NVIDIA P104 100 vs NVIDIA A40 PCIe
We compared two Desktop platform GPUs: 4GB VRAM P104 100 and 48GB VRAM A40 PCIe to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA A40 PCIe 's Advantages
Released 2 years and 10 months late
Boost Clock has increased by 0% (1740MHz vs 1733MHz)
More VRAM (48GB vs 4GB)
Larger VRAM bandwidth (695.8GB/s vs 320.3GB/s)
8832 additional rendering cores
Score
Benchmark
FP32 (float)
P104 100
6.655 TFLOPS
A40 PCIe
+462%
37.42 TFLOPS
Blender
P104 100
701
A40 PCIe
+469%
3990
Graphics Card
Dec 2017
Release Date
Oct 2020
Mining GPUs
Generation
Tesla
Desktop
Type
Desktop
PCIe 1.0 x4
Bus Interface
PCIe 4.0 x16
Clock Speeds
1607 MHz
Base Clock
1305 MHz
1733 MHz
Boost Clock
1740 MHz
1251 MHz
Memory Clock
1812 MHz
Memory
4GB
Memory Size
48GB
GDDR5X
Memory Type
GDDR6
256bit
Memory Bus
384bit
320.3GB/s
Bandwidth
695.8GB/s
Render Config
-
-
-
15
SM Count
84
1920
Shading Units
10752
120
TMUs
336
64
ROPs
112
-
Tensor Cores
336
-
RT Cores
84
48 KB (per SM)
L1 Cache
128 KB (per SM)
2 MB
L2 Cache
6 MB
-
-
-
Theoretical Performance
110.9 GPixel/s
Pixel Rate
194.9 GPixel/s
208.0 GTexel/s
Texture Rate
584.6 GTexel/s
104.0 GFLOPS
FP16 (half)
37.42 TFLOPS
6.655 TFLOPS
FP32 (float)
37.42 TFLOPS
208.0 GFLOPS
FP64 (double)
584.6 GFLOPS
Board Design
Unknown
TDP
300W
200 W
Suggested PSU
700 W
No outputs
Outputs
3x DisplayPort 1.4a
1x 8-pin
Power Connectors
8-pin EPS
Graphics Processor
GP104
GPU Name
GA102
GP104-100-A1
GPU Variant
-
Pascal
Architecture
Ampere
TSMC
Foundry
Samsung
16 nm
Process Size
8 nm
7.2 billion
Transistors
28.3 billion
314 mm²
Die Size
628 mm²
Graphics Features
12 (12_1)
DirectX
12 Ultimate (12_2)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.3
6.1
CUDA
8.6
6.8
Shader Model
6.6
