GPU Comparison
GEFORCE
NVIDIA Tesla K20m
CORE STATE
GK110
VRAM
5 GB
CLOCK SPEED
—
TDP
225 W
BUS WIDTH
320 bit
ARCHITECTURE
Kepler
PROCESS
28 nm
LAUNCH DATE
2013
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
Tesla K20m
Tesla T4
Core Specs
Shading Units
2,496
2,560
+2.6%
Shaders
2,496
2,560
+2.6%
TMUs
208
160
-23.1%
ROPs
40
64
+60.0%
SM Count
—
40
Clocks
Base Clock
—
585 MHz
Boost Clock
—
1590 MHz
GPU Clock
706 MHz
—
Memory Clock
1300 MHz
5.2 Gbps effective
1250 MHz
10 Gbps effective
Memory
Memory Size
5 GB
16 GB
VRAM (MB)
5,120
16,384
+220.0%
Memory Type
GDDR5
GDDR6
Memory Bus
320 bit
256 bit
Bandwidth
208.0 GB/s
320.0 GB/s
Cache
L1 Cache
16 KB (per SMX)
64 KB (per SM)
L2 Cache
1280 KB
4 MB
Performance
Pixel Rate
36.71 GPixel/s
101.8 GPixel/s
Texture Rate
146.8 GTexel/s
254.4 GTexel/s
FP32 (TFLOPS)
3.524 TFLOPS
8.141 TFLOPS
FP64 (TFLOPS)
1,174.8 GFLOPS (1:3)
254.4 GFLOPS (1:32)
FP16 (TFLOPS)
—
65.13 TFLOPS (8:1)
AI/RT
RT Cores
—
40
Tensor Cores
—
320
Power
TDP
225 W
70 W
TDP (W)
225
70
-68.9%
Suggested PSU
550 W
250 W
Power Connectors
1x 6-pin + 1x 8-pin
None
Architecture
Architecture
Kepler
Turing
GPU Name
GK110
TU104
Generation
Tesla Kepler
(Kxx)
Tesla Turing
(Txx)
Process Size
28 nm
12 nm
Transistors
7,080 million
13,600 million
Die Size
561 mm²
545 mm²
Foundry
TSMC
TSMC
Density
12.6M / mm²
25.0M / mm²
API Support
DirectX
12 (11_0)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.2.175
1.4
OpenCL
3.0
3.0
CUDA
3.5
7.5
Shader Model
6.5 (5.1)
6.8
Physical
Slot Width
Dual-slot
Single-slot
Length
267 mm
10.5 inches
168 mm
6.6 inches
Outputs
No outputs
No outputs
Bus Interface
PCIe 2.0 x16
PCIe 3.0 x16
Other
Launch Price
3,199 USD
—
Production
End-of-life
End-of-life
Predecessor
Tesla Fermi
Tesla Volta
Successor
Tesla Maxwell
Server Ampere