Whisper.cpp

Whisper.cpp is a port of OpenAI's Whisper model in C/C++. Whisper.cpp is developed by Georgi Gerganov for transcribing WAV audio files to text / speech recognition. Whisper.cpp supports ARM NEON, x86 AVX, and other advanced CPU features.

To run this test with the Phoronix Test Suite, the basic command is: phoronix-test-suite benchmark whisper-cpp.

Project Site

github.com

Source Repository

github.com

Test Created

16 June 2023

Last Updated

2 June 2024

Test Maintainer

Michael Larabel 

Test Type

System

Average Install Time

3 Minutes, 13 Seconds

Average Run Time

33 Minutes, 24 Seconds

Test Dependencies

C/C++ Compiler Toolchain

Accolades

10k+ Downloads

Supported Platforms


Public Result Uploads *Reported Installs **Reported Test Completions **Test Profile Page ViewsOpenBenchmarking.orgEventsWhisper.cpp Popularity Statisticspts/whisper-cpp2023.062023.072023.082023.092023.102023.112023.122024.012024.022024.032024.042024.052024.062024.072024.082024.092024.102024.1112002400360048006000
* Uploading of benchmark result data to OpenBenchmarking.org is always optional (opt-in) via the Phoronix Test Suite for users wishing to share their results publicly.
** Data based on those opting to upload their test results to OpenBenchmarking.org and users enabling the opt-in anonymous statistics reporting while running benchmarks from an Internet-connected platform.
Data updated weekly as of 20 November 2024.
ggml-small.en35.0%ggml-base.en35.8%ggml-medium.en29.2%Model Option PopularityOpenBenchmarking.org

Revision History

pts/whisper-cpp-1.1.0   [View Source]   Sun, 02 Jun 2024 10:53:16 GMT
Update against Whisper.cpp 1.6.2 upstream.

pts/whisper-cpp-1.0.0   [View Source]   Fri, 16 Jun 2023 17:44:41 GMT
Add initial whisper.cpp AI benchmark.

Suites Using This Test

Machine Learning

HPC - High Performance Computing


Performance Metrics

Analyze Test Configuration:

Whisper.cpp 1.4

Model: ggml-base.en - Input: 2016 State of the Union

OpenBenchmarking.org metrics for this test profile configuration based on 181 public results since 16 June 2023 with the latest data as of 26 May 2024.

Below is an overview of the generalized performance for components where there is sufficient statistically significant data based upon user-uploaded results. It is important to keep in mind particularly in the Linux/open-source space there can be vastly different OS configurations, with this overview intended to offer just general guidance as to the performance expectations.

Component
Percentile Rank
# Compatible Public Results
Seconds (Average)
Mid-Tier
75th
> 108
70th
11
115 +/- 10
66th
11
120 +/- 10
60th
5
136 +/- 8
59th
4
137 +/- 1
55th
4
149 +/- 1
51st
4
154 +/- 2
Median
50th
156
29th
5
203 +/- 27
Low-Tier
25th
> 234
25th
7
235 +/- 4
21st
4
246 +/- 1
11th
3
340 +/- 11
11th
4
349 +/- 13
4th
4
986 +/- 34
OpenBenchmarking.orgDistribution Of Public Results - Model: ggml-base.en - Input: 2016 State of the Union181 Results Range From 0 To 3775 Seconds76152228304380456532608684760836912988106411401216129213681444152015961672174818241900197620522128220422802356243225082584266027362812288829643040311631923268334434203496357236483724380020406080100

Based on OpenBenchmarking.org data, the selected test / test configuration (Whisper.cpp 1.4 - Model: ggml-base.en - Input: 2016 State of the Union) has an average run-time of 12 minutes. By default this test profile is set to run at least 3 times but may increase if the standard deviation exceeds pre-defined defaults or other calculations deem additional runs necessary for greater statistical accuracy of the result.

OpenBenchmarking.orgMinutesTime Required To Complete BenchmarkModel: ggml-base.en - Input: 2016 State of the UnionRun-Time1020304050Min: 1 / Avg: 11.61 / Max: 47

Based on public OpenBenchmarking.org results, the selected test / test configuration has an average standard deviation of 1%.

OpenBenchmarking.orgPercent, Fewer Is BetterAverage Deviation Between RunsModel: ggml-base.en - Input: 2016 State of the UnionDeviation3691215Min: 0 / Avg: 1 / Max: 11

Notable Instruction Set Usage

Notable instruction set extensions supported by this test, based on an automatic analysis by the Phoronix Test Suite / OpenBenchmarking.org analytics engine.

Instruction Set
Support
Instructions Detected
Used by default on supported hardware.
Found on Intel processors since Sandy Bridge (2011).
Found on AMD processors since Bulldozer (2011).

 
VZEROUPPER VBROADCASTSS VBROADCASTSD VINSERTF128 VEXTRACTF128 VPERMILPS
Used by default on supported hardware.
Found on Intel processors since Haswell (2013).
Found on AMD processors since Excavator (2016).

 
VPBROADCASTQ VINSERTI128 VEXTRACTI128 VPBROADCASTW VPERMQ VPBROADCASTB VPBROADCASTD VPSLLVD VPERMD VPSRLVD VPERM2I128
Advanced Vector Extensions 512 (AVX512)
Used by default on supported hardware.
 
(ZMM REGISTER USE)
AVX Vector Neural Network Instructions (AVX-VNNI)
Used by default on supported hardware.
 
VPDPBUSD
FMA (FMA)
Used by default on supported hardware.
Found on Intel processors since Haswell (2013).
Found on AMD processors since Bulldozer (2011).

 
VFMADD132PS VFMADD132SS VFMADD231SD VFMADD231PS VFMADD213PS VFMADD213SD VFMADD132SD VFNMADD231PS VFNMADD132PS VFMADD213SS VFMADD231SS VFNMADD231SS VFNMADD213SS VFNMADD213SD
Last automated analysis: 6 June 2024

This test profile binary relies on the shared libraries libm.so.6, libc.so.6.

Tested CPU Architectures

This benchmark has been successfully tested on the below mentioned architectures. The CPU architectures listed is where successful OpenBenchmarking.org result uploads occurred, namely for helping to determine if a given test is compatible with various alternative CPU architectures.

CPU Architecture
Kernel Identifier
Verified On
Intel / AMD x86 64-bit
x86_64
(Many Processors)
ARMv8 64-bit
aarch64
ARMv8 Cortex-A53 4-Core

Recent Test Results

OpenBenchmarking.org Results Compare

1 System - 3 Benchmark Results

AMD Ryzen 9 7940HS - Shenzhen Meigao Electronic Equipment F7BSC - AMD Device 14e8

cachyos rolling - 6.9.1-arch1-2 - Cinnamon 6.0.4

Find More Test Results