NVIDIA CMP 40HX vs NVIDIA P104 100
We compared two Desktop platform GPUs: 8GB VRAM CMP 40HX and 4GB VRAM P104 100 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA CMP 40HX 's Advantages
Released 3 years and 2 months late
More VRAM (8GB vs 4GB)
Larger VRAM bandwidth (448.0GB/s vs 320.3GB/s)
384 additional rendering cores
NVIDIA P104 100 's Advantages
Boost Clock has increased by 5% (1733MHz vs 1650MHz)
Score
Benchmark
FP32 (float)
CMP 40HX
+14%
7.603 TFLOPS
P104 100
6.655 TFLOPS
Blender
CMP 40HX
+67%
1174
P104 100
701
Graphics Card
Feb 2021
Release Date
Dec 2017
Mining GPUs
Generation
Mining GPUs
Desktop
Type
Desktop
PCIe 1.0 x4
Bus Interface
PCIe 1.0 x4
Clock Speeds
1470 MHz
Base Clock
1607 MHz
1650 MHz
Boost Clock
1733 MHz
1750 MHz
Memory Clock
1251 MHz
Memory
8GB
Memory Size
4GB
GDDR6
Memory Type
GDDR5X
256bit
Memory Bus
256bit
448.0GB/s
Bandwidth
320.3GB/s
Render Config
-
-
-
36
SM Count
15
2304
Shading Units
1920
144
TMUs
120
64
ROPs
64
288
Tensor Cores
-
36
RT Cores
-
64 KB (per SM)
L1 Cache
48 KB (per SM)
4 MB
L2 Cache
2 MB
-
-
-
Theoretical Performance
105.6 GPixel/s
Pixel Rate
110.9 GPixel/s
237.6 GTexel/s
Texture Rate
208.0 GTexel/s
15.21 TFLOPS
FP16 (half)
104.0 GFLOPS
7.603 TFLOPS
FP32 (float)
6.655 TFLOPS
237.6 GFLOPS
FP64 (double)
208.0 GFLOPS
Board Design
185W
TDP
Unknown
450 W
Suggested PSU
200 W
No outputs
Outputs
No outputs
1x 8-pin
Power Connectors
1x 8-pin
Graphics Processor
TU106
GPU Name
GP104
TU106-100-A1
GPU Variant
GP104-100-A1
Turing
Architecture
Pascal
TSMC
Foundry
TSMC
12 nm
Process Size
16 nm
10.8 billion
Transistors
7.2 billion
445 mm²
Die Size
314 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (12_1)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.3
7.5
CUDA
6.1
6.8
Shader Model
6.8
