Rtx 3000 wiki

Series of video cards by AMD

Not to be confused with the Radeon HD 6000 series, an earlier series of GPUs developed by AMD.

Release date	Radeon RX 6600 and 6600 XT August 11, 2021; 2 months ago (2021-08-11) Radeon RX 6700 XT March 18, 2021; 6 months ago (2021-03-18) Radeon RX 6800 and Radeon RX 6800 XT November 18, 2020; 10 months ago (2020-11-18) Radeon RX 6900 XT December 8, 2020; 10 months ago (2020-12-08) Radeon RX 6600M, 6700M, and 6800M June 1, 2021; 4 months ago (2021-06-01)
Codename	Navi 2X (nicknamed "Big Navi")
Architecture	RDNA 2
Transistors	11.06B (Navi 23)[1] 17.2B (Navi 22) 26.8B (Navi 21)
Fabrication process	TSMC7 nm
Mid-range	Radeon RX 6600M Radeon RX 6600 Radeon RX 6600 XT
High-end	Radeon RX 6700M Radeon RX 6700 XT Radeon RX 6800M Radeon RX 6800 Radeon RX 6800 XT
Enthusiast	Radeon RX 6900 XT
Predecessor	Radeon RX 5000 series

The Radeon RX 6000 series is a series of graphics processing units developed by AMD, based on their RDNA 2 architecture.^[2] It was announced on October 28, 2020^[3] and is the successor to the Radeon RX 5000 series. The lineup consists of the RX 6600, RX 6600 XT, RX 6700 XT, RX 6800, RX 6800 XT and RX 6900 XT for desktop computers, and the RX 6600M, RX 6700M, and RX 6800M for laptops.^[4][5]

The lineup is designed to compete with Nvidia's GeForce 30 series of cards. These GPUs are also the first generation of AMD GPUs that support hardware accelerated real-time ray tracing.^[6]

History[edit]

On September 14, 2020, AMD first teased the physical design of its RX 6000 series through a tweet shared on social messaging service Twitter. At the same time, the company launched a virtual island inside the video game Fortnite that contained a large-scale rendition of the RX 6000 hardware design, which players could freely explore using the game's "Creative" mode.^[7]

AMD officially unveiled the first three cards of the lineup, the RX 6800, RX 6800 XT, and RX 6900 XT, in an event titled "Where Gaming Begins: Ep. 2" on October 28.^[8] In the event, AMD announced the RX 6800 XT as its flagship graphics processor, comparing its performance to that of Nvidia's RTX 3080 graphics card in 1440p and 4K resolution gaming. The 6800 XT was announced with a price tag of $649 USD, which is $50 lower than the RTX 3080's starting price of $699.^[8] AMD then introduced the RX 6800 as a competitor to Nvidia's previous-generation RTX 2080 Ti, but with a significantly lower price tag of $579, compared to $999 for the 2080 Ti.^[8] Lastly, AMD introduced the RX 6900 XT as its top card of the lineup, claiming it produces similar performance to Nvidia's RTX 3090 but with lower power consumption, and announced a launch price of $999, $500 cheaper than the RTX 3090.^[8]

The Radeon RX 6800 and 6800 XT were released on November 18, 2020, and the RX 6900 XT was released on December 8, 2020.^[9][10]

On February 3, 2021, Gigabyte registered a range of RX 6700 XT graphics cards with the Eurasian Economic Commission (EEC), with the filing indicating that all seven registered models will ship with 12 GB of memory.^[11] On March 3, 2021, AMD officially announced the RX 6700 XT graphics card, set to compete with Nvidia's RTX 3060 Ti and 3070 cards.^[12] The card launched on March 18, 2021.^[4]

On May 31, 2021, AMD announced the RX 6000M series of GPUs designed for laptops.^[13][14] These include the RX 6600M, RX 6700M, and RX 6800M. These were made available beginning on June 1.^[13]

On June 23, Gigabyte registered six RX 6600 XT graphics cards with the EEC, indicating that they would all have 8 GB of memory.^[15][16][17] This was just over a month after ASRock had made similar filings for both the RX 6600 and 6600 XT, which listed 8 GB of memory for both models.^[18][19][20] On July 5, VideoCardz discovered that Taiwanese graphics card retailer PowerColor had already created product pages for the unannounced Radeon 6600 and 6600 XT GPUs.^[21][22]

On July 30, AMD announced the RX 6600 and 6600 XT GPUs, which were later released on August 11, 2021. The RX 6600 XT is available for $379 USD MSRP.^[23]

Availability issues[edit]

See also: 2020–21 global chip shortage

Much like Nvidia's competing GeForce 30 series, releases sold out almost immediately, due to a combination of low stock and scalping bots.^[24]

RX 6800 and 6800 XT[edit]

The RX 6800 and 6800 XT were launched on November 18, 2020, but due to low stock availability, they sold out at most retailers on the same day.^[9][25][26] American retailer Micro Center restricted all sales to in-store only, claiming that stock "will be extremely limited at launch".^[27]PCMag reported that stock for both cards on Newegg had sold out as early as 6:05 a.m. PST, and that stock was unavailable on AMD's own store website by 6:11 a.m.^[25] Retailer B&H Photo Video outright refused to accept orders for the graphics cards, stating in part, "we do not know and are therefore unable to provide a date or time for when these items will become available to purchase."^[25]

Scalpers were reportedly re-selling the GPUs on eBay for around US$1000 - $1500, roughly double the initial launch price.^[9] Some frustrated social media users attempted to derail the listings by using bots to make fake bids on the scalped GPUs at absurd prices; in one instance, an auction for an RX 6800 XT saw bids go as high as $70,000.^[28]

RX 6900 XT[edit]

The RX 6900 XT was launched on December 8, 2020, and similarly to the RX 6800 and 6800 XT, it sold out on the same day it was released.^[29][10] According to PCMag, it was listed as out of stock on Newegg by 6:02 a.m. PST, just two minutes after it went on sale. AMD's store website, struggling to keep up with the high number of visitors, kept displaying 503 Service Temporarily Unavailable errors to customers attempting to buy the graphics card; by 6:35 a.m., it was sold out.^[10]

Products[edit]

Desktop[edit]

Mobile[edit]

See also[edit]

References[edit]

External links[edit]

Short description
Brand of Nvidia graphics cards intended for use in workstations running professional applications

Quadro is Nvidia's brand for graphics cards intended for use in workstations running professional computer-aided design (CAD), computer-generated imagery (CGI), digital content creation (DCC) applications, scientific calculations and machine learning.

The GPU chips on Quadro-branded graphics cards are identical to those used on GeForce-branded graphics cards. Differences between the Quadro and GeForce cards include the use of ECC memory and enhanced floating point precision. These are desirable properties when the cards are used for calculations which, in contrast to graphics rendering, require reliability and precision.

The Nvidia Quadro product line directly competes with AMD's Radeon Pro line of professional workstation cards.^[1]

History

The Quadro line of GPU cards emerged in an effort towards market segmentation by Nvidia.(citation?) In introducing Quadro, Nvidia was able to charge a premium for essentially the same graphics hardware in professional markets, and direct resources to properly serve the needs of those markets. To differentiate their offerings, Nvidia used driver software and firmware to selectively enable features vital to segments of the workstation market, such as high-performance anti-aliased lines and two-sided lighting, in the Quadro product.(citation?) The Quadro line also received improved support through a certified driver program.(citation?) These features were of little value to the gamers that Nvidia's products already sold to, but their lack prevented high-end customers from using the less expensive products.

There are parallels between the market segmentation used to sell the Quadro line of products to workstation (DCC) markets and the Tesla line of products to engineering and HPC markets.

In a settlement of a patent infringement lawsuit between SGI and Nvidia, SGI acquired rights to speed-binned Nvidia graphics chips which they shipped under the VPro product label. These designs were completely separate from the SGI Odyssey based VPro products initially sold on their IRIX workstations which used a completely different bus. SGI's Nvidia-based VPro line included the VPro V3 (Geforce 256), VPro VR3 (Quadro), VPro V7 (Quadro2 MXR), and VPro VR7 (Quadro2 Pro).^[2][3]

Quadro SDI

Actual extra cards only for Quadro 4000 cards and higher:

Quadro Plex

Quadro Plex consists of a line of external servers for rendering videos. A Quadro Plex contains multiple Quadro FX video cards. A client computer connects to Quadro Plex (using PCI Express ×8 or ×16 interface card with interconnect cable) to initiate rendering. More data in Nvidia Tesla Cards.

Quadro SLI and SYNC

Scalable Link Interface, or SLI, is the next generation of Plex. SLI can improve Frame Rendering, FSAA.^[6][7]

Quadro SLI support Mosaic for 2 Cards and 8 Monitors.^[8]

With Quadro SYNC Card support of max. 16 Monitors (4 per Card) possible.^[9][10]

Most Cards have SLI-Bridge-Slot for 2, 3 or 4 cards on one main board.^[11]

Acceleration of scienctific calculations is possible with CUDA and OpenCL.^[12][13][14]

Nvidia has 4 types of SLI bridges:

More see SLI.

Quadro VCA

Nvidia supports SLI and supercomputing with its 8-GPU Visual Computing Appliance.^[19] Nvidia Iray,^[20][21] Chaosgroup V-Ray^[22] and Nvidia OptiX^[23] accelerate Raytracing for Maya, 3DS Max, Cinema4D, Rhinoceros and others. All software with CUDA or OpenCL, such as ANSYS, NASTRAN, ABAQUS, and OpenFoam, can benefit from VCA. The DGX-1 is available with 8 GP100 Cards.^[24]

More data in Nvidia Tesla Cards.

Quadro RTX

The Quadro RTX series is based on the Turing microarchitecture, and features real-time raytracing.^[25] This is accelerated by the use of new RT cores, which are designed to process quadtrees and spherical hierarchies, and speed up collision tests with individual triangles.

The raytracing performed by the RT cores can be used to produce reflections, refractions and shadows, replacing traditional raster techniques such as cube maps and depth maps. Instead of replacing rasterization entirely, however, the information gathered from ray-tracing can be used to augment the shading with information that is much more physically correct, especially regarding off-camera action.

Tensor cores further enhance the image produced by raytracing, and are used to de-noise a partially rendered image. The Tensor core performs the result of deep learning on supercomputers to codify how to, for example, perform better up scaling than a standard algorithm would, which results in higher perceived resolution. In the Tensor core's primary usage, a problem to be solved is analyzed on a supercomputer, which is taught by example what results are desired, and the supercomputer determines a method to use to achieve those results, which is then done with the consumer's Tensor core. These methods are delivered "over the air" to consumers.

RTX is also the name of the development platform introduced for the Quadro RTX series. RTX leverages Microsoft's DXR, OptiX and Vulkan for access to raytracing.^[26]

Turing is manufactured using TSMC's 12 nm FinFET semiconductor fabrication process.^[27] Quadro RTX also uses GDDR6 memory from Samsung Electronics.^[28]

Video cards

GeForce

Many of these cards use the same core as the game- and action-oriented GeForce video cards by Nvidia. Those cards that are identical to the desktop cards can be software modified to identify themselves as the equivalent Quadro cards and this allows optimized drivers intended for the Quadro cards to be installed on the system. While this may not offer all of the performance of the equivalent Quadro card,(citation?) it can improve performance in certain applications, but may require installing the MAXtreme driver for comparable speed.

The performance difference comes in the firmware controlling the card.(citation?) Given the importance of speed in a game, a system used for gaming can shut down textures, shading, or rendering after only approximating a final output—in order to keep the overall frame rate high. The algorithms on a CAD-oriented card tend rather to complete all rendering operations, even if that introduces delays or variations in the timing, prioritising accuracy and rendering quality over speed. A Geforce card focuses more on texture fillrates and high framerates with lighting and sound, but Quadro cards prioritize wireframe rendering and object interactions.

Software

With Caps Viewer (1.38 in 2018) all Windows Users can see data of the graphic Card, the installed Driver and can test some Features.^[29] GPU-Z reads also data of the graphic cards and the user can send some data for better database.^[30]

Quadro drivers

• Curie-Architecture Last drivers see Driver Portal of Nvidia^[31] (End-of-Life)
• Tesla-Architecture (G80+, GT2xx) in Legacy Mode Quadro Driver 340+: OpenGL 3.3, OpenCL 1.1, DirectX 10.0/10.1^[32] (End-of-Life)
• Fermi (GFxxx): OpenCL 1.1, OpenGL 4.5, some OpenGL 2016 Features with Quadro Driver 375,^[33] in legacy mode with version 391.74 (End-of-Life)
• Kepler (GKxxx): OpenCL 1.2, OpenGL 4.6, Vulkan 1.2 with Quadro Driver 450+^[34][35]
• Maxwell (GMxxx): OpenCL 1.2, OpenGL 4.6, Vulkan 1.2 with Quadro Driver 450+^[34]
• Pascal (GPxxx): OpenCL 1.2, OpenGL 4.6, Vulkan 1.2 with Quadro driver 450+^[34]
• Volta (GVxxx): OpenCL 1.2, OpenGL 4.6, Vulkan 1.2 with Quadro driver 450+^[34]
• Turing (TUxxx): OpenCL 1.2, OpenGL 4.6, Vulkan 1.2 with Quadro driver 450+^[34]

CUDA

• Tesla Architecture and later

Supported CUDA Level of GPU and Card.^[36]

• CUDA SDK 6.5 support for Compute Capability 1.0 - 5.x (Tesla, Fermi, Kepler, Maxwell) Last Version with support for Tesla-Architecture with Compute Capability 1.x
• CUDA SDK 7.5 support for Compute Capability 2.0 - 5.x (Fermi, Kepler, Maxwell)
• CUDA SDK 8.0 support for Compute Capability 2.0 - 6.x (Fermi, Kepler, Maxwell, Pascal) Last version with support for compute capability 2.x (Fermi)
• CUDA SDK 9.0/9.1/9.2 support for Compute Capability 3.0 - 7.2 (Kepler, Maxwell, Pascal, Volta)
• CUDA SDK 10.0/10.1/10.2 support for Compute Capability 3.0 - 7.5 (Kepler, Maxwell, Pascal, Volta, Turing) Last version with support for compute capability 3.x (Kepler).
• CUDA SDK 11.0 support for Compute Capability 5.2 - 7.5 (Maxwell, Pascal, Volta, Turing, Ampere)

For own Card Test see CUDA-Z Tool^[37]

Desktop PCI Express

Quadro FX (without CUDA, OpenCL, or Vulkan)

• Architecture Rankine, Curie
• Rankine (NV3x): DirectX 9.0a, Shader Model 2.0a, OpenGL 2.1
• Curie (NV4x, G7x): DirectX 9.0c, Shader Model 3.0, OpenGL 2.1

Quadro FX (with CUDA and OpenCL, but no Vulkan)

• Architecture Tesla (G80+, GT2xx) with OpenGL 3.3 and OpenCL 1.1
• Tesla (G80+): DirectX 10, Shader Model 4.0, only Single Precision (FP32) available for CUDA and OpenCL
• Tesla 2 (GT2xx): DirectX 10.1, Shader Model 4.1, Single Precision (FP32) available for CUDA and OpenCL (Double Precision (FP64) available for CUDA and OpenCL only for GT200 with Cuda Level 1.3 and higher)

Quadro

• Architecture Fermi (GFxxx), Kepler (GKxxx), Maxwell (GMxxx), Pascal (GPxxx), Volta (GVxxx) (except Quadro 400 with Tesla 2)
• All Cards with Display Port 1.1+ can support 10bit per Channel for OpenGL (HDR for Graphics Professional (Adobe Photoshop and more))
• Vulkan 1.1 available with Driver Windows 419.17, Linux 418.43 for Kepler, Maxwell, Pascal, Volta^[84]
• All Kepler, Maxwell, Pascal, Volta and later can do OpenGL 4.6 with Driver 418+^[85]
• All Quadro can do OpenCL 1.1. Kepler and later can do OpenCL 1.2 and incomplete OpenCL 2.0 with Driver 378.66 and higher.
• All can do Double Precision with Compute Capability 1.3 and higher (see CUDA)

¹ Nvidia Quadro 342.01 WHQL: support of OpenGL 3.3 and OpenCL 1.1 for legacy Tesla microarchitecture Quadros.^[32]

² Nvidia Quadro 377.83 WHQL: support of OpenGL 4.5, OpenCL 1.1 for legacy Fermi microarchitecture Quadros.^[33]

³ Nvidia Quadro 451.77 WHQL: support of OpenGL 4.6, OpenCL 1.2 for Kepler, Maxwell, Pascal & Volta.^[34]

⁴ OpenCL 1.1 is available for Tesla-Chips,^[141] OpenCL 1.0 for some Cards with G8x, G9x and GT200 by MAC OS X^[142]

Quadro RTX

^{[150][151][152][153]}

Desktop AGP

• Architecture Celsius (NV1x): DirectX 7, OpenGL 1.2 (1.3)
• Architecture Kelvin (NV2x): DirectX 8 (8.1), OpenGL 1.3 (1.5), Pixel Shader 1.1 (1.3)
• Architecture Rankine (NV3x): DirectX 9.0a, OpenGL 1.5 (2.1), Shader Model 2.0a
• Architecture Curie (NV4x): DirectX 9.0c, OpenGL 2.1, Shader Model 3.0

Desktop PCI

• Architecture Rankine (NV3x): DirectX 9.0a, OpenGL 1.5 (2.1), Shader Model 2.0a

For business NVS

The Nvidia Quadro NVS graphics processing units (GPUs) provide business graphics solutions for manufacturers of small, medium, and enterprise-level business workstations. The Nvidia Quadro NVS desktop solutions enable multi-display graphics for businesses such as financial traders.

• Architecture Celsius (NV1x): DirectX 7, OpenGL 1.2 (1.3)
• Architecture Kelvin (NV2x): DirectX 8 (8.1), OpenGL 1.3 (1.5), Pixel Shader 1.1 (1.3)
• Architecture Rankine (NV3x): DirectX 9.0a, OpenGL 1.5 (2.1), Shader Model 2.0a
• Architecture Curie (NV4x): DirectX 9.0c, OpenGL 2.1, Shader Model 3.0
• Architecture Tesla (G80+): DirectX 10.0, OpenGL 3.3, Shader Model 4.0, CUDA 1.0 or 1.1, OpenCL 1.1
• Architecture Tesla 2 (GT2xx): DirectX 10.1, OpenGL 3.3, Shader Model 4.1, CUDA 1.2 or 1.3, OpenCL 1.1
• Architecture Fermi (GFxxx): DirectX 11.0, OpenGL 4.6, Shader Model 5.0, CUDA 2.x, OpenCL 1.1
• Architecture Kepler (GKxxx): DirectX 11.2, OpenGL 4.6, Shader Model 5.0, CUDA 3.x, OpenCL 1.2, Vulkan 1.1
• Architecture Maxwell 1 (GM1xx): DirectX 12.0, OpenGL 4.6, Shader Model 5.0, CUDA 5.0, OpenCL 1.2, Vulkan 1.1

Mobile applications

Quadro FX M (without Vulkan)

• Architecture Rankine (NV3x), Curie (NV4x, G7x) and Tesla (G80+, GT2xx)

Quadro M

• Architecture Fermi, Kepler,^[220] Maxwell,^[221] Pascal
• Fermi, Kepler, Maxwell, Pascal can do OpenGL 4.6 with actual Driver 381+ for Linux or 390+ for windows^[85]
• All can do Double Precision with compute Capability 1.3 and higher
• Vulkan 1.0 and 1.1 with Kepler and later

Quadro RTX Mobile

Quadro NVS M

• Architecture Curie (NV4x, G7x): DirectX 9.0c, OpenGL 2.1, Shader Model 3.0
• Architecture Tesla (G80+): DirectX 10.0, OpenGL 3.3, Shader Model 4.0, CUDA 1.0 or 1.1, OpenCL 1.1
• Architecture Tesla 2 (GT2xx): DirectX 10.1, OpenGL 3.3, Shader Model 4.1, CUDA 1.2 or 1.3, OpenCL 1.1
• Architecture Fermi (GFxxx): DirectX 11.0, OpenGL 4.6, Shader Model 5.0, CUDA 2.x, OpenCL 1.1
• Architecture Kepler (GKxxx): DirectX 11.2, OpenGL 4.6, Shader Model 5.0, CUDA 3.x, OpenCL 1.2, Vulkan 1.1
• Architecture Maxwell 1 (GM1xx): DirectX 12.0, OpenGL 4.6, Shader Model 5.0, CUDA 5.0, OpenCL 1.2, Vulkan 1.1

NVENC Support Matrix

HW accelerated encode and decode are supported on NVIDIA Quadro products with Fermi, Kepler, Maxwell and Pascal generation GPUs.^[285]

See also

• AMD Radeon Pro – AMD's competing workstation graphics solution
• Comparison of Nvidia graphics processing units
• CUDA – Nvidia CUDA technology
• Nvidia Tesla – Nvidia's first dedicated general purpose GPU (graphics processing unit)
• Sun Visualization System – uses Nvidia Quadro FX for 3D rendering and graphics acceleration

References

For the screen reader known as "NVDA", see NonVisual Desktop Access.

American multinational technology company

Coordinates: 37°22′14.62″N121°57′49.46″W / 37.3707278°N 121.9637389°W / 37.3707278; -121.9637389

Nvidia Corporation^{[note 1]} (en-VID-ee-ə) is an American multinational technology company incorporated in Delaware and based in Santa Clara, California.^[2] It designs graphics processing units (GPUs) for the gaming and professional markets, as well as system on a chip units (SoCs) for the mobile computing and automotive market. Its primary GPU line, labeled "GeForce", is in direct competition with the GPUs of the "Radeon" brand by Advanced Micro Devices (AMD). Nvidia expanded its presence in the gaming industry with its handheld game consolesShield Portable, Shield Tablet, and Shield Android TV and its cloud gaming serviceGeForce Now. Its professional line of GPUs are used in workstations for applications in such fields as architecture, engineering and construction, media and entertainment, automotive, scientific research, and manufacturing design.^[3]

In addition to GPU manufacturing, Nvidia provides an application programming interface (API) called CUDA that allows the creation of massively parallel programs which utilize GPUs.^[4][5] They are deployed in supercomputing sites around the world.^[6][7] More recently, it has moved into the mobile computing market, where it produces Tegra mobile processors for smartphones and tablets as well as vehicle navigation and entertainment systems.^[8][9][10] In addition to AMD, its competitors include Intel and Qualcomm.^[11][12]

Nvidia announced plans on September 13, 2020 to acquire Arm Ltd. from SoftBank, pending regulatory approval, for a value of US$40 billion in stock and cash, which would be the largest semiconductor acquisition to date. SoftBank Group will acquire slightly less than a 10% stake in Nvidia, and Arm will maintain its headquarters in Cambridge.^{[13][14][15][16][17]}

History[edit]

Nvidia was founded on April 5, 1993,^[18][19][20] by Jensen Huang (CEO as of 2020^[update]), a Taiwanese American, previously director of CoreWare at LSI Logic and a microprocessor designer at Advanced Micro Devices (AMD), Chris Malachowsky, an electrical engineer who worked at Sun Microsystems, and Curtis Priem, previously a senior staff engineer and graphics chip designer at Sun Microsystems.^[21][22]

In 1993, the three co-founders believed that the proper direction for the next wave of computing was accelerated or graphics-based computing because it could solve problems that general-purpose computing could not. They also observed that video games were simultaneously one of the most computationally challenging problems and would have incredibly high sales volume. Video games became the company's flywheel to reach large markets and funding huge R&D to solve massive computational problems. With only $40,000 in the bank, the company was born.^[23] The company subsequently received $20 million of venture capital funding from Sequoia Capital and others.^[24] Nvidia initially had no name and the co-founders named all their files NV, as in "next version". The need to incorporate the company prompted the co-founders to review all words with those two letters, leading them to "invidia", the Latin word for "envy".^[23] Nvidia went public on January 22, 1999.^[25][26][27]

Releases and acquisitions[edit]

The release of the RIVA TNT in 1998 solidified Nvidia's reputation for developing capable graphics adapters. In late 1999, Nvidia released the GeForce 256 (NV10), most notably introducing on-board transformation and lighting (T&L) to consumer-level 3D hardware. Running at 120 MHz and featuring four-pixel pipelines, it implemented advanced video acceleration, motion compensation, and hardware sub-picture alpha blending. The GeForce outperformed existing products by a wide margin.

Due to the success of its products, Nvidia won the contract to develop the graphics hardware for Microsoft's Xbox game console, which earned Nvidia a $200 million advance. However, the project took many of its best engineers away from other projects. In the short term this did not matter, and the GeForce2 GTS shipped in the summer of 2000. In December 2000, Nvidia reached an agreement to acquire the intellectual assets of its one-time rival 3dfx, a pioneer in consumer 3D graphics technology leading the field from mid 1990s until 2000.^[28][29] The acquisition process was finalized in April 2002.^[30]

In July 2002, Nvidia acquired Exluna for an undisclosed sum. Exluna made software-rendering tools and the personnel were merged into the Cg project.^[31] In August 2003, Nvidia acquired MediaQ for approximately US$70 million.^[32] On April 22, 2004, Nvidia acquired iReady, also a provider of high performance TCP/IP and iSCSI offload solutions.^[33] In December 2004, it was announced that Nvidia would assist Sony with the design of the graphics processor (RSX) in the PlayStation 3 game console. On December 14, 2005, Nvidia acquired ULI Electronics, which at the time supplied third-party southbridge parts for chipsets to ATI, Nvidia's competitor.^[34] In March 2006, Nvidia acquired Hybrid Graphics.^[35] In December 2006, Nvidia, along with its main rival in the graphics industry AMD (which had acquired ATI), received subpoenas from the U.S. Department of Justice regarding possible antitrust violations in the graphics card industry.^[36]

Forbes named Nvidia its Company of the Year for 2007, citing the accomplishments it made during the said period as well as during the previous five years.^[37] On January 5, 2007, Nvidia announced that it had completed the acquisition of PortalPlayer, Inc.^[38] In February 2008, Nvidia acquired Ageia, developer of the PhysXphysics engine and physics processing unit. Nvidia announced that it planned to integrate the PhysX technology into its future GPU products.^[39][40]

In July 2008, Nvidia took a write-down of approximately $200 million on its first-quarter revenue, after reporting that certain mobile chipsets and GPUs produced by the company had "abnormal failure rates" due to manufacturing defects. Nvidia, however, did not reveal the affected products. In September 2008, Nvidia became the subject of a class action lawsuit over the defects, claiming that the faulty GPUs had been incorporated into certain laptop models manufactured by Apple Inc., Dell, and HP. In September 2010, Nvidia reached a settlement, in which it would reimburse owners of the affected laptops for repairs or, in some cases, replacement.^[41][42] On January 10, 2011, Nvidia signed a six-year, $1.5 billion cross-licensing agreement with Intel, ending all litigation between the two companies.^[43]

In November 2011, after initially unveiling it at Mobile World Congress, Nvidia released its Tegra 3ARMsystem-on-a-chip for mobile devices. Nvidia claimed that the chip featured the first-ever quad-core mobile CPU.^[44][45] In May 2011, it was announced that Nvidia had agreed to acquire Icera, a baseband chip making company in the UK, for $367 million.^[46] In January 2013, Nvidia unveiled the Tegra 4, as well as the Nvidia Shield, an Android-based handheld game console powered by the new system-on-chip.^[47] On July 29, 2013, Nvidia announced that they acquired PGI from STMicroelectronics.^[48]

Since 2014, Nvidia has diversified its business focusing on three markets: gaming, automotive electronics, and mobile devices.^[49]

On May 6, 2016, Nvidia unveiled the first GPUs of the GeForce 10 series, the GTX 1080 and 1070, based on the company's new Pascal microarchitecture. Nvidia claimed that both models outperformed its Maxwell-based Titan X model; the models incorporate GDDR5X and GDDR5 memory respectively, and use a 16 nm manufacturing process. The architecture also supports a new hardware feature known as simultaneous multi-projection (SMP), which is designed to improve the quality of multi-monitor and virtual reality rendering.^[50][51][52] Laptops that include these GPUs and are sufficiently thin – as of late 2017, under 0.8 inches (20 mm) – have been designated as meeting Nvidia's "Max-Q" design standard.^[53]

In July 2016, Nvidia agreed to a settlement for a false advertising lawsuit regarding its GTX 970 model, as the models were unable to use all of their advertised 4 GB of RAM due to limitations brought by the design of its hardware.^[54] In May 2017, Nvidia announced a partnership with Toyota which will use Nvidia's Drive PX-series artificial intelligence platform for its autonomous vehicles.^[55] In July 2017, Nvidia and Chinese search giant Baidu announced a far-reaching AI partnership that includes cloud computing, autonomous driving, consumer devices, and Baidu's open-source AI framework PaddlePaddle. Baidu unveiled that Nvidia's Drive PX 2 AI will be the foundation of its autonomous-vehicle platform.^[56]

Nvidia officially released the Titan V on December 7, 2017.^[57][58]

Nvidia officially released the Nvidia Quadro GV100 on March 27, 2018.^[59] Nvidia officially released the RTX 2080 GPUs in September 27, 2018. In 2018, Google announced that Nvidia's Tesla P4 graphic cards would be integrated into Google Cloud service's artificial intelligence.^[60]

In May 2018, on the Nvidia user forum, a thread was started^[61] asking the company to update users when they would release web drivers for its cards installed on legacy Mac Pro machines up to mid-2012 5,1 running the macOS Mojave operating system 10.14. Web drivers are required to enable graphics acceleration and multiple display monitor capabilities of the GPU. On its Mojave update info website, Apple stated that macOS Mojave would run on legacy machines with 'Metal compatible' graphics cards^[62] and listed Metal compatible GPUs, including some manufactured by Nvidia.^[63] However, this list did not include Metal compatible cards that currently work in macOS High Sierra using Nvidia developed web drivers. In September, Nvidia responded, "Apple fully control drivers for Mac OS. But if Apple allows, our engineers are ready and eager to help Apple deliver great drivers for Mac OS 10.14 (Mojave)."^[64] In October, Nvidia followed this up with another public announcement, "Apple fully controls drivers for Mac OS. Unfortunately, Nvidia currently cannot release a driver unless it is approved by Apple,"^[65] suggesting a possible rift between the two companies.^[66] By January 2019, with still no sign of the enabling web drivers, Apple Insider weighed into the controversy with a claim that Apple management "doesn't want Nvidia support in macOS".^[67] The following month, Apple Insider followed this up with another claim that Nvidia support was abandoned because of "relational issues in the past",^[68] and that Apple was developing its own GPU technology.^[69] Without Apple approved Nvidia web drivers, Apple users are faced with replacing their Nvidia cards with a competing supported brand, such as AMD Radeon from the list recommended by Apple.^[70]

On March 11, 2019, Nvidia announced a deal to buy Mellanox Technologies for $6.9 billion^[71] to substantially expand its footprint in the high-performance computing market. In May 2019, Nvidia announced new RTX Studio laptops. The creators say that the new laptop is going to be seven times faster than a top-end MacBook Pro with a Core i9 and AMD's Radeon Pro Vega 20 graphics in apps like Maya and RedCine-X Pro.^[72] In August 2019, Nvidia announced Minecraft RTX, an official Nvidia-developed patch for the game Minecraft adding real-time DXR raytracing exclusively to the Windows 10 version of the game. The whole game is, in Nvidia's words, "refit" with path tracing, which dramatically affects the way light, reflections, and shadows work inside the engine.^[73]

In May 2020, Nvidia's top scientists developed an open-sourceventilator in order to address the shortage resulting from the global coronavirus pandemic.^[74] On May 14, 2020, Nvidia officially announced their Ampere GPU microarchitecture and the Nvidia A100 GPU accelerator.^[75][76] In July 2020, it was reported that Nvidia was in talks with SoftBank to buy Arm, a UK-based chip designer, for $32 billion.^[77]

On September 1, 2020, Nvidia officially announced the GeForce 30 series based on the company's new Ampere microarchitecture.^[78][79]

On September 13, 2020, it was announced that Nvidia would buy Arm Holdings from SoftBank Group for $40 billion, subject to the usual scrutiny, with the latter retaining a 10% share of Nvidia.^{[16][15][80][81]}

In October 2020, Nvidia announced its plan to build the most powerful computer in Cambridge, England. Named Cambridge-1, the computer will employ AI to support healthcare research, with an expected completion by the end of 2020, at a cost of approximately £40 million. According to Jensen Huang, "The Cambridge-1 supercomputer will serve as a hub of innovation for the UK, and further the groundbreaking work being done by the nation’s researchers in critical healthcare and drug discovery."^[82]

Also in October 2020, along with the release of Nvidia RTX A6000, Nvidia announced it is retiring its workstation GPU brand Quadro, shifting product name to Nvidia RTX for future products and the manufacturing to be Nvidia Ampere architecture based.^[83]

In August 2021, the proposed takeover of Arm Holdings was stalled after the UK's Competition and Markets Authority raised "significant competition concerns".^[84]

Finances[edit]

For the fiscal year 2020, Nvidia reported earnings of US$2.796 billion, with an annual revenue of US$10.918 billion, a decline of 6.8% over the previous fiscal cycle. Nvidia's shares traded at over $531 per share, and its market capitalization was valued at over US$328.7 billion in January 2021.^[85]

For the Q2 of 2020, Nvidia reported sales of $3.87 billion, which was a 50% rise from the same period in 2019. The surge in sales and people's higher demand for computer technology. According to the financial chief of the company, Colette Kress, the effects of the pandemic will "likely reflect this evolution in enterprise workforce trends with a greater focus on technologies, such as Nvidia laptops and virtual workstations, that enable remote work and virtual collaboration."^[86]

GPU Technology Conference[edit]

Nvidia's GPU Technology Conference (GTC), now called Nvidia GTC, is a series of technical conferences held around the world.^[87] It originated in 2009 in San Jose, California, with an initial focus on the potential for solving computing challenges through GPUs.^[88] In recent years, the conference focus has shifted to various applications of artificial intelligence and deep learning, including: self-driving cars, healthcare, high performance computing, and Nvidia Deep Learning Institute (DLI) training.^[89] GTC 2018 attracted over 8400 attendees.^[87] GTC 2020 was converted to a digital event and drew roughly 59,000 registrants.^[90]

Product families[edit]

Nvidia's family includes graphics, wireless communication, PC processors, and automotive hardware/software.

Some families are listed below:

• GeForce, consumer-oriented graphics processing products
• Nvidia RTX, professional visual computing graphics processing products (replacing Quadro)
• NVS, multi-display business graphics solution
• Tegra, a system on a chip series for mobile devices
• Tesla, dedicated general-purpose GPU for high-end image generation applications in professional and scientific fields
• nForce, a motherboard chipset created by Nvidia for Intel (Celeron, Pentium and Core 2) and AMD (Athlon and Duron) microprocessors
• Nvidia GRID, a set of hardware and services by Nvidia for graphics virtualization
• Nvidia Shield, a range of gaming hardware including the Shield Portable, Shield Tablet and, most recently, the Shield Android TV
• Nvidia Drive automotive solutions, a range of hardware and software products for designers and manufacturers of autonomous vehicles. The Drive PX-series is a high performance computer platform aimed at autonomous driving through deep learning,^[91] while Driveworks is an operating system for driverless cars.^[92]
• BlueField, a range of Data Processing Units, initially inherited from their acquisition of Mellanox Technologies^[93][94]
• Nvidia Datacenter/Server class CPU, codenamed Nvidia Grace, coming in 2023^[95][96]

Open-source software support[edit]

See also: Free and open-source graphics device driver, Mesa 3D, nouveau (software), and OpenPOWER Foundation

Until September 23, 2013, Nvidia had not published any documentation for its advanced hardware,^[97] meaning that programmers could not write free and open-sourcedevice driver for its products without resorting to (clean room) reverse engineering.

Instead, Nvidia provides its own binary GeForce graphics drivers for X.Org and an open-source library that interfaces with the Linux, FreeBSD or Solaris kernels and the proprietary graphics software. Nvidia also provided but stopped supporting an obfuscated open-source driver that only supports two-dimensional hardware acceleration and ships with the X.Org distribution.^[98]

The proprietary nature of Nvidia's drivers has generated dissatisfaction within free-softwarecommunities.^[99] Some Linux and BSD users insist on using only open-source drivers and regard Nvidia's insistence on providing nothing more than a binary-only driver as inadequate, given that competing manufacturers such as Intel offer support and documentation for open-source developers and that others (like AMD) release partial documentation and provide some active development.^[100][101]

Because of the closed nature of the drivers, Nvidia video cards cannot deliver adequate features on some platforms and architectures given that the company only provides x86/x64 and ARMv7-A driver builds.^[102] As a result, support for 3D graphics acceleration in Linux on PowerPC does not exist, nor does support for Linux on the hypervisor-restricted PlayStation 3 console.

Some users claim that Nvidia's Linux drivers impose artificial restrictions, like limiting the number of monitors that can be used at the same time, but the company has not commented on these accusations.^[103]

In 2014, with Maxwell GPUs, Nvidia started to require firmware by them to unlock all features of its graphics cards. Up to now, this state has not changed and makes writing open-source drivers difficult.^{[104][105][106]}

Deep learning[edit]

Nvidia GPUs are used in deep learning, and accelerated analytics due to Nvidia's API CUDA which allows programmers to utilize the higher number of cores present in GPUs to parallelizeBLAS operations which are extensively used in machine learning algorithms.^[107] They were included in many Tesla vehicles before Elon Musk announced at Tesla Autonomy Day in 2019 that the company developed its own SoC and Full Self-Driving computer now and would stop using Nvidia hardware for their vehicles.^[108][109] These GPUs are used by researchers, laboratories, tech companies and enterprise companies.^[110] In 2009, Nvidia was involved in what was called the "big bang" of deep learning, "as deep-learning neural networks were combined with Nvidia graphics processing units (GPUs)".^[111] That year, the Google Brain used Nvidia GPUs to create Deep Neural Networks capable of machine learning, where Andrew Ng determined that GPUs could increase the speed of deep-learning systems by about 100 times.^[112]

DGX[edit]

Main article: Nvidia DGX

DGX is a line of supercomputers by Nvidia.

In April 2016, Nvidia produced the DGX-1 based on an 8 GPU cluster, to improve the ability of users to use deep learning by combining GPUs with integrated deep learning software.^[113] It also developed Nvidia Tesla K80 and P100 GPU-based virtual machines, which are available through Google Cloud, which Google installed in November 2016.^[114]Microsoft added GPU servers in a preview offering of its N series based on Nvidia's Tesla K80s, each containing 4992 processing cores. Later that year, AWS's P2 instance was produced using up to 16 Nvidia Tesla K80 GPUs. That month Nvidia also partnered with IBM to create a software kit that boosts the AI capabilities of Watson,^[115] called IBM PowerAI.^[116][117] Nvidia also offers its own NVIDIA Deep Learning software development kit.^[118] In 2017, the GPUs were also brought online at the Riken Center for Advanced Intelligence Project for Fujitsu.^[119] The company's deep learning technology led to a boost in its 2017 earnings.^[120]

In May 2018, researchers at the artificial intelligence department of Nvidia realized the possibility that a robot can learn to perform a job simply by observing the person doing the same job. They have created a system that, after a short revision and testing, can already be used to control the universal robots of the next generation. In addition to GPU manufacturing, Nvidia provides parallel processing capabilities to researchers and scientists that allow them to efficiently run high-performance applications.^[121]

Inception Program[edit]

Nvidia's Inception Program was created to support startups making exceptional advances in the fields of artificial intelligence and data science. Award winners are announced at Nvidia's GTC Conference. As of March 2018, there were 2,800 startups in the Inception Program. As of August 2021, NVIDIA Inception has surpassed 8,500 members in 90 countries, with cumulative funding of $60 billion (USD).^{[122][123][124]}

2018 winners[edit]

• Subtle Medical (healthcare)
• AiFi (enterprise)
• Kinema Systems (autonomous vehicles)

2017 winners[edit]

• Genetesis (social innovation)
• Athelas (hottest emerging)
• Deep Instinct (most disruptive)

Controversies[edit]

Maxwell advertising dispute[edit]

GTX 970 hardware specifications[edit]

Issues with the GeForce GTX 970's specifications were first brought up by users when they found out that the cards, while featuring 4 GB of memory, rarely accessed memory over the 3.5 GB boundary. Further testing and investigation eventually led to Nvidia issuing a statement that the card's initially announced specifications had been altered without notice before the card was made commercially available, and that the card took a performance hit once memory over the 3.5 GB limit were put into use.^{[125][126][127]}

The card's back-end hardware specifications, initially announced as being identical to those of the GeForce GTX 980, differed in the amount of L2 cache (1.75 MB versus 2 MB in the GeForce GTX 980) and the number of ROPs (56 versus 64 in the 980). Additionally, it was revealed that the card was designed to access its memory as a 3.5 GB section, plus a 0.5 GB one, access to the latter being 7 times slower than the first one.^[128] The company then went on to promise a specific driver modification in order to alleviate the performance issues produced by the cutbacks suffered by the card.^[129] However, Nvidia later clarified that the promise had been a miscommunication and there would be no specific driver update for the GTX 970.^[130] Nvidia claimed that it would assist customers who wanted refunds in obtaining them.^[131] On February 26, 2015, Nvidia CEO Jen-Hsun Huang went on record in Nvidia's official blog to apologize for the incident.^[132] In February 2015 a class-action lawsuit alleging false advertising was filed against Nvidia and Gigabyte Technology in the U.S. District Court for Northern California.^[133][134]

Nvidia revealed that it is able to disable individual units, each containing 256KB of L2 cache and 8 ROPs, without disabling whole memory controllers.^[135] This comes at the cost of dividing the memory bus into high speed and low speed segments that cannot be accessed at the same time unless one segment is reading while the other segment is writing because the L2/ROP unit managing both of the GDDR5 controllers shares the read return channel and the write data bus between the two GDDR5 controllers and itself.^[135] This is used in the GeForce GTX 970, which therefore can be described as having 3.5 GB in its high speed segment on a 224-bit bus and 0.5 GB in a low speed segment on a 32-bit bus.^[135]

On July 27, 2016, Nvidia agreed to a preliminary settlement of the U.S. class action lawsuit,^[133] offering a $30 refund on GTX 970 purchases. The agreed upon refund represents the portion of the cost of the storage and performance capabilities the consumers assumed they were obtaining when they purchased the card.^[136]

Async compute support[edit]

While the Maxwell series was marketed as fully DirectX 12 compliant,^[137][138] Oxide Games, developer of Ashes of the Singularity, uncovered that Maxwell-based cards do not perform well when async compute is utilized.^{[139][140][141][137]}

It appears that while this core feature is in fact exposed by the driver,^[142] Nvidia partially implemented it through a driver-based shim, coming at a high performance cost.^[141] Unlike AMD's competing GCN-based graphics cards which include a full implementation of hardware-based asynchronous compute,^[143][144] Nvidia planned to rely on the driver to implement a software queue and a software distributor to forward asynchronous tasks to the hardware schedulers, capable of distributing the workload to the correct units.^[145] Asynchronous compute on Maxwell therefore requires that both a game and the GPU driver be specifically coded for asynchronous compute on Maxwell in order to enable this capability.^[146] The 3DMark Time Spy benchmark shows no noticeable performance difference between asynchronous compute being enabled or disabled.^[146] Asynchronous compute is disabled by the driver for Maxwell.^[146]

Oxide claims that this led to Nvidia pressuring them not to include the asynchronous compute feature in their benchmark at all, so that the 900 series would not be at a disadvantage against AMD's products which implement asynchronous compute in hardware.^[140]

Maxwell requires that the GPU be statically partitioned for asynchronous compute to allow tasks to run concurrently.^[147] Each partition is assigned to a hardware queue. If any of the queues that are assigned to a partition empty out or are unable to submit work for any reason (e.g. a task in the queue must be delayed until a hazard is resolved), the partition and all of the resources in that partition reserved for that queue will idle.^[147] Asynchronous compute therefore could easily hurt performance on Maxwell if it is not coded to work with Maxwell's static scheduler.^[147] Furthermore, graphics tasks saturate Nvidia GPUs much more easily than they do to AMD's GCN-based GPUs which are much more heavily weighted towards compute, so Nvidia GPUs have fewer scheduling holes that could be filled by asynchronous compute than AMD's.^[147] For these reasons, the driver forces a Maxwell GPU to place all tasks into one queue and execute each task in serial, and give each task the undivided resources of the GPU no matter whether or not each task can saturate the GPU or not.^[147]

GeForce Partner Program[edit]

Main article: GeForce Partner Program

The Nvidia GeForce Partner Program was a marketing program designed to provide partnering companies with benefits such as public relations support, video game bundling, and marketing development funds.^[148] The program proved to be controversial, with complaints about it possibly being an anti-competitive practice.^[149]

First announced in a blog post on March 1, 2018,^[150] it was canceled on May 4, 2018.^[151]

Hardware Unboxed controversy[edit]

On December 10, 2020, Nvidia told popular YouTube tech reviewer Steven Walton of Hardware Unboxed that it would no longer supply him with GeForce Founders Edition graphics card review units.^[152][153] In a Twitter message, Hardware Unboxed said, "Nvidia have officially decided to ban us from receiving GeForce Founders Edition GPU review samples. Their reasoning is that we are focusing on rasterization instead of ray tracing. They have said they will revisit this 'should your editorial direction change.'"^[154]

In emails that were disclosed by Walton from Nvidia Senior PR Manager Bryan Del Rizzo, Nvidia had said:

...your GPU reviews and recommendations have continued to focus singularly on rasterization performance, and you have largely discounted all of the other technologies we offer gamers. It is very clear from your community commentary that you do not see things the same way that we, gamers, and the rest of the industry do.^[155]

TechSpot, partner site of Hardware Unboxed, said, "this and other related incidents raise serious questions around journalistic independence and what they are expecting of reviewers when they are sent products for an unbiased opinion."^[155]

A number of prominent technology reviewers came out strongly against Nvidia's move.^[156][157]Linus Sebastian, of Linus Tech Tips, titled the episode of his popular weekly WAN Show, "NVIDIA might ACTUALLY be EVIL..."^[158] and was highly critical of the company's move to dictate specific outcomes of technology reviews.^[159] The popular review site Gamers Nexus said it was, "Nvidia's latest decision to shoot both its feet: They've now made it so that any reviewers covering RT will become subject to scrutiny from untrusting viewers who will suspect subversion by the company. Shortsighted self-own from NVIDIA."^[160]

Two days later, Nvidia reversed their stance.^[161][162] Hardware Unboxed sent out a Twitter message, "I just received an email from Nvidia apologizing for the previous email & they've now walked everything back."^[163][156] On December 14, Hardware Unboxed released a video explaining the controversy from their viewpoint.^[164] Via Twitter, they also shared a second apology sent by Nvidia's Del Rizzo that said "to withhold samples because I didn't agree with your commentary is simply inexcusable and crossed the line."^[165][166]

See also[edit]

Notes[edit]

References[edit]

Wikipedia list article

This list contains general information about graphics processing units (GPUs) and video cards from Nvidia, based on official specifications. In addition some Nvidia motherboards come with integrated onboard GPUs. Limited/Special/Collectors' Editions or AIB versions are not included.

Field explanations[edit]

The fields in the table listed below describe the following:

• Model – The marketing name for the processor, assigned by Nvidia.
• Launch – Date of release for the processor.
• Code name – The internal engineering codename for the processor (typically designated by an NVXY name and later GXY where X is the series number and Y is the schedule of the project for that generation).
• Fab – Fabrication process. Average feature size of components of the processor.
• Bus interface – Bus by which the graphics processor is attached to the system (typically an expansion slot, such as PCI, AGP, or PCI-Express).
• Memory – The amount of graphics memory available to the processor.
• SM Count – Number of streaming multiprocessors.^[1]
• Core clock – The factory core clock frequency; while some manufacturers adjust clocks lower and higher, this number will always be the reference clocks used by Nvidia.
• Memory clock – The factory effective memory clock frequency (while some manufacturers adjust clocks lower and higher, this number will always be the reference clocks used by Nvidia). All DDR/GDDR memories operate at half this frequency, except for GDDR5, which operates at one quarter of this frequency.
• Core config – The layout of the graphics pipeline, in terms of functional units. Over time the number, type, and variety of functional units in the GPU core has changed significantly; before each section in the list there is an explanation as to what functional units are present in each generation of processors. In later models, shaders are integrated into a unified shader architecture, where any one shader can perform any of the functions listed.
• Fillrate – Maximum theoretical fill rate in textured pixels per second. This number is generally used as a maximum throughput number for the GPU and generally, a higher fill rate corresponds to a more powerful (and faster) GPU.
• Memory subsection
  • Bandwidth – Maximum theoretical bandwidth for the processor at factory clock with factory bus width. GHz = 10⁹ Hz.
  • Bus type – Type of memory bus or buses used.
  • Bus width – Maximum bit width of the memory bus or buses used. This will always be a factory bus width.
• API support section
  • Direct3D – Maximum version of Direct3D fully supported.
  • OpenGL – Maximum version of OpenGL fully supported.
  • OpenCL – Maximum version of OpenCL fully supported.
  • Vulkan – Maximum version of Vulkan fully supported.
• Features – Added features that are not standard as a part of the two graphics libraries.

Desktop GPUs[edit]

Pre-GeForce[edit]

See Fahrenheit (microarchitecture)

GeForce 256 series[edit]

Further information: GeForce 256 and Celsius (microarchitecture)

• All models are made via 220 nm fabrication process
• All models support Direct3D 7.0 and OpenGL 1.2
• All models support hardware Transform and Lighting (T&L) and Cube Environment Mapping

GeForce2 series[edit]

Further information: GeForce 2 series and Celsius (microarchitecture)

• All models support Direct3D 7 and OpenGL 1.2
• All models support TwinView Dual-Display Architecture, Second Generation Transform and Lighting (T&L), Nvidia Shading Rasterizer (NSR), High-Definition Video Processor (HDVP)
• GeForce2 MX models support Digital Vibrance Control (DVC)

GeForce3 series[edit]

Further information: GeForce 3 series and Kelvin (microarchitecture)

• All models are made via 150 nm fabrication process
• All models support Direct3D 8.0 and OpenGL 1.3
• All models support 3D Textures, Lightspeed Memory Architecture (LMA), nFiniteFX Engine, Shadow Buffers

GeForce4 series[edit]

Further information: GeForce 4 series and Kelvin (microarchitecture)

• All models are manufactured with a 150 nm manufacturing process
• All models support Accuview Antialiasing (AA), Lightspeed Memory Architecture II (LMA II), nView

• ¹Pixel shaders: vertex shaders: texture mapping units: render output units
• ² GeForce4 Ti4400 8x: Card manufacturers utilizing this chip, labeled the card as a Ti4800SE. The surface of the chip has "Ti-8x" printed on it.
• ³ GeForce4 Ti4600 8x: Card manufacturers utilizing this chip, labeled the card as a Ti4600, and in some cases as a Ti4800. The surface of the chip has "Ti-8x" printed on it, as well as "4800" printed at the bottom.

GeForce FX (5xxx) series[edit]

Further information: GeForce FX series and Rankine (microarchitecture)

• All models support Direct3D 9.0a and OpenGL 1.5 (2.1 (software) with latest drivers)
• The GeForce FX series runs vertex shaders in an array

GeForce 6 (6xxx) series[edit]

Further information: GeForce 6 series and Curie (microarchitecture)

• All models support Direct3D 9.0c and OpenGL 2.1
• All models support Transparency AA (starting with version 91.47 of the ForceWare drivers) and PureVideo

Features[edit]

GeForce 7 (7xxx) series[edit]

Further information: GeForce 7 series and Curie (microarchitecture)

• All models support Direct3D 9.0c and OpenGL 2.1
• All models support Transparency AA (starting with version 91.47 of the ForceWare drivers)