AMD Radeon Instinct MI325X vs PNY A100 Liquid Cooled

Comparison of AMD Radeon Instinct MI325X with 288 GB HBM3e and 19,456 cores vs PNY A100 Liquid Cooled with 80 GB HBM2e and 6,912 cores.

Loading...

Performance Rating

AMD Radeon Instinct MI325X outperforms PNY A100 Liquid Cooled by 290.17% in the overall GPU ARK performance rating

A100 A100
H200 H200
MI325X MI325X

AMD Radeon Instinct MI325X

100.0

AMD Radeon Instinct MI325X

100.0
RX 7900 XTX RX 7900 XTX
MI250 MI250
Instinct MI300X Instinct MI300X

PNY A100 Liquid Cooled

25.6

PNY A100 Liquid Cooled

25.6

Contents:

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

Memory

Memory Size

🔥 +260% 288 ГБ
80 ГБ

Memory Type

HBM3e HBM2e

Memory Bandwidth

🔥 +431% 10.3 TB/s
1.94 TB/s

Memory Bus Width

8,192 бит 5,120 бит

ML Performance

FP16 (Half Precision)

🔥 +738% 653.7 TFLOPS
77.97 TFLOPS

BF16 (Brain Float)

No
🔥 311.84 TFLOPS

TF32 (TensorFloat)

No
🔥 155.92

Compute Power

FP32 (Single Precision)

🔥 +319% 81.72 TFLOPS
19.49 TFLOPS

FP64 (Double Precision)

🔥 +738% 81.72 TFLOPS
9.746 TFLOPS

CUDA Cores

🔥 +181% 19,456
6,912

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 3.0 Ampere

SM (Streaming Multiprocessor)

No
🔥 108

PCIe Version

PCIe 5.0 x16 PCIe 4.0 x16

ML Software Support

CUDA Version

No 8.0

Clocks & Performance

Base Clock

1,000
🔥 +6% 1,065

Boost Clock

🔥 +49% 2,100
1,410

Memory Clock

🔥 +67% 2,525
1,512

Power Consumption

TDP/TGP

1000 W
🔥 -70% 300 W

Recommended PSU

1400 W
🔥 -50% 700 W

Power Connector

None 8-pin EPS

Rendering

Texture Units (TMU)

🔥 +181% 1,216
432

ROP

No No

L2 Cache

16 MB
🔥 +400% 80 MB

Benchmarks

MLPerf, llama2-70b-99.9 (Dummy)

3 596 tokens/s

MLPerf, llama2-70b-99.9 (fp8)

1 946 tokens/s

llama.cpp, llama-2-7b-Q4_0

22.4 tokens/s

MLPerf, mixtral-8x7b (fp8)

6 975 tokens/s

Additional

Slots

OAM Module Single-slot

Release Date

Oct. 15, 2024 June 28, 2021

Display Outputs

No outputs
No outputs

Renting is cheaper than buying