Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[mthreads] base/benchmarks: support s5000 #787

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

[mthreads] base/benchmarks: support s5000 #787

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9db29e8
[mthreads]base/toolkits: recovery MPI_intra test code
gliangMT Dec 13, 2024
59bfbcb
Merge branch 'FlagOpen:main' into main
gliangMT Dec 18, 2024
aac2190
Merge branch 'FlagOpen:main' into main
gliangMT Feb 24, 2025
7c70f41
[mthreads]base/benchmarks: support s5000
gliangMT Feb 25, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
44 changes: 44 additions & 0 deletions base/benchmarks/computation-BF16/mthreads/S5000/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,44 @@
# 参评AI芯片信息

* 厂商：Mthreads

## 服务器1

- 产品名称：S5000
- 产品型号：MTT S5000
- TDP：/

# 所用服务器配置

* 服务器数量：1

## 服务器1

* 单服务器内使用卡数：8
* 服务器型号：/
* 操作系统版本：Ubuntu 22.04.4 LTS
* 操作系统内核：Linux 5.15.0-105-generic
* CPU：/
* docker版本：24.0.7
* 内存：2TiB
* 服务器间AI芯片直连规格及带宽：此评测样例无需服务器间通信

# 评测结果

## 核心评测结果

| 评测项 | BF16算力测试值(8卡平均) | BF16算力标定值(8卡平均) | 测试标定比例(8卡平均) |
| ---- | ---------------- | ---------------- | ------------- |
| 评测结果 | / | / | / |

## 能耗监控结果

| 监控项 | 系统平均功耗 | 系统最大功耗 | 系统功耗标准差 | 单机TDP | 单卡平均功耗(8卡平均) | 单卡最大功耗(8卡最大) | 单卡功耗标准差(8卡平均) | 单卡TDP |
| ---- | ------------ | ------------ | ------------- | ----- | ------------- | ------------- | -------------- | ----- |
| 监控结果 | / | / | / | / | / | / | / | / |

## 其他重要监控结果

| 监控项 | 系统平均CPU占用 | 系统平均内存占用 | 单卡平均温度(8卡平均) | 单卡平均显存占用(8卡平均) |
| ---- | --------------- | -------------- | ------------- | --------------- |
| 监控结果 | / | / | / | / |
6 changes: 6 additions & 0 deletions base/benchmarks/computation-BF16/mthreads/S5000/case_config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
M: 6144
N: 192000
K: 4096
WARMUP: 100
ITERS: 12000
DIST_BACKEND: "mccl"
1 change: 1 addition & 0 deletions base/benchmarks/computation-BF16/mthreads/S5000/env.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
echo "MTHREADS PLACEHOLDER ENV.SH"
1 change: 1 addition & 0 deletions base/benchmarks/computation-BF16/mthreads/S5000/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
loguru
44 changes: 44 additions & 0 deletions base/benchmarks/computation-FP16/mthreads/S5000/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,44 @@
# 参评AI芯片信息

* 厂商：Mthreads

## 服务器1

- 产品名称：S5000
- 产品型号：MTT S5000
- TDP：/

# 所用服务器配置

* 服务器数量：1

## 服务器1

* 单服务器内使用卡数：8
* 服务器型号：/
* 操作系统版本：Ubuntu 22.04.4 LTS
* 操作系统内核：Linux 5.15.0-105-generic
* CPU：/
* docker版本：24.0.7
* 内存：2TiB
* 服务器间AI芯片直连规格及带宽：此评测样例无需服务器间通信

# 评测结果

## 核心评测结果

| 评测项 | FP16算力测试值(8卡平均) | FP16算力标定值(8卡平均) | 测试标定比例(8卡平均) |
| ---- | ---------------- | ---------------- | ------------- |
| 评测结果 | / | / | / |

## 能耗监控结果

| 监控项 | 系统平均功耗 | 系统最大功耗 | 系统功耗标准差 | 单机TDP | 单卡平均功耗(8卡平均) | 单卡最大功耗(8卡最大) | 单卡功耗标准差(8卡平均) | 单卡TDP |
| ---- | ------------ | ------------ | ------------- | ----- | ------------- | ------------- | -------------- | ----- |
| 监控结果 | / | / | / | / | / | / | / | / |

## 其他重要监控结果

| 监控项 | 系统平均CPU占用 | 系统平均内存占用 | 单卡平均温度(8卡平均) | 单卡平均显存占用(8卡平均) |
| ---- | --------------- | -------------- | ------------- | --------------- |
| 监控结果 | / | / | / | / |
6 changes: 6 additions & 0 deletions base/benchmarks/computation-FP16/mthreads/S5000/case_config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
M: 6144
N: 192000
K: 4096
WARMUP: 100
ITERS: 12000
DIST_BACKEND: "mccl"
1 change: 1 addition & 0 deletions base/benchmarks/computation-FP16/mthreads/S5000/env.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
echo "MTHREADS PLACEHOLDER ENV.SH"
1 change: 1 addition & 0 deletions base/benchmarks/computation-FP16/mthreads/S5000/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
loguru
44 changes: 44 additions & 0 deletions base/benchmarks/computation-FP32/mthreads/S5000/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,44 @@
# 参评AI芯片信息

* 厂商：Mthreads

## 服务器1

- 产品名称：S5000
- 产品型号：MTT S5000
- TDP：/

# 所用服务器配置

* 服务器数量：1

## 服务器1

* 单服务器内使用卡数：8
* 服务器型号：/
* 操作系统版本：Ubuntu 22.04.4 LTS
* 操作系统内核：Linux 5.15.0-105-generic
* CPU：/
* docker版本：24.0.7
* 内存：2TiB
* 服务器间AI芯片直连规格及带宽：此评测样例无需服务器间通信

# 评测结果

## 核心评测结果

| 评测项 | FP32算力测试值(16卡平均) | FP32算力标定值(16卡平均) | 测试标定比例(16卡平均) |
| ---- | ---------------- | ---------------- | ------------- |
| 评测结果 | / | / | / |

## 能耗监控结果

| 监控项 | 系统平均功耗(2机平均) | 系统最大功耗(2机最大) | 系统功耗标准差(2机最大) | 单机TDP | 单卡平均功耗(16卡平均) | 单卡最大功耗(16卡最大) | 单卡功耗标准差(16卡最大) | 单卡TDP |
| ---- | ------------ | ------------ | ------------- | ----- | ------------- | ------------- | -------------- | ----- |
| 监控结果 | / | / | / | / | / | / | / | / |

## 其他重要监控结果

| 监控项 | 系统平均CPU占用(2机平均) | 系统平均内存占用(2机平均) | 单卡平均温度(16卡平均) | 单卡平均显存占用(16卡平均) |
| ---- | --------------- | -------------- | ------------- | --------------- |
| 监控结果 | / | / | / | / |
6 changes: 6 additions & 0 deletions base/benchmarks/computation-FP32/mthreads/S5000/case_config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
M: 8192
N: 8192
K: 8192
DIST_BACKEND: "mccl"
WARMUP: 100
ITERS: 10000
2 changes: 2 additions & 0 deletions base/benchmarks/computation-FP32/mthreads/S5000/env.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,2 @@
echo "MTHREADS PLACEHOLDER ENV.SH"
export MUSA_KERNEL_TIMEOUT=3600000
1 change: 1 addition & 0 deletions base/benchmarks/computation-FP32/mthreads/S5000/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
loguru
44 changes: 44 additions & 0 deletions base/benchmarks/computation-FP64/mthreads/S5000/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,44 @@
# 参评AI芯片信息

* 厂商：Mthreads

## 服务器1

- 产品名称：S5000
- 产品型号：MTT S5000
- TDP：/

# 所用服务器配置

* 服务器数量：1

## 服务器1

* 单服务器内使用卡数：8
* 服务器型号：/
* 操作系统版本：Ubuntu 22.04.4 LTS
* 操作系统内核：Linux 5.15.0-105-generic
* CPU：/
* docker版本：24.0.7
* 内存：2TiB
* 服务器间AI芯片直连规格及带宽：此评测样例无需服务器间通信

# 评测结果

## 核心评测结果

| 评测项 | FP64算力测试值(8卡平均) | FP64算力标定值(8卡平均) | 测试标定比例(8卡平均) |
| ---- | ---------------- | ---------------- | ------------- |
| 评测结果 | / | / | / |

## 能耗监控结果

| 监控项 | 系统平均功耗 | 系统最大功耗 | 系统功耗标准差 | 单机TDP | 单卡平均功耗(8卡平均) | 单卡最大功耗(8卡最大) | 单卡功耗标准差(8卡平均) | 单卡TDP |
| ---- | ------------ | ------------ | ------------- | ----- | ------------- | ------------- | -------------- | ----- |
| 监控结果 | / | / | / | / | / | / | / | / |

## 其他重要监控结果

| 监控项 | 系统平均CPU占用 | 系统平均内存占用 | 单卡平均温度(8卡平均) | 单卡平均显存占用(8卡平均) |
| ---- | --------------- | -------------- | ------------- | --------------- |
| 监控结果 | / | / | / | / |
6 changes: 6 additions & 0 deletions base/benchmarks/computation-FP64/mthreads/S5000/case_config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
M: 16384
N: 16384
K: 16384
WARMUP: 10
ITERS: 600
DIST_BACKEND: "mccl"
1 change: 1 addition & 0 deletions base/benchmarks/computation-FP64/mthreads/S5000/env.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
echo "MTHREADS PLACEHOLDER ENV.SH"
1 change: 1 addition & 0 deletions base/benchmarks/computation-FP64/mthreads/S5000/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
loguru
29 changes: 29 additions & 0 deletions base/benchmarks/computation-FP8/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
# 评测原理

1. 使用computation-bound的算子GEMM来评测芯片FP8算力
2. 此算子为计算密集型，在业界被广泛用于测试算力

# 适配修改规范

本评测样例配置文件如下：

```yaml
M: 4096
N: 4096
K: 4096
WARMUP: 100
ITERS: 10000
DIST_BACKEND: "mpi"
```

1. M、N、K为GEMM算子的配置。本评测样例以[M,N]矩阵和[N,K]矩阵相乘作为计算内容。厂商可在正整数范围内任意调整此三项配置，发挥自身能力

例如，英伟达A800-80-SXM芯片采用M=8192、N=8192、K=8192

2. WARMUP为预热所需迭代次数。厂商可在正整数范围内任意调整此值。WARMUP迭代部分不计入性能计算

3. ITERS为正式评测迭代次数。厂商可在正整数范围内调整此值，同时保证总运行时间大于等于6分钟

4. DIST_BACKEND为通讯库。在本评测样例中，仅供初始化使用，无通信算子。厂商可任意调整为适用于自己的通讯库

例如，英伟达A800-80-SXM芯片采用DIST_BACKEND="nccl"
6 changes: 6 additions & 0 deletions base/benchmarks/computation-FP8/case_config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
M: 4096
N: 4096
K: 4096
WARMUP: 100
ITERS: 10000
DIST_BACKEND: "mpi"
Loading