Report an Error

NVIDIA Tesla K20Xm

Graphics Processor
GK110
Cores
2688
TMUs
224
ROPs
48
Memory Size
6 GB
Memory Type
GDDR5
Bus Width
384 bit
GPU Chip
GPU
The Tesla K20Xm was an enthusiast-class professional graphics card by NVIDIA, launched on November 12th, 2012. Built on the 28 nm process, and based on the GK110 graphics processor, the card supports DirectX 12. The GK110 graphics processor is a large chip with a die area of 561 mm² and 7,080 million transistors. It features 2688 shading units, 224 texture mapping units, and 48 ROPs. NVIDIA has paired 6 GB GDDR5 memory with the Tesla K20Xm, which are connected using a 384-bit memory interface. The GPU is operating at a frequency of 732 MHz, memory is running at 1300 MHz (5.2 Gbps effective).
Being a dual-slot card, its power draw is rated at 235 W maximum. This device has no display connectivity, as it is not designed to have monitors connected to it. Tesla K20Xm is connected to the rest of the system using a PCI-Express 3.0 x16 interface. The card measures 267 mm in length, and features a dual-slot cooling solution. Its price at launch was 7699 US Dollars.

Graphics Processor

GPU Name
GK110
Architecture
Kepler
Foundry
TSMC
Process Size
28 nm
Transistors
7,080 million
Density
12.6M / mm²
Die Size
561 mm²
Chip Package
BGA-2152

Graphics Card

Release Date
Nov 12th, 2012
Generation
Tesla Kepler
(Kxx)
Predecessor
Tesla Fermi
Successor
Tesla Maxwell
Production
End-of-life
Launch Price
7,699 USD
Bus Interface
PCIe 3.0 x16

Relative Performance

GeForce 210
3%
GeForce 9400 GT
3%
Radeon HD 4550
4%
Radeon HD 5450
5%
Radeon HD 6450
6%
GeForce GT 520
7%
GeForce GT 220
8%
GeForce GT 430
11%
Radeon HD 5570
11%
Radeon HD 4670
12%
GeForce GT 440
12%
GeForce GT 240
13%
GeForce 9600 GT
15%
Radeon HD 5670
15%
GeForce GT 640
17%
GeForce 9800 GT
17%
Radeon HD 6670
18%
Radeon HD 4830
18%
Radeon HD 4770
20%
Radeon HD 4850
21%
GeForce GTS 250
21%
GeForce GTS 450
22%
Radeon HD 5750
23%
Radeon HD 7750
26%
Radeon HD 5770
26%
GeForce GTX 260
26%
Radeon HD 4870
27%
Radeon Vega 8
27%
GeForce GTX 550 Ti
27%
GeForce GTX 650
27%
Radeon HD 5830
29%
Radeon HD 6790
30%
Radeon HD 4890
31%
GeForce GTX 460
32%
Radeon RX Vega 11
32%
GeForce GT 1030
33%
GeForce GTX 275
33%
Radeon HD 7770 GHz Edition
33%
GeForce GTX 280
33%
GeForce GTX 465
34%
Radeon HD 6850
35%
GeForce GTX 285
36%
Radeon HD 5850
39%
GeForce GTX 650 Ti
39%
Radeon HD 7790
39%
Radeon RX 550
41%
GeForce GTX 470
41%
Radeon HD 6870
43%
Radeon HD 5870
45%
GeForce GTX 560 Ti
47%
Radeon HD 4870 X2
47%
GeForce GTX 750 Ti
48%
Radeon HD 6950
48%
GeForce GTX 295
50%
GeForce GTX 650 Ti Boost
50%
Radeon HD 7850
51%
GeForce GTX 480
52%
Radeon HD 6970
54%
Radeon R7 265
55%
Radeon RX 460
56%
Radeon R7 370
56%
GeForce GTX 660
56%
GeForce GTX 570
56%
Radeon RX 560
57%
GeForce GTX 950
61%
Radeon HD 7870 GHz Edition
61%
Radeon R9 270X
63%
GeForce GTX 660 Ti
64%
GeForce GTX 580
64%
GeForce GTX 1050
64%
Radeon HD 7950
65%
GeForce GTX 1630
66%
Radeon HD 5970
66%
GeForce GTX 760
68%
GeForce GTX 670
73%
GeForce GTX 960
75%
Radeon R9 380
77%
Radeon R9 285
77%
Radeon HD 7970
78%
GeForce GTX 680
79%
GeForce GTX 1050 Ti
81%
GeForce GTX 770
83%
Radeon HD 6990
85%
Radeon R9 280X
87%
GeForce GTX 590
88%
Radeon HD 7970 GHz Edition
88%
GeForce GTX 780
97%
Arc A380
98%
Radeon RX 6400
99%
Tesla K20Xm
100%
GeForce GTX 1650
101%
Radeon RX 470
107%
Radeon R9 290
109%
Radeon RX 570
115%
Radeon R9 390
116%
Radeon R9 290X
117%
GeForce GTX 970
118%
GeForce GTX TITAN
120%
GeForce GTX 780 Ti
121%
Radeon R9 390X
123%
Radeon HD 7990
124%
Radeon RX 480
124%
GeForce GTX 690
125%
GeForce GTX 1060 6 GB
129%
Radeon RX 5500 OEM
130%
Radeon RX 580
132%
Radeon RX 5500 XT
133%
Radeon RX 6500 XT
134%
GeForce GTX 980
135%
GeForce GTX 1650 SUPER
136%
Radeon R9 FURY
136%
Radeon RX 590
142%
Radeon R9 295X2
150%
GeForce GTX 1660
151%
Radeon R9 FURY X
153%
GeForce GTX 980 Ti
155%
GeForce GTX TITAN X
159%
GeForce GTX 1660 SUPER
164%
GeForce GTX 1070
173%
GeForce GTX 1660 Ti
173%
GeForce RTX 3050 8 GB
176%
Radeon RX Vega 56
184%
GeForce GTX 1070 Ti
196%
Radeon RX 5600 XT
198%
Radeon RX Vega 64
201%
GeForce GTX 1080
204%
GeForce RTX 2060
204%
Radeon RX 5700
212%
Arc A580
213%
Radeon RX 6600
222%
GeForce RTX 2060 SUPER
229%
GeForce RTX 2070
232%
Radeon RX 5700 XT
232%
GeForce RTX 3060 12 GB
237%
Arc A750
245%
Radeon VII
246%
TITAN X Pascal
252%
Radeon RX 6600 XT
255%
GeForce GTX 1080 Ti
260%
Arc A770
262%
GeForce RTX 2070 SUPER
263%
Radeon RX 6650 XT
276%
Radeon RX 7600
279%
GeForce RTX 4060
281%
GeForce RTX 2080
282%
Radeon RX 7600 XT
288%
GeForce RTX 2080 SUPER
296%
GeForce RTX 3060 Ti
303%
Radeon RX 6700 XT
314%
Radeon RX 6750 XT
333%
GeForce RTX 4060 Ti 8 GB
337%
GeForce RTX 2080 Ti
339%
GeForce RTX 3070
355%
GeForce RTX 3070 Ti
380%
Radeon RX 7700 XT
384%
Radeon RX 6800
392%
GeForce RTX 4070
434%
Radeon RX 6800 XT
445%
GeForce RTX 3080
462%
Radeon RX 7800 XT
466%
Radeon RX 6900 XT
479%
GeForce RTX 4070 SUPER
503%
Radeon RX 7900 GRE
513%
GeForce RTX 3080 Ti
517%
Radeon RX 6950 XT
518%
GeForce RTX 3090
527%
GeForce RTX 4070 Ti
540%
GeForce RTX 4070 Ti SUPER
590%
Radeon RX 7900 XT
606%
GeForce RTX 3090 Ti
611%
GeForce RTX 4080
686%
GeForce RTX 4080 SUPER
696%
Radeon RX 7900 XTX
703%
GeForce RTX 4090
862%
Based on TPU review data: "Performance Summary" at 1920x1080, 4K for 2080 Ti and faster.
Performance estimated based on architecture, shader count and clocks.

Clock Speeds

GPU Clock
732 MHz
Memory Clock
1300 MHz
5.2 Gbps effective

Memory

Memory Size
6 GB
Memory Type
GDDR5
Memory Bus
384 bit
Bandwidth
249.6 GB/s

Render Config

Shading Units
2688
TMUs
224
ROPs
48
SMX Count
14
L1 Cache
16 KB (per SMX)
L2 Cache
1536 KB

Theoretical Performance

Pixel Rate
40.99 GPixel/s
Texture Rate
164.0 GTexel/s
FP32 (float)
3.935 TFLOPS
FP64 (double)
1,312 GFLOPS (1:3)

Board Design

Slot Width
Dual-slot
Length
267 mm
10.5 inches
TDP
235 W
Suggested PSU
550 W
Outputs
No outputs
Board Number
P2081 SKU 200

Graphics Features

DirectX
12 (11_0)
OpenGL
4.6
OpenCL
3.0
Vulkan
1.2.175
CUDA
3.5
Shader Model
6.5 (5.1)

GK110 GPU Notes

NVENC: 1st Gen
NVDEC: 1st Gen
PureVideo HD: VP5
VDPAU: Feature Set D

L1 Cache is configurable from 16 KB up to 48 KB per SMX

GK110 has 5 GPCs each capable of 8 pixels per clock. This limits complete GPU to 40 pixels per clock and because of that it can't feed all 48 ROPs when they all require data at the same time.
Additional ROPs however can be used for MSAA (because it doesn't require additional data from rasterizers, while giving more work to ROPs.)

Latest Drivers:
Windows XP / Server 2003 x64:
Quadro Release R319 U2 (321.01)

Windows Vista:
GeForce Release 365.19
Quadro Release R346 U7 (348.40) / R352 BETA (352.86)

Windows 7 / 8 / 8.1 / 10 / 11 (x32 / x64):
GeForce Release 391.35 / 475.06
Quadro Release R390 U9 (392.37) / R470 U16 (474.82)
Data Center Release 427.11
Tesla Release 441.22

Windows 10 / 11 x64:
GeForce Release 475.06
Quadro Release R470 U16 (474.82)
Data Center Release 474.82
Nov 2nd, 2024 18:48 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts