GPU Comparison
GEFORCE
NVIDIA PG506-232
CORE STATE
GA100
VRAM
24 GB
CLOCK SPEED
1440 MHz
TDP
165 W
BUS WIDTH
3072 bit
ARCHITECTURE
Ampere
PROCESS
7 nm
LAUNCH DATE
2021
VS
GEFORCE
Tesla T4
CORE STATE
TU104
VRAM
16 GB
CLOCK SPEED
1590 MHz
TDP
70 W
BUS WIDTH
256 bit
ARCHITECTURE
Turing
PROCESS
12 nm
LAUNCH DATE
2018
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
PG506-232
Tesla T4
Core Specs
Shading Units
3,584
2,560
-28.6%
Shaders
3,584
2,560
-28.6%
TMUs
224
160
-28.6%
ROPs
96
64
-33.3%
SM Count
56
40
-28.6%
Clocks
Base Clock
930 MHz
585 MHz
Boost Clock
1440 MHz
1590 MHz
Memory Clock
1215 MHz
2.4 Gbps effective
1250 MHz
10 Gbps effective
Memory
Memory Size
24 GB
16 GB
VRAM (MB)
24,576
16,384
-33.3%
Memory Type
HBM2
GDDR6
Memory Bus
3072 bit
256 bit
Bandwidth
933.1 GB/s
320.0 GB/s
Cache
L1 Cache
192 KB (per SM)
64 KB (per SM)
L2 Cache
24 MB
4 MB
Performance
Pixel Rate
138.2 GPixel/s
101.8 GPixel/s
Texture Rate
322.6 GTexel/s
254.4 GTexel/s
FP32 (TFLOPS)
10.32 TFLOPS
8.141 TFLOPS
FP64 (TFLOPS)
5.161 TFLOPS (1:2)
254.4 GFLOPS (1:32)
FP16 (TFLOPS)
10.32 TFLOPS (1:1)
65.13 TFLOPS (8:1)
AI/RT
RT Cores
—
40
Tensor Cores
224
320
+42.9%
Power
TDP
165 W
70 W
TDP (W)
165
70
-57.6%
Suggested PSU
450 W
250 W
Power Connectors
8-pin EPS
None
Architecture
Architecture
Ampere
Turing
GPU Name
GA100
TU104
Generation
Server Ampere
(Axx)
Tesla Turing
(Txx)
Process Size
7 nm
12 nm
Transistors
54,200 million
13,600 million
Die Size
826 mm²
545 mm²
Foundry
TSMC
TSMC
Density
65.6M / mm²
25.0M / mm²
API Support
DirectX
—
12 Ultimate (12_2)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
8.0
7.5
Shader Model
—
6.8
Physical
Slot Width
Dual-slot
Single-slot
Length
267 mm
10.5 inches
168 mm
6.6 inches
Height
112 mm
4.4 inches
—
Outputs
No outputs
No outputs
Bus Interface
PCIe 4.0 x16
PCIe 3.0 x16
Other
Production
End-of-life
End-of-life
Predecessor
Tesla Turing
Tesla Volta
Successor
Server Ada
Server Ampere