Report an Error

NVIDIA Tesla P40

Graphics Processor
GP102
Cores
3840
TMUs
240
ROPs
96
Memory Size
24 GB
Memory Type
GDDR5
Bus Width
384 bit
GPU Chip
GPU
The Tesla P40 was a data center GPU by NVIDIA, launched on September 13th, 2016. Built on the 16 nm process, and based on the GP102 graphics processor, the GPU supports DirectX 12. The GP102 graphics processor is a large chip with a die area of 471 mm² and 11.8 billion transistors. It features 3840 shading units, 240 texture mapping units, and 96 ROPs. NVIDIA has paired 24 GB GDDR5 memory with the Tesla P40, which is connected using a 384-bit memory interface. The GPU is operating at a frequency of 1303 MHz, which can be boosted up to 1531 MHz, memory is running at 1808 MHz (7.2 Gbps effective).
Being a dual-slot card, the NVIDIA Tesla P40 draws power from an 8-pin EPS power connector, with power draw rated at 250 W maximum. This card has no display connectivity, as it is not designed to have monitors connected to it. Tesla P40 is connected to the rest of the system using a PCI-Express 3.0 x16 interface. The card a dual-slot cooling solution. Its price at launch was 5,699 US Dollars.

Graphics Processor

GPU Name
GP102
Architecture
Pascal
Foundry
TSMC
Process Type
16FF+ FinFET
Process Size
16 nm
Transistors
11.8 billion
Density
25.1M / mm²
Die Size
471 mm²
Chip Package
FCBGA-2397
Package Size
45 mm × 45 mm

Graphics Card

Release Date
Sep 13th, 2016
Generation
Tesla Pascal
(Pxx)
Predecessor
Tesla Maxwell
Successor
Tesla Volta
Production
End-of-life
Driver Support
Legacy
Launch Price
5,699 USD
Bus Interface
PCIe 3.0 x16

Relative Performance

GeForce RTX 5090
489%
GeForce RTX 4090
373%
GeForce RTX 5080
322%
GeForce RTX 4080 SUPER
288%
GeForce RTX 4080
285%
Radeon RX 7900 XTX
285%
GeForce RTX 5070 Ti
281%
Radeon RX 9070 XT
269%
Radeon RX 7900 XT
246%
GeForce RTX 3090 Ti
241%
GeForce RTX 4070 Ti SUPER
241%
Radeon RX 9070
241%
GeForce RTX 4070 Ti
220%
GeForce RTX 5070
218%
GeForce RTX 3090
215%
GeForce RTX 3080 Ti
212%
GeForce RTX 4070 SUPER
205%
Radeon RX 7900 GRE
202%
Radeon RX 9070 GRE
200%
Radeon RX 6950 XT
199%
GeForce RTX 4070
196%
Radeon RX 6900 XT
194%
GeForce RTX 3080
191%
Radeon RX 7800 XT
191%
Radeon RX 6800 XT
184%
GeForce RTX 5060 Ti 16 GB
171%
GeForce RTX 5060 Ti 8 GB
170%
Radeon RX 7700 XT
167%
Radeon RX 9060 XT 16 GB
162%
GeForce RTX 3070 Ti
162%
Radeon RX 6800
156%
Radeon RX 9060 XT 8 GB
153%
GeForce RTX 3070
151%
GeForce RTX 2080 Ti
150%
GeForce RTX 4060 Ti 8 GB
149%
GeForce RTX 5060
147%
Radeon RX 6750 XT
143%
GeForce RTX 3060 Ti
132%
Radeon RX 6700 XT
130%
GeForce RTX 2080 SUPER
121%
Radeon RX 7600 XT
120%
GeForce RTX 4060
119%
Arc B580
119%
GeForce RTX 5050
116%
GeForce RTX 2080
115%
Radeon RX 6650 XT
113%
Radeon RX 7600
113%
GeForce RTX 2070 SUPER
107%
GeForce GTX 1080 Ti
106%
Arc A770
106%
GeForce RTX 3060 12 GB
102%
Radeon RX 6600 XT
101%
Radeon VII
100%
Tesla P40
100%
Arc A750
98%
Steam Machine GPU
96%
Radeon RX 5700 XT
96%
GeForce RTX 2070
95%
Radeon RX 6600
90%
Arc A580
88%
GeForce RTX 2060 SUPER
88%
Radeon RX Vega 64
84%
GeForce RTX 2060
81%
Radeon RX 5700
81%
GeForce GTX 1080
79%
GeForce GTX 1070 Ti
78%
Radeon RX 5600 XT
77%
Radeon RX Vega 56
73%
GeForce GTX 1070
71%
GeForce GTX 1660 SUPER
67%
GeForce GTX 1660 Ti
67%
GeForce GTX 980 Ti
63%
GeForce RTX 3050 8 GB
63%
Radeon R9 FURY X
62%
GeForce GTX 1660
62%
Radeon RX 590
58%
Radeon R9 FURY
57%
GeForce GTX 980
56%
GeForce GTX 1650 SUPER
55%
Radeon RX 6500 XT
55%
Radeon RX 5500 XT
54%
Radeon RX 580
54%
GeForce GTX 1060 6 GB
52%
Radeon R9 390X
52%
GeForce GTX 690
51%
Radeon RX 480
51%
Radeon HD 7990
50%
GeForce GTX 780 Ti
50%
GeForce GTX 970
48%
Radeon R9 290X
48%
Radeon R9 390
47%
Radeon RX 570
47%
Radeon R9 290
44%
Radeon RX 470
44%
GeForce GTX 1650
41%
Radeon RX 6400
40%
Arc A380
40%
GeForce GTX 780
40%
Radeon HD 7970 GHz Edition
36%
GeForce GTX 590
36%
Radeon R9 280X
35%
Radeon HD 6990
35%
GeForce GTX 1050 Ti
33%
GeForce GTX 680
32%
Radeon R9 380
32%
Radeon HD 7970
32%
GeForce GTX 770
32%
Radeon R9 285
32%
GeForce GTX 960
31%
GeForce GTX 670
30%
Radeon HD 5970
27%
GeForce GTX 1630
27%
Radeon HD 7950
26%
GeForce GTX 1050
26%
GeForce GTX 760
26%
GeForce GTX 580
26%
GeForce GTX 660 Ti
26%
Radeon R9 270X
26%
Radeon HD 7870 GHz Edition
25%
GeForce GTX 950
25%
Radeon RX 560
23%
GeForce GTX 570
23%
GeForce GTX 660
23%
Radeon RX 460
23%
Radeon R7 370
23%
Radeon HD 6970
22%
GeForce GTX 480
21%
Radeon HD 7850
21%
GeForce GTX 295
20%
Radeon HD 6950
20%
GeForce GTX 750 Ti
20%
Radeon HD 4870 X2
19%
GeForce GTX 560 Ti
19%
Radeon HD 5870
19%
Radeon HD 6870
17%
GeForce GTX 470
17%
GeForce GTX 650 Ti
16%
Radeon HD 5850
16%
GeForce GTX 285
15%
Radeon HD 6850
14%
GeForce GTX 280
13%
Radeon HD 7770 GHz Edition
13%
GeForce GTX 275
13%
GeForce GT 1030
13%
GeForce GTX 460
13%
Radeon HD 4890
13%
Radeon HD 6790
12%
Radeon HD 5830
12%
GeForce GTX 650
11%
GeForce GTX 550 Ti
11%
Radeon HD 4870
11%
GeForce GTX 260
11%
Radeon HD 5770
11%
Radeon HD 7750
10%
Radeon HD 5750
9%
GeForce GTS 450
9%
GeForce GTS 250
9%
Radeon HD 4850
8%
Radeon HD 4770
8%
Radeon HD 4830
7%
Radeon HD 6670
7%
GeForce 9800 GT
7%
GeForce GT 640
7%
Radeon HD 5670
6%
GeForce 9600 GT
6%
GeForce GT 240
5%
GeForce GT 440
5%
Radeon HD 4670
5%
Radeon HD 5570
4%
GeForce GT 430
4%
GeForce GT 220
3%
GeForce GT 520
3%
Radeon HD 6450
3%
Radeon HD 5450
2%
Radeon HD 4550
2%
GeForce 9400 GT
1%
GeForce 210
1%
Based on TPU review data: "Performance Summary" at 1920x1080, 4K for RTX 3080 and faster.
Performance estimated based on architecture, shader count and clocks.

Clock Speeds

Base Clock
1303 MHz
Boost Clock
1531 MHz
Memory Clock
1808 MHz
7.2 Gbps effective

Memory

Memory Size
24 GB
Memory Type
GDDR5
Memory Bus
384 bit
Bandwidth
347.1 GB/s

Render Config

Shading Units
3840
TMUs
240
ROPs
96
SM Count
30
L1 Cache
48 KB (per SM)
L2 Cache
3 MB

Theoretical Performance

Pixel Rate
147.0 GPixel/s
Texture Rate
367.4 GTexel/s
FP16
183.7 GFLOPS (1:64)
FP32
11.76 TFLOPS
FP64
367.4 GFLOPS (1:32)

Numeric Format Support

Vector
INT4
INT8
INT16
INT32
FP16
BF16
FP32
FP64
Matrix
Not Supported

Board Design

Slot Width
Dual-slot
Length
267 mm
10.5 inches
Height
111 mm
4.4 inches
TDP
250 W
Suggested PSU
600 W
Outputs
No outputs
Power Connectors
8-pin EPS
Board Number
PG610 SKU 200

Graphics Features

DirectX
12 (12_1)
OpenGL
4.6
OpenCL
3.0
Vulkan
1.4
CUDA
6.1
Shader Model
6.8

GP102 GPU Notes

Super Resolution: FSR 3.1.5, XeSS 3
Frame Generation: FSR 3.1.6, XeSS 3
Ray Reconstruction: No Support
NVENC: 6th Gen
NVDEC: 3rd Gen
PureVideo HD: VP8
VDPAU: Feature Set H

Latest Supported Drivers:
Windows 7 / 8 / 8.1 (x32 / x64):
GeForce Release 391.35 / 474.89
Quadro Release R390 U9 (392.37) / R440 U4 (441.66)
Data Center Release 427.11

Windows 10 / 11 (x32 / x64):
GeForce Release 391.35 / 582.78
Quadro Release R390 U9 (392.37) / R580 U11 (582.78)
Data Center Release 582.78

Windows Server 2016 / 2019:
Data Center Release 529.19 / 539.72

Windows Server 2022 / 2025:
Data Center Release 582.78

Linux (X32 / X64):
GeForce Release 390.157 / 580.178.04
Quadro Release R390 (390.157) / R580 (580.178.04)
Data Center Release 580.178.04
Oct 28th, 2026 08:58 EDT change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts