NVIDIA L20 vs Moore Threads MTT S3000
We compared two Desktop platform GPUs: 48GB VRAM L20 and 32GB VRAM Moore Threads MTT S3000 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA L20 's Advantages
Boost Clock2520MHz
More VRAM (48GB vs 32GB)
Larger VRAM bandwidth (864.0GB/s vs 448.0GB/s)
7680 additional rendering cores
Moore Threads MTT S3000 's Advantages
Lower TDP (250W vs 275W)
Score
Benchmark
FP32 (float)
L20
+281%
59.35 TFLOPS
Moore Threads MTT S3000
15.56 TFLOPS
Graphics Card
Nov 2023
Release Date
Unknown
Tesla Ada
Generation
MUSA-Chunxiao
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 5.0 x16
Clock Speeds
1440 MHz
Base Clock
1900 MHz
2520 MHz
Boost Clock
0 MHz
2250 MHz
Memory Clock
1750 MHz
Memory
48GB
Memory Size
32GB
GDDR6
Memory Type
GDDR6
384bit
Memory Bus
256bit
864.0GB/s
Bandwidth
448.0GB/s
Render Config
-
Compute Units
32
92
SM Count
-
11776
Shading Units
4096
368
TMUs
256
128
ROPs
256
368
Tensor Cores
-
92
RT Cores
-
128 KB (per SM)
L1 Cache
-
96 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
322.6 GPixel/s
Pixel Rate
486.4 GPixel/s
927.4 GTexel/s
Texture Rate
486.4 GTexel/s
59.35 TFLOPS
FP16 (half)
31.13 TFLOPS
59.35 TFLOPS
FP32 (float)
15.56 TFLOPS
927.4 GFLOPS
FP64 (double)
243.2 GFLOPS
Board Design
275W
TDP
250W
600 W
Suggested PSU
600 W
4x DisplayPort 1.4a
Outputs
2x DisplayPort 1.4a
1x 16-pin
Power Connectors
1x 8-pin
Graphics Processor
AD102
GPU Name
MTT Gen 2
AD102
GPU Variant
-
Ada Lovelace
Architecture
MUSA-Chunxiao
TSMC
Foundry
TSMC
5 nm
Process Size
12 nm
76.3 billion
Transistors
22 billion
609 mm²
Die Size
Unknown
Graphics Features
12 Ultimate (12_2)
DirectX
12 Ultimate (12_2)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.4
8.9
CUDA
-
6.7
Shader Model
6.5
