NVIDIA GeForce RTX 3080 vs NVIDIA GRID A100A

What is the difference between NVIDIA GeForce RTX 3080 and NVIDIA GRID A100A. Find out which graphics card has better performance.

Graphics Processor (GPU)

GA102GPU NameGA100
AmpereArchitectureAmpere
SamsungFoundryTSMC
8 nmProcess Size7 nm
28,300 millionTransistors54,200 million
628 mm²Die Size826 mm²

Graphics Card

Sep 1st, 2020Release DateMay 14th, 2020
GeForce 30FamilyGRID
(Ax)
ActiveProductionActive
PCIe 4.0 x16Bus InterfacePCIe 4.0 x16

Memory

10 GBMemory Size48 GB
GDDR6XMemory TypeHBM2e
320 bitMemory Bus6144 bit
760.3 GB/sBandwidth1,866 GB/s

Performance

164.2 GPixel/sPixel fillrate193.0 GPixel/s
465.1 GTexel/sTexture fillrate434.2 GTexel/s
29.77 TFLOPS (1:1)FP16 (half) performance55.57 TFLOPS (4:1)
29.77 TFLOPSFP32 (float) performance13.89 TFLOPS
465.1 GFLOPS (1:64)FP64 (double) performance6.947 TFLOPS (1:2)

Clock Speeds

1440 MHzBase Clock900 MHz
1710 MHzBoost Clock1005 MHz
1188 MHz
19 Gbps effective
Memory Clock1215 MHz
2.4 Gbps effective

Render Config

8704Shading Units6912
272TMUs432
96ROPs192
128 KB (per SM)L1 Cache192 KB (per SM)
5 MBL2 Cache48 MB
68SM Count108
272Tensor Cores432

Board Design

Dual-slotSlot WidthIGP
320 WThermal design power (TDP)400 W
700 WSuggested PSU800 W
1x HDMI
3x DisplayPort
Display ConnectorsNo outputs
1x 12-pinPower ConnectorsNone

API support

12 Ultimate (12_2)DirectXN/A
4.6OpenGLN/A
3.0OpenCL3.0
1.2VulkanN/A
6.6Shader ModelN/A
8.6CUDA8.0