AMD Radeon Instinct MI300X vs NVIDIA GeForce RTX 5070 Mobile

Comparison of AMD Radeon Instinct MI300X with 192 GB HBM3 and 19,456 cores vs NVIDIA GeForce RTX 5070 Mobile with 8 GB GDDR7 and 4,608 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

AMD Radeon Instinct MI300X

94.1

AMD Radeon Instinct MI300X

94.1
MI250 MI250
Instinct MI300X Instinct MI300X
RX 7900 XTX RX 7900 XTX

NVIDIA GeForce RTX 5070 Mobile

NVIDIA GeForce RTX 5070 Mobile

Contents:

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

Memory

Memory Size

🔥 +2,300% 192 ГБ
8 GB

Memory Type

HBM3 GDDR7

Memory Bandwidth

🔥 10.3 TB/s
384.0 GB/s

Memory Bus Width

8,192 бит 128 бит

ML Performance

FP16 (Half Precision)

🔥 +4,724% 653.7 TFLOPS
13.55 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +503% 81.72 TFLOPS
13.55 TFLOPS

FP64 (Double Precision)

🔥 +38,502% 81.72 TFLOPS
0.2117 TFLOPS

CUDA Cores

🔥 +322% 19,456
4,608

RT Cores

No 36

Architecture & Compatibility

GPU Architecture

CDNA 3.0 Blackwell 2.0

SM (Streaming Multiprocessor)

No 36

PCIe Version

PCIe 5.0 x16 PCIe 5.0 x16

ML Software Support

CUDA Version

No 12.0

Clocks & Performance

Base Clock

🔥 +4% 1,000
960

Boost Clock

🔥 +43% 2,100
1,470

Memory Clock

🔥 +68% 2,525
1,500

Power Consumption

TDP/TGP

750 W
🔥 -93% 50 W

Recommended PSU

1150 W No

Power Connector

None None

Rendering

Texture Units (TMU)

🔥 +744% 1,216
144

ROP

No 36

L2 Cache

🔥 16 MB
32 MB

Benchmarks

MLPerf, llama2-70b-99.9 (UNSET)

1 983 tokens/s

MLPerf, llama2-70b-99.9 (fp16)

1 740 tokens/s

MLPerf, llama2-70b-99.9 (fp8)

1 057 tokens/s

MLPerf, llama3.1-405b (UNSET)

30.4 tokens/s

MLPerf, llama3.1-405b (fp16)

34.8 tokens/s

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

232.9 tokens/s

MLPerf, mixtral-8x7b (fp8)

5 975 tokens/s

Additional

Slots

OAM Module IGP

Release Date

Dec. 6, 2023 April 12, 2025

Display Outputs

No outputs
Portable Device Dependent

Renting is cheaper than buying