Nvidia GeForce 10 series

from Wikipedia, the free encyclopedia
Logo of the Nvidia GeForce 10 series
Nvidia GeForce GTX 1070 "Founders Edition"

The GeForce 10 series is a series of graphics chips from Nvidia and the successor to the GeForce 900 series . All graphics processors in this series are based on the Pascal architecture and support the Shader model 5.0 (SM 5.0) according to DirectX 12.1. It was replaced in the high-end segment by the 20 series , which also supports ray tracing , and for the mainstream area by the 16 series .

description

GeForce GTX 1080 Ti, Titan X and Titan Xp

MSI GeForce GT 1030 with passive cooling

On August 2, 2016 Nvidia introduced the successor of the Titanium X GeForce GTX Titan X ago. Since both cards are very similar in terms of their names (they only differ in that they omit the “GeForce GTX” prefix, which curiously still exists on the card itself), this caused some confusion. While the previous generation Titan X still used the GM200 GPU, the new version uses the GP102-400-A1 graphics processor.

The GP102-GPU represents the graphics processor of the Pascal generation for the enthusiast sector. It is 471 mm² in size, consists of around 12 billion transistors and is manufactured at TSMC using the 16 nm FinFET manufacturing process. From a technical point of view, the GP102 with 96 raster, 3840 shader and 240 texture units is a 50% larger variant of the GP104. In previous generations, Nvidia used a graphics processor for the high-end / enthusiast sector and the professional market, which is why GPUs like the GK110 have numerous HPC functions that are superfluous for 3D applications. Already in the last generation, Nvidia introduced a pure gamer chip for the enthusiast market with the GM200, due to the production-related limitations of the older 28 nm manufacturing technology. In the Pascal generation, these limitations disappeared again with the new 16 nm production, but Nvidia still continued to develop two different graphics processors for the professional and the enthusiast market (the GP100 and the GP102). Although both processors have the same number of shader, texture and raster units, they are vastly different. The GP100 uses a different cluster structure which differs from all other GPUs of the Pascal generation, has numerous HPC functions and has an HBM2 memory interface .

With the Titan X , two clusters are deactivated on the GP102-400-A1, which means that 3584 shader and 224 texture units are still available (the full configuration is initially only available on the Quadro P6000 ). The Titan X also has a 12 GB GDDR5X video memory that is connected via a 384-bit memory interface. Compared to the GeForce GTX 1080, the Titan X achieves around 25 percent higher performance under Full HD, which increases to around 31% under 4K. Corresponding to earlier Titan models, the Titan X again took the position as the fastest single GPU graphics card on the market, favored by the lack of AMD competing products. The fact that Nvidia did not sell the Titan X through board partners, but only through its own online shop for a list price of € 1299, caused criticism. After the Titan X had become obsolete with the GeForce GTX 1080 Ti, Nvidia replaced it with the Titan Xp . The GP102-450-A1 uses this in its full configuration and with slightly increased clock rates.

On March 9, 2017, Nvidia introduced the GeForce GTX 1080 Ti . This uses the GP102-350-K1-A1 graphics processor, which comes with two deactivated shader clusters (SMX). Thus, the GTX 1080 Ti has 3584 shader and 224 texture units, just like the Titan X before. In contrast to this, the video memory has also been cut: The GeForce GTX 1080 Ti still has 11 GB of GDDR5X-VRAM on a 352-bit memory interface. In combination with slightly higher clock rates, the GeForce GTX 1080 Ti achieves a slightly better performance than the Titan X despite the smaller video memory. Nvidia set the list price for the GeForce GTX 1080 Ti at 699 US dollars and waived a price premium for the Founders Edition. This makes it the cheapest graphics card with the GP102 graphics processor and at the same time the only one for which Nvidia has allowed board partners to design their own.

GeForce GTX 1070, 1070 Ti and 1080

Nvidia presented the first graphics card of the GeForce 10 series, the GeForce GTX 1080, on May 6, 2016 and announced the start of sales for May 27. The GeForce GTX 1070 followed on June 10, 2016. Both cards use the GP104 graphics processor. A special feature of the cards was that Nvidia offered them in a so-called Founders Edition . However, it was only a matter of the reference design, which Nvidia now offered under a different name with an extended support period and increased list price and sold through its own online shop. The last circumstance in particular - the higher list price (379 to 449 US dollars for the GeForce GTX 1070) compared to designs by the board partners - caused considerable confusion and criticism in the specialist press. The criticism of the Founders Edition was also caused by the increased noise development - compared to the designs of the board partners: On the GeForce GTX 1080 this reached up to 4.2 sone under full load, despite the vapor chamber cooling.

The GP104-GPU represents the graphics processor of the Pascal generation for the high-end sector. It is 314 mm² in size, consists of 7.2 billion transistors and is manufactured at TSMC using the 16 nm FinFET manufacturing process. It has 2560 shader and 160 texture units organized in 20 clusters. Furthermore, the GP104 has 4 raster engines, 64 raster units on 8 partitions and a 2048 kB L2 cache. Compared to its direct predecessor, the GM204, the GP104 now also supports HDMI 2.0b (including HDCP 2.2 ) and DisplayPort 1.4 (although this was not yet certified at the time of release).

The GeForce GTX 1080 was released by Nvidia on May 27, 2016, with the GP104-400-A1 being used in full and having an 8 GB GDDR5X video memory. Under Ultra HD, the card has a 64% to 73 % higher performance than its predecessor, the GeForce GTX 980 , the GeForce GTX 980 Ti a 30% to 37% and the AMD Radeon R9 Fury X a 26% to 33% higher performance . The GeForce GTX 1080 established itself briefly as the fastest single GPU card on the market until Nvidia presented the Titan X. The high performance in combination with the further improved energy efficiency led to correspondingly positive reviews in the trade press. On the other hand, the list price for the Founders Edition was again increased to 599 and 699 US dollars (the list price for the GeForce GTX 980 was still 549 US dollars), favored by the lack of competitive offers from AMD. It wasn't until the GeForce GTX 1080 Ti was launched that Nvidia officially lowered the list price to $ 499. On February 28, 2017, Nvidia introduced the GP104-410-A1 GPU on the GeForce GTX 1080, which supports the GDDR5X video memory up to 11 Gbps.

The GeForce GTX 1070 followed the GTX 1080 on June 10, 2016. The GP104-200-A1 graphics processor used is, however, partially severely cut: 5 of the 20 shader clusters are deactivated, i.e. H. a complete GPC unit is not available. In contrast to its direct predecessor, the GeForce GTX 970 , Nvidia did not cut the video memory, but instead of the GDDR5X, the conventional GDDR5 memory is used. In this configuration, the GeForce GTX 1070 achieved 82 to 86% of the performance of the GTX 1080. Compared to the top models of the previous generation, the GeForce GTX 980 Ti and the Radeon R9 Fury X , it achieved a performance gain of 3 to 9% under Ultra HD. As with the GeForce GTX 1080, Nvidia also increased the list price for the GTX 1070 by 50 US dollars compared to its direct predecessor (from 329 to 379 US dollars), for which no counterpart was issued by AMD.

The GeForce GTX 1070 Ti , released on October 26, 2017, tries the balancing act of not cannibalizing the more powerful GTX 1080 sister model and at the same time facing AMD's Radeon RX Vega 56 . Out of consideration for the GTX 1080, Nvidia had imposed restrictions on the production partners with regard to factory overclocking or increasing the power limit. As a result, all Ti cards sold have a base clock of 1,607 MHz in the BIOS, but the models from Nvidia partners can exceed the performance of Nvidia's reference card using tuning tools and a more efficient cooling design.

GeForce GTX 1060

The GP106-GPU represents the graphics processor of the Pascal generation for the middle class sector. It is 200 mm² in size, consists of 4.4 billion transistors and is also manufactured at TSMC in the 16 nm FinFET production process. From a technical point of view, the GP106 is a straight halving of the GP104: It has 1280 shader and 80 texture units that are organized in 10 clusters. Furthermore, the GP106 has 2 raster engines, 48 ​​raster units on 6 partitions, and a 2048 kB L2 cache. In contrast to the graphics processors of earlier generations from the performance sector, the GP106 no longer supports SLI .

On July 19, 2016, Nvidia presented the GeForce GTX 1060 , which uses the GP106-400-A1 in its full configuration and has over 6 GB of VRAM. In contrast to the other previously presented graphics cards in the series, the Radeon RX 480 was a direct competitor from AMD . Compared to this, the GeForce GTX 1060 was able to show around 6% higher performance under Full HD, compared to its direct predecessor, the GeForce GTX 960 (4 GB), around 75% higher performance. Since the difference in performance roughly corresponded to the difference in price (€ 269 to € 279), both cards were rated positively in the trade press for their price-performance ratio. A previously assumed disadvantage of the GeForce GTX 1060 compared to the Radeon RX 480 was the smaller video memory (6 to 8 GB). In the practical tests - due to the better memory management and the improved color compression - no disadvantage could be found. In contrast, the Radeon RX 480 could show a higher performance under DirectX 12 and Vulkan , which means that it is potentially more future-proof. Nvidia came under fire when it launched a 3 GB version of the GeForce GTX 1060 in early August. In some cases, a clear memory limitation could be determined. The fact that the GP106-300-A1 is not used in its full configuration here, but a shader cluster is deactivated, which is not immediately recognizable to the customer, also had a negative effect on the performance; Both models can only be distinguished by the memory size. Analogous to the GeForce GTX 1080, Nvidia introduced the GP106-410-A1 GPU on the GeForce GTX 1060 on February 28, 2017, with which the GDDR5 video memory is supported up to 9 Gbps. At the end of December 2017, Nvidia only presented a 5 GB version of the GeForce GTX 1060 for the Asian market. This uses the GP106-350-K3-A1 in full configuration, but with a memory interface reduced to 160 bits.

GeForce GTX 1050 and 1050 Ti

On October 25, 2016, Nvidia presented the GeForce GTX 1050 and GeForce GTX 1050 Ti graphics cards , which use the GP107 graphics processor and are equipped with the previously released Pascal graphics cards of the Nvidia GeForce 10 series for performance and Clearly sell the high-end market. Nvidia is targeting the price segments below 200 euros, so the GeForce GTX 1050 will come onto the market for 125 euros. A reference model like the Founders Edition of the GeForce GTX 1080 and GeForce GTX 1070 will not be available for the GeForce GTX 1050. With the GeForce GTX 1050 and the more powerful variant GTX 1050 Ti Nvidia positioned leistungstechnisch between AMD 's Radeon RX 470 and Radeon RX 460 . Compared to the top models of the GeForce 10 series, the GTX 1050 with the GP107 has to accept significant performance cuts in gaming applications: On the smaller die base area of ​​132 mm² (see GP104 / GTX 1080: 314 mm²) the GP107 has significantly less grid, Shader and texture units. The memory size of the GTX 1050 has been reduced to 2 GB GDDR5, while the GTX 1050 Ti with 4 GB GDDR5 is even more generously equipped than the 3 GB version of the GeForce GTX 1060. Even when overclocked, the performance was only achieved with reduced detail levels in satisfactory results for the games. Performance tests were limited to the popular Full HD resolution with 1920 × 1080 pixels. The lower graphics performance results in lower power consumption and a lower noise level than Nvidia's top models.

GeForce GT 1030

The entry-level model GeForce GT 1030 was presented on May 17, 2017 . This uses the GP108 graphics processor, which is manufactured by Samsung using the 14 nm LPP process. There is no reference design of the GT 1030, only custom models from various manufacturers are sold. Nvidia positions this graphics card in the price segment under 80 euros. With only 384 shaders and a 2 GB GDDR5 memory, the graphics performance is severely limited. Nvidia's partners also sell GT-1030 cards with DDR4 memory, which are significantly slower, but also more energy-efficient. The GT 1030 supports neither SLI nor G-Sync, nor does it have special VR technologies such as simultaneous multi-projection . Due to the low TDP of only 30 W, no additional power plug is required from the power supply unit. Most versions can thus be built compactly and passively cooled and are then almost inaudible. The GT 1030 is particularly suitable for office and multimedia applications and graphically less demanding games such as Dota 2 or Overwatch. In games with high graphics requirements such as Battlefield 1, the details must be greatly reduced under Full HD (1920 × 1080 pixels) in order to achieve a smooth refresh rate .

Data overview

Graphics processors

Graphics
chip
production units L2
cache
API support Video
pro-
cessor
Bus
interface
stelle
production
process
transis-
interfere
The -
area
ROP
particle
functions
ROPs Unified shaders Texture units DirectX OpenGL OpenCL CUDA -
Level
volcano
Stream
processors
Shader -
cluster
TAUs TMUs
GP100 16 nm 15.3 billion 610 mm² 06th 96 3840 60 240 240 4096 KB 12.1 4.5 1.2 6.1 1.0 k. A. PCIe 3.0
GP102 12.0 billion 471 mm² 12 96 3840 30th 240 240 3072 KB VP8
GP104 07.2 billion 314 mm² 08th 64 2560 20th 160 160 2048 KB
GP106 04.4 billion 200 mm² 06th 48 1280 10 080 080 1536 KB
GP107 14 nm 03.3 billion 132 mm² 04th 32 0768 06th 048 048 1024 KB
GP108 01.8 billion 070 mm² 02 16 0384 03 024 024 0512 KB

Desktop model data

model Official
launch
Graphics processor (GPU) Graphics memory Performance data Power consumption
Type Active units Chip clock
(MHz)
Memory
size
Memory
clock
(MHz)
Storage
interface
Computing power
(in GFlops )
Fill rate Memory
bandwidth
MGCP
Readings
ROPs Shader -
cluster
ALUs Texture
units
Base Boost SP
(MAD)
DP
(FMA)
Pixels
(GP / s)
Texel
(GT / s)
Idle 3D
load
GeForce GT 1030 March 12, 2018 GP108-310-A1 16 03 0384 024 1151 1379 02 GB DDR4GX 1050 064 bit 01059 033 022.1 033.1 016.8 GB / s 020 W k. A. k. A.
17th May 2017 GP108-300-A1 1227 1468 02 GB GDDR5X 1500 (3000) 01127 035 023.5 035.2 048 GB / s 030 W k. A. k. A.
GeForce GTX 1050 (2 GB) Oct 25, 2016 GP107-300-A1 32 05 0640 040 1354 1455 02 GB GDDR5X 1750 (3500) 128 bit 01862 058 046.6 058.2 112 GB / s 075 W 05 W 055 W
GeForce GTX 1050 (3 GB) May 21, 2018 GP107-301-K1-A1 24 06th 0768 048 1392 1518 03 GB GDDR5X 1750 (3500) 096 bits 02332 073 036.4 072.9 084 GB / s 075 W k. A. k. A.
GeForce GTX 1050 Ti Oct 25, 2016 GP107-400-A1 32 06th 0768 048 1290 1382 04 GB GDDR5X 1750 (3500) 128 bit 02123 066 044.2 066.3 112 GB / s 075 W 05 W 058 W
GeForce GTX 1060 (3 GB) 18 Aug 2016 GP106-300-A1 48 09 1152 072 1506 1708 03 GB GDDR5X 2000 (4000) 192 bits 03935 123 082.0 123.0 192 GB / s 120 W 07 W 115 W
GeForce GTX 1060 (5 GB) Dec 26, 2017 GP106-350-K3-A1 40 10 1280 080 1506 1708 05 GB GDDR5X 2000 (4000) 160 bits 04372 137 068.3 136.6 160 GB / s 120 W k. A. k. A.
GeForce GTX 1060 (6 GB) 19 Jul 2016 GP106-400-A1 48 10 1280 080 1506 1708 06 GB GDDR5X 2000 (4000) 192 bits 04372 137 082.0 136.6 192 GB / s 120 W 06 W 118 W
28 Feb 2017 GP106-410-A1 2250 (4500) 216 GB / s k. A. k. A.
GeForce GTX 1070 Jun 10, 2016 GP104-200-A1 64 15th 1920 120 1506 1683 08 GB GDDR5X 2000 (4000) 256 bit 06463 202 107.7 202.0 256 GB / s 150 W 07 W 147 W
GeForce GTX 1070 Ti Oct. 26, 2017 GP104-300-A1 64 19th 2432 152 1607 1683 08 GB GDDR5X 2000 (4000) 256 bit 08186 256 107.7 255.8 256 GB / s 180 W 09 W. 176 W
GeForce GTX 1080 May 27, 2016 GP104-400-A1 64 20th 2560 160 1607 1733 08 GB GDDR5X 2500 (5000) 256 bit 08873 277 110.9 277.3 320 GB / s 180 W 07 W 173 W
28 Feb 2017 GP104-410-A1 2750 (5500) 352 GB / s k. A. k. A.
GeForce GTX 1080 Ti 0March 9, 2017 GP102-350-K1-A1 88 28 3584 224 1480 1582 11 GB GDDR5X 2750 (5500) 352 bits 11340 354 139.2 354.4 484 GB / s 250 W 11 W. 230 W
Titan X 0Aug 2, 2016 GP102-400-A1 96 28 3584 224 1417 1531 12 GB GDDR5X 2500 (5000) 384 bits 10974 343 147.0 342.9 480 GB / s 250 W 11 W. 220 W
Titan Xp 0Apr 6, 2017 GP102-450-A1 96 30th 3840 240 1404 1582 12 GB GDDR5X 2850 (5700) 384 bits 12150 380 151.9 379.7 547 GB / s 250 W 12 W. 245 W

Notebook model data

model Official
launch
Graphics processor (GPU) Graphics memory Performance data MGCP
Type Active units Chip clock
(MHz)
Memory
size
Memory
clock
(MHz)
Storage
interface
Computing power
(in GFlops )
Fill rate Memory
bandwidth
ROPs Shader -
cluster
ALUs Texture
units
Base Boost SP
(MAD)
DP
(FMA)
Pixels
(GP / s)
Texel
(GT / s)
GeForce GTX 1050 0Jan. 3, 2017 GP107: N17P-G0-A1 32 05 0640 040 1354 1493 4 GB GDDR5X 1750 (3500) 128 bit 1911 054 047.7 59.7 112 GB / s 75 W
0Feb. 1, 2017 GP106
0Jan. 3, 2018 GP107: N17P-G0-A1 0999-1189 1139-1328 34-40 W
GeForce GTX 1050 Ti 0Jan. 3, 2017 GP107: N17P-G1-A1 32 06th 0768 048 1493 1620 4 GB GDDR5X 1750 (3500) 128 bit 2488 078 051.8 077.7 112 GB / s 75 W
0Feb. 1, 2017 GP106
0Jan. 3, 2018 GP107: N17P-G1-A1 1151-1290 1290-1417 40-46 W.
GeForce GTX 1060 15 Aug 2016 GP106: N17E-G1-A1 48 10 1280 080 1404 1670 6 GB GDDR5X 2000 (4000) 192 bits 4275 134 080.2 133.6 192 GB / s 80 W
27 Jun 2017 1063-1265 1341-1480 60-70 W
GeForce GTX 1070 15 Aug 2016 GP104: N17E-G2-A1 64 16 2048 128 1442 1645 8 GB GDDR5X 2000 (4000) 256 bit 6738 211 105.3 210.6 256 GB / s 110 W
27 Jun 2017 1101-1215 1265-1379 80-90 W
GeForce GTX 1080 15 Aug 2016 GP104: N17E-G3-A1 64 20th 2560 160 1556 1733 8 GB GDDR5X 2500 (5000) 256 bit 8873 277 110.9 277.3 320 GB / s 150 W
27 Jun 2017 1101-1290 1278-1468 90-110 W.

Remarks

  1. a b The date indicated is the date of the public presentation, not the date of availability of the models.
  2. a b The specified performance values ​​for the computing power via the stream processors, the pixel and texel fill rate, and the memory bandwidth are theoretical maximum values ​​(with boost clock) that are not directly comparable with the performance values ​​of other architectures. The overall performance of a graphics card depends, among other things, on how well the available resources can be used or fully utilized. There are also other factors that are not listed here that affect performance.
  3. a b c d The specified clock rates are the reference data recommended or specified by Nvidia; the I / O clock is specified for the memory clock (the effective clock in brackets). However, the exact clock rate can deviate by a few megahertz due to different clock generators, and the final definition of the clock rates is in the hands of the respective graphics card manufacturer. It is therefore entirely possible that there are or will be graphics card models with different clock rates.
  4. a b The MGCP value given by Nvidia does not necessarily correspond to the maximum power consumption. This value is also not necessarily comparable with the TDP value of the competitor AMD.
  5. The measured values ​​listed in the table relate to the pure power consumption of graphics cards that correspond to the Nvidia reference design. A special measuring device is required to measure these values; Depending on the measurement technology used and the given measurement conditions, including the program used to generate the 3D load, the values ​​can fluctuate between different devices. Therefore, measured value ranges are given here, each representing the lowest, typical and highest measured values ​​from different sources.

Web links

Commons : Nvidia GeForce 10 series  - collection of images, videos and audio files

Individual evidence

  1. a b Nvidia Titan X in the test: The GP102, which is not allowed to be a (extremely fast) GeForce. PC Games Hardware, September 11, 2016, accessed September 15, 2016 .
  2. Nvidia Titan Xp: GP102 fully equipped with 12 GB GDDR5X for 1,349 euros. ComputerBase, April 6, 2017, accessed May 16, 2017 .
  3. GTX 1080 Ti: Benchmarks, also with overclocking. PC Games Hardware, March 10, 2017, accessed May 16, 2017 .
  4. Launch analysis nVidia GeForce GTX 1080 Ti (page 3). 3DCenter.org, March 23, 2017, accessed May 16, 2017 .
  5. Nvidia GeForce GTX 1080 and 1070: Founders Editions as "Early Access". PC Games Hardware, May 11, 2016, accessed September 7, 2016 .
  6. GeForce GTX 1080 in the test: First 16 nm king with 2 GHz in OC operation (paragraph: Nvidia GeForce GTX 1080 in the test: loudness, power consumption, cooling). PC Games Hardware, October 7, 2016, accessed October 20, 2016 .
  7. GeForce GTX 1080 in the test: A bang in 16 nm with 1.8 GHz and GDDR5X (page 4). Computerbase, May 17, 2016, accessed September 17, 2016 .
  8. GeForce GTX 1080 in the test: A bang in 16 nm with 1.8 GHz and GDDR5X (page 7). Computerbase, May 17, 2016, accessed September 7, 2016 .
  9. a b c d GeForce GTX 1080 Supercharged With Next-Gen 11 Gbps GDDR5X Memory. Nvidia Corporation, February 28, 2017, accessed May 16, 2017 .
  10. GeForce GTX 1070 in the test: Efficient, high performance at a high price (page 4). Computerbase, June 9, 2016, accessed September 7, 2016 .
  11. GeForce GTX 1070 Ti in the test: Courage for the gap - with success? [Update]. PC Games Hardware, November 3, 2017, accessed November 9, 2017 .
  12. GeForce GTX 1060 in the test: Three GP106 in a duel with the Radeon RX 480 (page 1). Computerbase, July 19, 2016, accessed September 15, 2016 .
  13. GeForce GTX 1060 in the test: Three GP106 in a duel with the Radeon RX 480 (page 3). Computerbase, July 19, 2016, accessed September 15, 2016 .
  14. GeForce GTX 1060 in the test: Three GP106 in a duel with the Radeon RX 480 (page 7). Computerbase, July 19, 2016, accessed September 15, 2016 .
  15. AMD overtakes nVidia under DirectX 12 & Vulkan. 3dCenter.org, July 24, 2016, accessed September 15, 2016 .
  16. Asus GeForce GTX 1060 3 GB in the test: 3,072 MB VRAM is not always enough (page 4). Computerbase, September 12, 2016, accessed September 15, 2016 .
  17. a b NVIDIA GeForce GTX 1060 5GB. VideoCardz, December 16, 2017, accessed October 30, 2018 .
  18. GeForce GTX 1050 (Ti) in the test: Pascal's economy model from 125 euros. PC Games Hardware, October 25, 2016, accessed October 25, 2016 .
  19. GeForce GT 1030 Graphics Card | GeForce | NVIDIA. Retrieved March 11, 2019 .
  20. Entry-level graphics cards: RX 560 and GT 1030 available. PC Games Hardware, May 17, 2017, accessed November 29, 2017 .
  21. Graphics card: watch out for the Geforce GT 1030 with DDR4 memory - Golem.de. Retrieved on August 20, 2019 (German).
  22. Wolfgang Andermahr: RX 550 and GT 1030 in the test: GPUs for Dota 2 or Overwatch and UHD-YouTube. Computerbase, May 23, 2017, accessed November 29, 2017 .
  23. NVIDIA GeForce GT 1030. Nvidia Corporation, accessed April 9, 2018 .
  24. a b c NVIDIA GeForce GTX 1050. Nvidia Corporation, accessed January 10, 2017 .
  25. a b c d e f g h i j k l m n o p q r s t Power consumption of current and past graphics cards. 3DCenter.org, February 23, 2014, accessed May 13, 2017 .
  26. a b NVIDIA GeForce GTX 1060. Nvidia Corporation, July 7, 2016, accessed July 8, 2016 .
  27. NVIDIA GeForce GTX 1070. Nvidia Corporation, May 18, 2016, accessed May 19, 2016 .
  28. NVIDIA GeForce GTX 1070 Ti. (No longer available online.) Nvidia Corporation, October 26, 2017, formerly in the original ; accessed on October 26, 2017 (English).  ( Page no longer available , search in web archivesInfo: The link was automatically marked as defective. Please check the link according to the instructions and then remove this notice.@1@ 2Template: Dead Link / www.geforce.com  
  29. NVIDIA GeForce GTX 1080. Nvidia Corporation, May 6, 2016, accessed May 7, 2016 .
  30. NVIDIA GeForce GTX 1080 Ti. Nvidia Corporation, accessed March 1, 2017 .
  31. NVIDIA TITAN X. Nvidia Corporation, July 21, 2016, accessed July 22, 2016 .
  32. NVIDIA TITAN Xp. Nvidia Corporation, April 6, 2017, accessed April 6, 2017 .
  33. a b c d e GeForce GTX 10-Series Notebooks. Nvidia Corporation, August 16, 2016, accessed January 10, 2017 .