.0214.991.4610954.49972364095.092643117248.299362814.19OpenBenchmarking.org

Llama.cpp

Llama.cpp is a port of Facebook's LLaMA model in C/C++ developed by Georgi Gerganov. Llama.cpp allows the inference of LLaMA and other supported models in C/C++. For CPU inference Llama.cpp supports AVX2/AVX-512, ARM NEON, and other modern ISAs along with features like OpenBLAS usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-70b-chat.Q5_0.ggufabcf0.32850.6570.98551.3141.6425SE +/- 0.00, N = 31.451.451.451.461. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

CacheBench

This is a performance test of CacheBench, which is part of LLCbench. CacheBench is designed to test the memory and cache bandwidth performance Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Readbcaf2K4K6K8K10KSE +/- 14.16, N = 310857.2410863.4210914.5410954.50MIN: 10840.61 / MAX: 10931.91MIN: 10840.77 / MAX: 10909.02MIN: 10859.9 / MAX: 10971.13MIN: 10928.99 / MAX: 10978.81. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Writecafb14K28K42K56K70KSE +/- 64.28, N = 363783.8563817.9764095.0964184.07MIN: 58980.46 / MAX: 64725.85MIN: 59125.08 / MAX: 64846.48MIN: 58078.31 / MAX: 65198.73MIN: 59624.93 / MAX: 65076.131. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / Writeacbf30K60K90K120K150KSE +/- 213.45, N = 3115970.80116206.00117095.82117248.30MIN: 96405.85 / MAX: 127016.83MIN: 97578.52 / MAX: 126854.22MIN: 98368.14 / MAX: 127792.96MIN: 96453.27 / MAX: 127988.941. (CC) gcc options: -O3 -lrt

Llama.cpp

Llama.cpp is a port of Facebook's LLaMA model in C/C++ developed by Georgi Gerganov. Llama.cpp allows the inference of LLaMA and other supported models in C/C++. For CPU inference Llama.cpp supports AVX2/AVX-512, ARM NEON, and other modern ISAs along with features like OpenBLAS usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-13b.Q4_0.gguffbca246810SE +/- 0.06, N = 38.008.018.028.211. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-7b.Q4_0.gguffabc48121620SE +/- 0.00, N = 314.1914.2314.2314.991. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas