AMD Radeon Instinct MI325X vs NVIDIA P102-100

Comparison of AMD Radeon Instinct MI325X with 288 GB HBM3e and 19,456 cores vs NVIDIA P102-100 with 5 GB GDDR5X and 3,200 cores.

Loading...

Performance Rating

AMD Radeon Instinct MI325X outperforms NVIDIA P102-100 by 1,355.6% 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

NVIDIA P102-100

6.9

NVIDIA P102-100

6.9

Contents:

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

Memory

Memory Size

🔥 +5,660% 288 ГБ
5 ГБ

Memory Type

HBM3e GDDR5X

Memory Bandwidth

🔥 10.3 TB/s
440.3 GB/s

Memory Bus Width

8,192 бит 320 бит

ML Performance

FP16 (Half Precision)

🔥 +288% 653.7 TFLOPS
168.3 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +659% 81.72 TFLOPS
10.77 TFLOPS

FP64 (Double Precision)

🔥 +24,178% 81.72 TFLOPS
0.3366 TFLOPS

CUDA Cores

🔥 +508% 19,456
3,200

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 3.0 Pascal

SM (Streaming Multiprocessor)

No 25

PCIe Version

PCIe 5.0 x16 PCIe 1.0 x4

ML Software Support

CUDA Version

No 6.1

Clocks & Performance

Base Clock

🔥 1,000
1,582

Boost Clock

🔥 +25% 2,100
1,683

Memory Clock

🔥 +84% 2,525
1,376

Power Consumption

TDP/TGP

1000 W
🔥 -75% 250 W

Recommended PSU

1400 W
🔥 -57% 600 W

Power Connector

None 2x 8-pin

Rendering

Texture Units (TMU)

🔥 +508% 1,216
200

ROP

No No

L2 Cache

🔥 +540% 16 MB
2.5 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 Dual-slot

Release Date

Oct. 15, 2024 Feb. 12, 2018

Display Outputs

No outputs
No outputs

Renting is cheaper than buying