NVIDIA GRID K2 vs AMD Instinct MI300A
We compared two Professional market GPUs: 4GB VRAM GRID K2 and 128GB VRAM AMD Instinct MI300A to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA GRID K2 's Advantages
Lower TDP (225W vs 760W)
AMD Instinct MI300A 's Advantages
Released 10 years and 7 months late
Boost Clock2100MHz
More VRAM (128GB vs 4GB)
Larger VRAM bandwidth (5300GB/s vs 160.0GB/s)
13056 additional rendering cores
Score
Benchmark
FP32 (float)
GRID K2
2.289 TFLOPS
AMD Instinct MI300A
+5256%
122.6 TFLOPS
Graphics Card
May 2013
Release Date
Dec 2023
GRID
Generation
Instinct
Professional
Type
Professional
PCIe 3.0 x16
Bus Interface
PCIe 5.0 x16
Clock Speeds
-
Base Clock
1000 MHz
-
Boost Clock
2100 MHz
1250 MHz
Memory Clock
5200 MHz
Memory
4GB
Memory Size
128GB
GDDR5
Memory Type
HBM3
256bit
Memory Bus
8192bit
160.0GB/s
Bandwidth
5300GB/s
Render Config
-
Compute Units
228
-
-
-
1536
Shading Units
14592
128
TMUs
880
32
ROPs
0
-
-
-
-
-
-
16 KB (per SMX)
L1 Cache
16 KB (per CU)
512 KB
L2 Cache
16 MB
-
-
-
Theoretical Performance
23.84 GPixel/s
Pixel Rate
0 MPixel/s
95.36 GTexel/s
Texture Rate
1496 GTexel/s
-
FP16 (half)
980.6 TFLOPS
2.289 TFLOPS
FP32 (float)
122.6 TFLOPS
95.36 GFLOPS
FP64 (double)
61.3 TFLOPS
Board Design
225W
TDP
760W
550 W
Suggested PSU
1000 W
No outputs
Outputs
No outputs
1x 6-pin + 1x 8-pin
Power Connectors
None
Graphics Processor
GK104
GPU Name
MI300
GK104-895-A2
GPU Variant
-
Kepler
Architecture
CDNA 3.0
TSMC
Foundry
TSMC
28 nm
Process Size
5 nm
3.54 billion
Transistors
146 billion
294 mm²
Die Size
1017 mm²
Graphics Features
12 (11_0)
DirectX
N/A
4.6
OpenGL
N/A
3.0
OpenCL
3.0
1.2.175
Vulkan
N/A
3.0
CUDA
-
6.5 (5.1)
Shader Model
N/A
