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 K20c
CORE STATE
GK110
VRAM
5 GB
CLOCK SPEED
—
TDP
225 W
BUS WIDTH
320 bit
ARCHITECTURE
Kepler
PROCESS
28 nm
LAUNCH DATE
2012
PERFORMANCE BENCHMARKS
geekbench_opencl
DETAILED SPECIFICATIONS
SPECIFICATION
PG506-232
Tesla K20c
Core Specs
Shading Units
3,584
2,496
-30.4%
Shaders
3,584
2,496
-30.4%
TMUs
224
208
-7.1%
ROPs
96
40
-58.3%
SM Count
56
—
Clocks
Base Clock
930 MHz
—
Boost Clock
1440 MHz
—
GPU Clock
—
706 MHz
Memory Clock
1215 MHz
2.4 Gbps effective
1300 MHz
5.2 Gbps effective
Memory
Memory Size
24 GB
5 GB
VRAM (MB)
24,576
5,120
-79.2%
Memory Type
HBM2
GDDR5
Memory Bus
3072 bit
320 bit
Bandwidth
933.1 GB/s
208.0 GB/s
Cache
L1 Cache
192 KB (per SM)
16 KB (per SMX)
L2 Cache
24 MB
1280 KB
Performance
Pixel Rate
138.2 GPixel/s
36.71 GPixel/s
Texture Rate
322.6 GTexel/s
146.8 GTexel/s
FP32 (TFLOPS)
10.32 TFLOPS
3.524 TFLOPS
FP64 (TFLOPS)
5.161 TFLOPS (1:2)
1,174.8 GFLOPS (1:3)
FP16 (TFLOPS)
10.32 TFLOPS (1:1)
—
AI/RT
Tensor Cores
224
—
Power
TDP
165 W
225 W
TDP (W)
165
225
+36.4%
Suggested PSU
450 W
550 W
Power Connectors
8-pin EPS
1x 6-pin + 1x 8-pin
Architecture
Architecture
Ampere
Kepler
GPU Name
GA100
GK110
Generation
Server Ampere
(Axx)
Tesla Kepler
(Kxx)
Process Size
7 nm
28 nm
Transistors
54,200 million
7,080 million
Die Size
826 mm²
561 mm²
Foundry
TSMC
TSMC
Density
65.6M / mm²
12.6M / mm²
API Support
DirectX
—
12 (11_0)
OpenGL
—
4.6
Vulkan
—
1.2.175
OpenCL
3.0
3.0
CUDA
8.0
3.5
Shader Model
—
6.5 (5.1)
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm
10.5 inches
267 mm
10.5 inches
Height
112 mm
4.4 inches
—
Outputs
No outputs
No outputs
Bus Interface
PCIe 4.0 x16
PCIe 2.0 x16
Other
Launch Price
—
3,199 USD
Production
End-of-life
End-of-life
Predecessor
Tesla Turing
Tesla Fermi
Successor
Server Ada
Tesla Maxwell