GitLab

**Description**
Benchmarks: Micro benchmarks - Update hipblaslt metric unit to tflops

**Description**
Adapt to hipblasLt data type changes.

**Description**
Bug fix for rocm docker build and deploy.

**Description**
Introduce option for fine-grained GPU memory for AMD GPUs, which should
include GPU P2P performance.

**Description**
Make metrics of dist-inference-cpp aligned with PyTorch version.

---------
Co-authored-by: Yuting Jiang <yutingjiang@microsoft.com>

**Description**
Add data type option for NCCL and RCCL tests.

**Description**
Add in-place metrics for NCCL/RCCL benchmark for latency measurement.

**Description**
Support different data types in different hipblasLt versions for
dist_inference

**Description**
`add_compile_options` will not work for ROCm build, change it to setting
`CMAKE_CXX_FLAGS`.

**Description**
Add distributed inference benchmark cpp implementation.

**Description**
upgrade to rocm5.7 dockerfile.

---------
Co-authored-by: yukirora <yuting.jiang@microsoft.com>

Benchmarks: Micro benchmark - Add one-to-all, all-to-one, all-to-all support to gpu_copy_bw_performance (#588)

**Description**
Add one-to-all, all-to-one, all-to-all support to
gpu_copy_bw_performance, and fix performance bug in gpu_copy

**Description**
Revise NCCL/RCCL benchmarks to graph mode add latency metrics.

**Description**
Benchmarks: micro benchmark - Support cpu-gpu and gpu-cpu in
ib-validation

**Major Revision**
- Support cpu-gpu and gpu-cpu in ib-validation


**Minor Revision**
- support multi msg size, multi direction, multi ib commands in
ib-validation

Upgrade Docker image to CUDA 12.2 for H100:
* upgrade base image to 23.10
* fix onnxruntime version in python3.10
* fix compilation errors

**Description**
add initialization options for rocm gemm flops.

**Description**
hipblaslt function benchmark and rebase cublaslt function benchmark.

**Description**
add int8 support for cublaslt function.

**Description**
remove cp ptx file in gpu burn test since the command is run inside
self.args.bin_dir dir.


https://github.com/microsoft/superbenchmark/blob/d246bab430adeb461072918a551b2e2b68c9bce5/superbench/benchmarks/micro_benchmarks/micro_base.py#L183

**Description**
source code for evaluating NVDEC decoding performance.

---------
Co-authored-by: yukirora <yuting.jiang@microsoft.com>

**Description**
add source code for DirectXRenderPerf.

---------
Co-authored-by: yukirora <yuting.jiang@microsoft.com>

**Description**
Cherry-pick bug fixes from v0.9.0 to main.

**Major Revision**
- CI/CD: pipeline - clean more disk space to fix rocm building image
pipeline(#555 )
- Benchmarks: bug fix - use absolute path for input file in
DirectXEncodingLatency(#554)
- CI/CD - add push win docker image on release branch in pipeline (#552)
- Docs - Upgrade version and release note(#557)

**Description**
add python code for DirectXGPUEncodingLatency.

**Description**
add python code for DirectXGPUCopy.

**Description**
add python code for DirecXGPUMemBw.

**Description**
add python code for DirectX core flops and init DirectX test pipeline.

**Major Revision**
- add python code for DirectX core flops 
- init DirectX test pipeline


**Minor Revision**
- add test for DirectX core flops

**Description**
add auto selecting algorithm support for cudnn functions.

**Major Revision**
- add auto selecting algorithm support for cudnn functions in source
code
- add 'auto_algo' option in benchmark
- add related test

* Update result parsing for newer tensorrt versions
* Update arguments when load torchvision models

**Description**
Add source code of DirectxGPUCopy microbenchmark.

**Description**
Add source code of DirectxGPUMemBw microbenchmark.

---------
Co-authored-by: v-junlinlv <v-junlinlv@microsoft.com>

**Description**
Add source code of DirectXGPUCoreFLOPs microbenchmark.

---------
Co-authored-by: v-junlinlv <v-junlinlv@microsoft.com>

**Description**
This commit revises distributed inference benchmark to give a unified
step time result by taking maximum step times of different GPUs.

**Description**

Cherry-pick bug fixes from v0.8.0 to main.

**Major Revisions**

* Monitor - Fix the cgroup version checking logic (#502)
* Benchmark - Fix matrix size overflow issue in cuBLASLt GEMM (#503)
* Fix wrong torch usage in communication wrapper for Distributed
Inference Benchmark (#505)
* Analyzer: Fix bug in python3.8 due to pandas api change (#504)
* Bug - Fix bug to get metric from cmd when error happens (#506)
* Monitor - Collect realtime GPU power when benchmarking (#507)
* Add num_workers argument in model benchmark (#511)
* Remove unreachable condition when write host list (#512)
* Update cuda11.8 image to cuda12.1 based on nvcr23.03 (#513)
* Doc - Fix wrong unit of cpu-memory-bw-latency in doc (#515)
* Docs - Upgrade version and release note (#508)
Co-authored-by: guoshzhao <guzhao@microsoft.com>
Co-authored-by: Ziyue Yang <ziyyang@microsoft.com>
Co-authored-by: Yuting Jiang <yutingjiang@microsoft.com>

**Description**
This PR adds a micro-benchmark of distributed model inference workloads.

**Major Revision**
- Add a new micro-benchmark dist-inference.
- Add corresponding example and unit tests.
- Update configuration files to include this new micro-benchmark.
- Update micro-benchmark README.

---------
Co-authored-by: Peng Cheng <chengpeng5555@outlook.com>

Support batch and shape range with multiplication factors in cublaslt
gemm benchmark.

**Description**

- Adding HPL benchmark

---------
Co-authored-by: Ubuntu <azureuser@sbtestvm.jzlku1oskncengjiado35wf1hd.ax.internal.cloudapp.net>
Co-authored-by: Peng Cheng <chengpeng5555@outlook.com>

**Description**
Support error tolerance in micro-benchmark for CuDNN function


**Major Revision**
- revise micro_base to support running the remaining commands run when
one command failed in the microbenchmark
- make error tolerance as true in cudnn functions

Support FP64/TF32/FP16/BF16 in cublaslt (batch) GEMM.

Benchmarks: Revision - Support flexible warmup and non-random data initialization in cublas-benchmark  (#479)

**Description**
revise cublas-benchmark for flexible warmup and fill data with fixed
number for perf test to improve the running efficiency.

**Major Revision**
- remove num_in_steps for warmup to support more flexible warmup setting
for users
- Add support to generate input with fixed number for perf test

**Description**

- Added stream benchmark
- Added stream unit test
- Added stream example
- Modified docker files to build stream

---------
Co-authored-by: Ubuntu <azureuser@sbtestvm.jzlku1oskncengjiado35wf1hd.ax.internal.cloudapp.net>
Co-authored-by: Peng Cheng <chengpeng5555@outlook.com>
Co-authored-by: Yifan Xiong <xiongyf@yandex.com>