NVIDIA A100 PCIe vs NVIDIA RTX 4000 SFF Ada Generation
AI GPU
We compared a GPU: 40GB VRAM A100 PCIe and a Professional market GPU: 16GB VRAM RTX 4000 SFF Ada Generation to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA A100 PCIe 's Advantages
Released 1 years and 7 months late
More VRAM (40GB vs 16GB)
Larger VRAM bandwidth (1555GB/s vs 1.02TB/s)
3072 additional rendering cores
Lower TDP (250W vs 300W)
NVIDIA RTX 4000 SFF Ada Generation 's Advantages
Boost Clock has increased by 24% (1746MHz vs 1410MHz)
Score
Benchmark
FP32 (float)
A100 PCIe
+45%
19.49 TFLOPS
RTX 4000 SFF Ada Generation
13.41 TFLOPS
Graphics Card
Jun 2020
Release Date
Nov 2018
Tesla
Generation
Radeon Instinct
AI GPU
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
765 MHz
Base Clock
1200 MHz
1410 MHz
Boost Clock
1746 MHz
1215 MHz
Memory Clock
1000 MHz
Memory
40GB
Memory Size
16GB
HBM2e
Memory Type
HBM2
5120bit
Memory Bus
4096bit
1555GB/s
Bandwidth
1.02TB/s
Render Config
-
Compute Units
60
108
SM Count
48
6912
Shading Units
3840
432
TMUs
240
160
ROPs
64
432
Tensor Cores
192
-
RT Cores
48
192 KB (per SM)
L1 Cache
16 KB (per CU)
40 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
225.6 GPixel/s
Pixel Rate
111.7 GPixel/s
609.1 GTexel/s
Texture Rate
419.0 GTexel/s
77.97 TFLOPS
FP16 (half)
26.82 TFLOPS
19.49 TFLOPS
FP32 (float)
13.41 TFLOPS
9.746 TFLOPS
FP64 (double)
6.705 TFLOPS
Board Design
250W
TDP
300W
600 W
Suggested PSU
700 W
No outputs
Outputs
1x mini-DisplayPort 1.4a
8-pin EPS
Power Connectors
2x 8-pin
Graphics Processor
GA100
GPU Name
Vega 20
-
GPU Variant
Vega 20 GLXT
Ampere
Architecture
GCN 5.1
TSMC
Foundry
TSMC
7 nm
Process Size
7 nm
54.2 billion
Transistors
13.23 billion
826 mm²
Die Size
331 mm²
Graphics Features
N/A
DirectX
12 (12_1)
N/A
OpenGL
4.6
3.0
OpenCL
2.1
N/A
Vulkan
1.3
8.0
CUDA
8.9
N/A
Shader Model
6.7
