Home GPU Comparison NVIDIA H100 PCIe vs NVIDIA A10G

NVIDIA H100 PCIe vs NVIDIA A10G

AI GPU We compared a GPU: 80GB VRAM H100 PCIe and a Professional market GPU: 12GB VRAM A10G to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.

Main Differences

NVIDIA H100 PCIe 's Advantages
Released 11 months late
Boost Clock has increased by 3% (1755MHz vs 1710MHz)
More VRAM (80GB vs 12GB)
Larger VRAM bandwidth (2039GB/s vs 600.2GB/s)
5376 additional rendering cores
NVIDIA A10G 's Advantages
Lower TDP (150W vs 350W)

Score

Benchmark

FP32 (float)
H100 PCIe +62%
51.22 TFLOPS
A10G
31.52 TFLOPS
Blender
H100 PCIe +32%
4845
A10G
3664
VS

Graphics Card

Mar 2022
Release Date
Apr 2021
Tesla Hopper
Generation
Tesla
AI GPU
Type
Professional
PCIe 5.0 x16
Bus Interface
PCIe 4.0 x16

Clock Speeds

1095 MHz
Base Clock
1320 MHz
1755 MHz
Boost Clock
1710 MHz
1593 MHz
Memory Clock
1563 MHz

Memory

80GB
Memory Size
12GB
HBM2e
Memory Type
GDDR6
5120bit
Memory Bus
384bit
2039GB/s
Bandwidth
600.2GB/s

Render Config

114
SM Count
72
-
Compute Units
-
14592
Shading Units
9216
456
TMUs
288
24
ROPs
96
456
Tensor Cores
288
-
RT Cores
72
256 KB (per SM)
L1 Cache
128 KB (per SM)
50 MB
L2 Cache
6 MB

Theoretical Performance

42.12 GPixel/s
Pixel Rate
164.2 GPixel/s
800.3 GTexel/s
Texture Rate
492.5 GTexel/s
204.9 TFLOPS
FP16 (half)
31.52 TFLOPS
51.22 TFLOPS
FP32 (float)
31.52 TFLOPS
25.61 TFLOPS
FP64 (double)
985.0 GFLOPS

Graphics Processor

GH100
GPU Name
GA102
-
GPU Variant
GA102-890-A1
Hopper
Architecture
Ampere
TSMC
Foundry
Samsung
4 nm
Process Size
8 nm
80 billion
Transistors
28.3 billion
814 mm²
Die Size
628 mm²

Board Design

350W
TDP
150W
750 W
Suggested PSU
450 W
No outputs
Outputs
No outputs
1x 16-pin
Power Connectors
8-pin EPS

Graphics Features

N/A
DirectX
12 Ultimate (12_2)
N/A
OpenGL
4.6
3.0
OpenCL
3.0
N/A
Vulkan
1.3
9.0
CUDA
8.6
N/A
Shader Model
6.6

Related GPU Comparisons

© 2024 - TopCPU.net   Contact Us Privacy Policy