NVIDIA GRID A100A vs AMD Radeon Instinct MI100
We compared two Professional market GPUs: 48GB VRAM GRID A100A and 32GB VRAM Radeon Instinct MI100 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA GRID A100A 's Advantages
More VRAM (48GB vs 32GB)
Larger VRAM bandwidth (1866GB/s vs 1229GB/s)
AMD Radeon Instinct MI100 's Advantages
Released 6 months late
Boost Clock has increased by 49% (1502MHz vs 1005MHz)
768 additional rendering cores
Lower TDP (300W vs 400W)
Score
Benchmark
FP32 (float)
GRID A100A
13.89 TFLOPS
Radeon Instinct MI100
+66%
23.07 TFLOPS
Graphics Card
May 2020
Release Date
Nov 2020
GRID
Generation
Radeon Instinct
Professional
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
900 MHz
Base Clock
1000 MHz
1005 MHz
Boost Clock
1502 MHz
1215 MHz
Memory Clock
1200 MHz
Memory
48GB
Memory Size
32GB
HBM2e
Memory Type
HBM2
6144bit
Memory Bus
4096bit
1866GB/s
Bandwidth
1229GB/s
Render Config
-
Compute Units
120
108
SM Count
-
6912
Shading Units
7680
432
TMUs
480
192
ROPs
64
432
Tensor Cores
-
-
-
-
192 KB (per SM)
L1 Cache
16 KB (per CU)
48 MB
L2 Cache
8 MB
-
-
-
Theoretical Performance
193.0 GPixel/s
Pixel Rate
96.13 GPixel/s
434.2 GTexel/s
Texture Rate
721.0 GTexel/s
55.57 TFLOPS
FP16 (half)
184.6 TFLOPS
13.89 TFLOPS
FP32 (float)
23.07 TFLOPS
6.947 TFLOPS
FP64 (double)
11.54 TFLOPS
Board Design
400W
TDP
300W
800 W
Suggested PSU
700 W
No outputs
Outputs
No outputs
None
Power Connectors
2x 8-pin
Graphics Processor
GA100
GPU Name
Arcturus
-
GPU Variant
Arcturus XL
Ampere
Architecture
CDNA 1.0
TSMC
Foundry
TSMC
7 nm
Process Size
7 nm
54.2 billion
Transistors
25.6 billion
826 mm²
Die Size
750 mm²
Graphics Features
N/A
DirectX
N/A
N/A
OpenGL
N/A
3.0
OpenCL
2.1
N/A
Vulkan
N/A
8.0
CUDA
-
N/A
Shader Model
N/A
