AMD FireStream 9370 vs NVIDIA PG506 232
We compared two Desktop platform GPUs: 4GB VRAM FireStream 9370 and 24GB VRAM PG506 232 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA PG506 232 's Advantages
Released 10 years and 10 months late
Boost Clock1440MHz
More VRAM (24GB vs 4GB)
Larger VRAM bandwidth (933.1GB/s vs 147.2GB/s)
1984 additional rendering cores
Lower TDP (165W vs 225W)
Score
Benchmark
FP32 (float)
FireStream 9370
2.64 TFLOPS
PG506 232
+290%
10.32 TFLOPS
Graphics Card
Jun 2010
Release Date
Apr 2021
FireStream
Generation
Tesla
Desktop
Type
Desktop
PCIe 2.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
-
Base Clock
930 MHz
-
Boost Clock
1440 MHz
1150 MHz
Memory Clock
1215 MHz
Memory
4GB
Memory Size
24GB
GDDR5
Memory Type
HBM2
256bit
Memory Bus
3072bit
147.2GB/s
Bandwidth
933.1GB/s
Render Config
20
Compute Units
-
-
SM Count
56
1600
Shading Units
3584
80
TMUs
224
32
ROPs
96
-
Tensor Cores
224
-
-
-
8 KB (per CU)
L1 Cache
192 KB (per SM)
512 KB
L2 Cache
24 MB
-
-
-
Theoretical Performance
26.40 GPixel/s
Pixel Rate
138.2 GPixel/s
66.00 GTexel/s
Texture Rate
322.6 GTexel/s
-
FP16 (half)
10.32 TFLOPS
2.640 TFLOPS
FP32 (float)
10.32 TFLOPS
528.0 GFLOPS
FP64 (double)
5.161 TFLOPS
Board Design
225W
TDP
165W
550 W
Suggested PSU
450 W
1x DisplayPort 1.1
Outputs
No outputs
1x 8-pin
Power Connectors
8-pin EPS
Graphics Processor
Cypress
GPU Name
GA100
Cypress XT GL
GPU Variant
-
TeraScale 2
Architecture
Ampere
TSMC
Foundry
TSMC
40 nm
Process Size
7 nm
2.154 billion
Transistors
54.2 billion
334 mm²
Die Size
826 mm²
Graphics Features
11.2 (11_0)
DirectX
N/A
4.4
OpenGL
N/A
1.2
OpenCL
3.0
N/A
Vulkan
N/A
-
CUDA
8.0
5.0
Shader Model
N/A
