AMD Radeon Instinct MI325X vs NVIDIA RTX PRO 3000 Blackwell Mobile

Comparison AMD Radeon Instinct MI325X with 288 GB HBM3e and 19,456 cores vs NVIDIA RTX PRO 3000 Blackwell Mobile with 12 GB GDDR7 and 5,888 cores.

Loading...

Performance Rating

AMD Radeon Instinct MI325X outperforms NVIDIA RTX PRO 3000 Blackwell Mobile by 805.8% 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 RTX PRO 3000 Blackwell Mobile

11.0

NVIDIA RTX PRO 3000 Blackwell Mobile

11.0

Expert Comparison

AMD Radeon Instinct MI325X имеет значительно большую пропускную способность и больше ядер и памяти, что делает его более мощным для вычислений и больших данных. Напротив, NVIDIA RTX PRO 3000 Blackwell Mobile более энергоэффективен и подходит для мобильных решений, хотя и значительно уступает по производительности в стационарных приложениях.

Contents:

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

Memory

Memory Size

🔥 +2,300% 288 ГБ
12 GB

Memory Type

HBM3e GDDR7

Memory Bandwidth

🔥 10.3 TB/s
672.0 GB/s

Memory Bus Width

8,192 бит 192 бит

ML Performance

FP16 (Half Precision)

🔥 +3,359% 653.7 TFLOPS
18.9 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +332% 81.72 TFLOPS
18.9 TFLOPS

FP64 (Double Precision)

🔥 +27,574% 81.72 TFLOPS
0.2953 TFLOPS

CUDA Cores

🔥 +230% 19,456
5,888

RT Cores

No 46

Architecture & Compatibility

GPU Architecture

CDNA 3.0 Blackwell 2.0

SM (Streaming Multiprocessor)

No 46

PCIe Version

PCIe 5.0 x16 PCIe 5.0 x16

ML Software Support

CUDA Version

No 12.0

CUDA Toolkit (first supported)

v12

CUDA Toolkit status

Supported Supported

Clocks & Performance

Base Clock

🔥 +18% 1,000
848

Boost Clock

🔥 +31% 2,100
1,605

Memory Clock

🔥 +44% 2,525
1,750

Power Consumption

Recommended PSU

1400 W No

Power Connector

None None

TDP/TGP

1000 W
🔥 -94% 60 W

Rendering

Texture Units (TMU)

🔥 +561% 1,216
184

ROP

No 46

L2 Cache

🔥 16 MB
48 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 IGP

Release Date

Oct. 12, 2024 No

Display Outputs

No outputs
Portable Device Dependent

Renting is cheaper than buying