AMD Radeon Instinct MI325X vs BIOSTAR RX Vega 56

Comparison of AMD Radeon Instinct MI325X with 288 GB HBM3e and 19,456 cores vs BIOSTAR RX Vega 56 with 8 GB HBM2 and 3,584 cores.

Loading...

Performance Rating

AMD Radeon Instinct MI325X outperforms BIOSTAR RX Vega 56 by 1,056.07% 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

BIOSTAR RX Vega 56

8.7

BIOSTAR RX Vega 56

8.7

Contents:

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

Memory

Memory Size

🔥 +3,500% 288 ГБ
8 ГБ

Memory Type

HBM3e HBM2

Memory Bandwidth

🔥 10.3 TB/s
409.6 GB/s

Memory Bus Width

8,192 бит 2,048 бит

ML Performance

FP16 (Half Precision)

🔥 +3,000% 653.7 TFLOPS
21.09 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +675% 81.72 TFLOPS
10.54 TFLOPS

FP64 (Double Precision)

🔥 +12,301% 81.72 TFLOPS
0.659 TFLOPS

CUDA Cores

🔥 +443% 19,456
3,584

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 3.0 GCN 5.0

SM (Streaming Multiprocessor)

No No

PCIe Version

PCIe 5.0 x16 PCIe 3.0 x16

ML Software Support

CUDA Version

No No

Clocks & Performance

Base Clock

🔥 1,000
1,156

Boost Clock

🔥 +43% 2,100
1,471

Memory Clock

🔥 +216% 2,525
800

Power Consumption

TDP/TGP

1000 W
🔥 -79% 210 W

Recommended PSU

1400 W
🔥 -61% 550 W

Power Connector

None 2x 8-pin

Rendering

Texture Units (TMU)

🔥 +443% 1,216
224

ROP

No No

L2 Cache

🔥 +300% 16 MB
4 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 Aug. 14, 2017

Display Outputs

No outputs
1x HDMI 2.0b
3x DisplayPort 1.4a

Renting is cheaper than buying