AMD FirePro W5170M vs NVIDIA H200 SXM 141 GB

Comparison of AMD FirePro W5170M with 2 GB GDDR5 and 640 cores vs NVIDIA H200 SXM 141 GB with 141 GB HBM3e and 16,896 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

AMD FirePro W5170M

AMD FirePro W5170M

RX 7900 XTX RX 7900 XTX
MI250 MI250
Instinct MI300X Instinct MI300X

NVIDIA H200 SXM 141 GB

67.4

NVIDIA H200 SXM 141 GB

67.4

Contents:

Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering Benchmarks Additional

Memory

Memory Size

2 GB
🔥 +6,950% 141 ГБ

Memory Type

GDDR5 HBM3e

Memory Bandwidth

72.00 GB/s
🔥 4.89 TB/s

Memory Bus Width

128 бит 6,144 бит

ML Performance

FP16 (Half Precision)

No
🔥 267.6 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

1.184 TFLOPS
🔥 +5,551% 66.91 TFLOPS

FP64 (Double Precision)

0.074 TFLOPS
🔥 +45,103% 33.45 TFLOPS

CUDA Cores

640
🔥 +2,540% 16,896

RT Cores

No No

Architecture & Compatibility

GPU Architecture

GCN 1.0 Hopper

SM (Streaming Multiprocessor)

No
🔥 132

PCIe Version

MXM-A (3.0) PCIe 5.0 x16

ML Software Support

CUDA Version

No 9.0

Clocks & Performance

Base Clock

900
🔥 +67% 1,500

Boost Clock

925
🔥 +114% 1,980

Memory Clock

1,125
🔥 +42% 1,593

Power Consumption

TDP/TGP

unknown 700 W

Recommended PSU

No 1100 W

Power Connector

None 8-pin EPS

Rendering

Texture Units (TMU)

40
🔥 +1,220% 528

ROP

No No

L2 Cache

256 KB
🔥 50 MB

Benchmarks

MLPerf, llama2-70b-99.9 (UNSET)

3 534 tokens/s

MLPerf, llama2-70b-99.9 (fp16)

3 553 tokens/s

MLPerf, llama2-70b-99.9 (fp8)

2 444 tokens/s

MLPerf, llama3.1-405b (fp16)

40.8 tokens/s

MLPerf, llama3.1-405b (fp8)

25.3 tokens/s

MLPerf, llama3.1-8b (fp8)

5 161 tokens/s

MLPerf, deepseek-r1 (fp8)

1 113 tokens/s

MLPerf, mixtral-8x7b (fp8)

7 132 tokens/s

Additional

Slots

MXM Module
🔥 SXM Module

Release Date

Aug. 25, 2014 Nov. 18, 2024

Display Outputs

Portable Device Dependent
No outputs

Renting is cheaper than buying