AMD Radeon Instinct MI355X vs NVIDIA H100 SXM5 80 GB

Comparison of AMD Radeon Instinct MI355X with 288 GB HBM3e and 16,384 cores vs NVIDIA H100 SXM5 80 GB with 80 GB HBM3 and 16,896 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

AMD Radeon Instinct MI355X

AMD Radeon Instinct MI355X

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

NVIDIA H100 SXM5 80 GB

NVIDIA H100 SXM5 80 GB

Contents:

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

Memory

Memory Size

288 GB 80 GB

Memory Type

HBM3e HBM3

Memory Bandwidth

8.19 TB/s 3.36 TB/s

Memory Bus Width

8,192 бит 5,120 бит

ML Performance

FP16 (Half Precision)

629.1 TFLOPS 267.6 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

78.64 TFLOPS 66.91 TFLOPS

FP64 (Double Precision)

78.64 TFLOPS 33.45 TFLOPS

CUDA Cores

16,384 16,896

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 4.0 Hopper

SM (Streaming Multiprocessor)

No 132

PCIe Version

PCIe 5.0 x16 PCIe 5.0 x16

ML Software Support

CUDA Version

No 9.0

Clocks & Performance

Base Clock

1,000 1,590

Boost Clock

2,400 1,980

Memory Clock

2,000 1,313

Power Consumption

TDP/TGP

1400 W
🔥 -50% 700 W

Recommended PSU

1800 W
🔥 -39% 1100 W

Power Connector

None None

Rendering

Texture Units (TMU)

1,024 528

ROP

No No

L2 Cache

32 MB 50 MB

Benchmarks

MLPerf, llama2-70b-99.9 (mxfp4)

10 129 tokens/s

MLPerf, llama3.1-405b (mxfp4)

252.4 tokens/s

MLPerf, mixtral-8x7b (fp8)

6 362 tokens/s

Additional

Slots

OAM Module SXM Module

Release Date

March 12, 2025 Oct. 13, 2022

Display Outputs

No outputs
No outputs

Renting is cheaper than buying