AMD Radeon Instinct MI325X vs NVIDIA RTX PRO 6000 Blackwell Server
We compared two Desktop platform GPUs: 288GB VRAM Radeon Instinct MI325X and 96GB VRAM RTX PRO 6000 Blackwell Server to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
AMD Radeon Instinct MI325X 's Advantages
More VRAM (288GB vs 96GB)
Larger VRAM bandwidth (10.3TB/s vs 1.79TB/s)
NVIDIA RTX PRO 6000 Blackwell Server 's Advantages
Boost Clock has increased by 25% (2617MHz vs 2100MHz)
4608 additional rendering cores
Lower TDP (600W vs 1000W)
Score
Benchmark
FP32 (float)
Radeon Instinct MI325X
81.72 TFLOPS
RTX PRO 6000 Blackwell Server
+54%
126 TFLOPS
Graphics Card
Oct 2024
Release Date
Mar 2025
Radeon Instinct
Generation
Server Blackwell
Desktop
Type
Desktop
PCIe 5.0 x16
Bus Interface
PCIe 5.0 x16
Clock Speeds
1000 MHz
Base Clock
1590 MHz
2100 MHz
Boost Clock
2617 MHz
2525 MHz
Memory Clock
1750 MHz
Memory
288GB
Memory Size
96GB
HBM3e
Memory Type
GDDR7
8192bit
Memory Bus
512bit
10.3TB/s
Bandwidth
1.79TB/s
Render Config
304
Compute Units
-
-
SM Count
188
19456
Shading Units
24064
1216
TMUs
752
0
ROPs
192
1216
Tensor Cores
752
-
RT Cores
188
16 KB (per CU)
L1 Cache
128 KB (per SM)
16 MB
L2 Cache
128 MB
256 MB
L3 Cache
-
Theoretical Performance
0 MPixel/s
Pixel Rate
502.5 GPixel/s
2553.6 GTexel/s
Texture Rate
1968 GTexel/s
653.7 TFLOPS
FP16 (half)
126.0 TFLOPS
81.72 TFLOPS
FP32 (float)
126.0 TFLOPS
81.72 TFLOPS
FP64 (double)
1.968 TFLOPS
Board Design
1000W
TDP
600W
1400 W
Suggested PSU
1000 W
No outputs
Outputs
4x DisplayPort 2.1b
None
Power Connectors
1x 16-pin
Graphics Processor
Aqua Vanjaram
GPU Name
GB202
-
-
-
CDNA 3.0
Architecture
Blackwell 2.0
TSMC
Foundry
TSMC
5 nm
Process Size
5 nm
153 billion
Transistors
92.2 billion
1017 mm²
Die Size
750 mm²
Graphics Features
N/A
DirectX
12 Ultimate (12_2)
N/A
OpenGL
4.6
3.0
OpenCL
3.0
N/A
Vulkan
1.4
-
CUDA
12.0
N/A
Shader Model
6.8
