GeForce RTX 3090 and A100 PCIe - video card comparison
GeForce RTX 3090
A100 PCIe
Architecture
Ampere (2020−2022)
Ampere (2020−2022)
GPU code name
Ampere GA102
GA100
Market segment
Desktop
Desktop
Release date
September 2020
June 2020
Interface
PCIe 4.0 x16
PCIe 4.0 x16
Supplementary power connectors
1x 12-pin
1x 8-pin
Boost clock speed
1700 MHz
-
Core clock speed
1400 MHz
1410 MHz
Manufacturing process technology
8 nm
7 nm
Number of transistors
28,300 million
54,200 million
Pipelines / CUDA cores
10496
6912
Power consumption (TDP)
350 Watt
250 Watt
Texture fill rate
556.0
609.1
Memory type
GDDR6X
HBM2e
Maximum RAM amount
24 GB
40 GB
Memory bus width
384 Bit
5120 Bit
Memory clock speed
19500 MHz
2.4 GB/s
Memory bandwidth
936.2 GB/s
1,555 GB/s
Shared memory
-
-
Display Connectors
1x HDMI, 3x DisplayPort
No outputs
HDMI
+
-
DirectX
12 Ultimate (12_2)
12 Ultimate (12_2)
CUDA
8.5
8.0
OpenCL
2.0
2.0
OpenGL
4.6
4.6
Shader Model
6.5
6.5
Vulkan
1.2
1.2