NVIDIA PG506 232 vs Moore Threads MTT S4000
We compared two Desktop platform GPUs: 24GB VRAM PG506 232 and 48GB VRAM Moore Threads MTT S4000 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA PG506 232 's Advantages
Boost Clock1440MHz
Larger VRAM bandwidth (933.1GB/s vs 768.0GB/s)
Lower TDP (165W vs 450W)
Moore Threads MTT S4000 's Advantages
More VRAM (48GB vs 24GB)
4608 additional rendering cores
Score
Benchmark
FP32 (float)
PG506 232
10.32 TFLOPS
Moore Threads MTT S4000
+138%
24.58 TFLOPS
Graphics Card
Apr 2021
Release Date
Unknown
Tesla
Generation
MUSA-Chunxiao
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 5.0 x16
Clock Speeds
930 MHz
Base Clock
1500 MHz
1440 MHz
Boost Clock
-
1215 MHz
Memory Clock
2000 MHz
Memory
24GB
Memory Size
48GB
HBM2
Memory Type
GDDR6
3072bit
Memory Bus
384bit
933.1GB/s
Bandwidth
768.0GB/s
Render Config
-
Compute Units
64
56
SM Count
-
3584
Shading Units
8192
224
TMUs
512
96
ROPs
512
224
Tensor Cores
128
-
-
-
192 KB (per SM)
L1 Cache
-
24 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
138.2 GPixel/s
Pixel Rate
768.0 GPixel/s
322.6 GTexel/s
Texture Rate
768.0 GTexel/s
10.32 TFLOPS
FP16 (half)
49.15 TFLOPS
10.32 TFLOPS
FP32 (float)
24.58 TFLOPS
5.161 TFLOPS
FP64 (double)
384.0 GFLOPS
Board Design
165W
TDP
450W
450 W
Suggested PSU
850 W
No outputs
Outputs
4x DisplayPort 1.4a
8-pin EPS
Power Connectors
2x 8-pin
Graphics Processor
GA100
GPU Name
MTT Gen 2
-
-
-
Ampere
Architecture
MUSA-Chunxiao
TSMC
Foundry
TSMC
7 nm
Process Size
12 nm
54.2 billion
Transistors
22 billion
826 mm²
Die Size
Unknown
Graphics Features
N/A
DirectX
12 Ultimate (12_2)
N/A
OpenGL
4.6
3.0
OpenCL
3.0
N/A
Vulkan
1.4
8.0
CUDA
-
N/A
Shader Model
6.5
