NVIDIA CMP 50HX vs NVIDIA GeForce GTX 660
We compared two Desktop platform GPUs: 10GB VRAM CMP 50HX and 2GB VRAM GeForce GTX 660 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA CMP 50HX 's Advantages
Released 8 years and 9 months late
Boost Clock has increased by 50% (1545MHz vs 1032MHz)
More VRAM (10GB vs 2GB)
Larger VRAM bandwidth (560.0GB/s vs 144.2GB/s)
2624 additional rendering cores
NVIDIA GeForce GTX 660 's Advantages
Lower TDP (140W vs 250W)
Score
Benchmark
FP32 (float)
CMP 50HX
+458%
11.07 TFLOPS
GeForce GTX 660
1.981 TFLOPS
Blender
CMP 50HX
+1146%
1384
GeForce GTX 660
111
Graphics Card
Jun 2021
Release Date
Sep 2012
Mining GPUs
Generation
GeForce 600
Desktop
Type
Desktop
PCIe 1.0 x4
Bus Interface
PCIe 3.0 x16
Clock Speeds
1350 MHz
Base Clock
980 MHz
1545 MHz
Boost Clock
1032 MHz
1750 MHz
Memory Clock
1502 MHz
Memory
10GB
Memory Size
2GB
GDDR6
Memory Type
GDDR5
320bit
Memory Bus
192bit
560.0GB/s
Bandwidth
144.2GB/s
Render Config
-
-
-
56
SM Count
-
3584
Shading Units
960
192
TMUs
80
80
ROPs
24
448
Tensor Cores
-
56
RT Cores
-
64 KB (per SM)
L1 Cache
16 KB (per SMX)
5 MB
L2 Cache
384 KB
-
-
-
Theoretical Performance
123.6 GPixel/s
Pixel Rate
20.64 GPixel/s
296.6 GTexel/s
Texture Rate
82.56 GTexel/s
22.15 TFLOPS
FP16 (half)
-
11.07 TFLOPS
FP32 (float)
1.981 TFLOPS
346.1 GFLOPS
FP64 (double)
82.56 GFLOPS
Board Design
250W
TDP
140W
600 W
Suggested PSU
300 W
No outputs
Outputs
2x DVI
1x HDMI 1.4a
1x DisplayPort 1.2
2x 8-pin
Power Connectors
1x 6-pin
Graphics Processor
TU102
GPU Name
GK106
TU102-100-A1
GPU Variant
GK106-400-A1
Turing
Architecture
Kepler
TSMC
Foundry
TSMC
12 nm
Process Size
28 nm
18.6 billion
Transistors
2.54 billion
754 mm²
Die Size
221 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (11_0)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.1
7.5
CUDA
3.0
6.8
Shader Model
5.1
