 | This articleneeds additional citations forverification. Please helpimprove this article byadding citations to reliable sources. Unsourced material may be challenged and removed. Find sources: "NV1" – news ·newspapers ·books ·scholar ·JSTOR(April 2020) (Learn how and when to remove this message) |
.jpg) NV1 chip manufactured by SGS-Thomson under the model name STG2000 | |
| Release date | May 22, 1995 (1995-05-22)[1] |
|---|---|
| Manufactured by | SGS-Thomson Microelectronics |
| Designed by | Nvidia |
| Marketed by | Diamond Multimedia |
| Codename | NV1 |
| Fabrication process | 500 nm |
| History | |
| Successor | RIVA 128 |
| Support status | |
| Unsupported | |
TheNV1 wasNvidia's firstgraphics accelerator, introduced in May 1995 and released later that year as a multimediaPCI card.[2] Manufactured bySGS-Thomson Microelectronics, sometimes under the model nameSTG2000, the chip was sold in retail byDiamond as theDiamond Edge 3D card. The NV1 stood out for its use ofquadratic texture mapping, a departure from the triangularprimitives favored by competitors. The use of quadratics made it possible to port games from theSega Saturn; however, after the NV1 was introduced, Microsoft announced thatDirectX would exclusively support triangle primitives.[3] As a result, the NV1 failed to gain traction in the market.[4]
In addition to its2D/3D graphics core andVideo RAM orFPM DRAM memory, the NV1 card also integrated the functionality of a 32-channel playback-onlysound card, and had ajoystick port, along with ports for two Sega Saturn controllers. As such, it was marketed as a "multimedia card" that was a replacement for both a graphics card and aSound Blaster-compatible audio card inIBM PC compatible systems. However, this made it more expensive, and with many computer owners owning a sound card, the all-in-one design further hurt its market appeal.
TheNV2 was a follow-up chip developed for Sega'sDreamcast, but was ultimately abandoned. Nvidia shifted focus with its next product, theRIVA 128, which adopted triangle primitives and dropped the audio functionality. This alignment with Direct3D and a more streamlined design made the RIVA 128 a success.
Several Sega Saturn games saw NV1-compatible conversions on the PC such asPanzer Dragoon andVirtua Fighter Remix. However, the NV1 struggled in a market place full of several competing proprietary standards, and was marginalized by emerging trianglepolygon-based 2D/3D accelerators such as the low-costS3 Graphics ViRGE,Matrox Mystique,ATI Rage, andRendition Vérité V1000 among other early entrants. It ultimately did not sell well, despite being a promising and interesting device.[peacock prose]
NV1's biggest initial problem was its cost and overall quality. Although it offered credible 3D performance, its use of quadratic surfaces was not popular, and was quite different than typical polygon rendering. The audio portion of the card received merely acceptable reviews, with theGeneral MIDI receiving lukewarm responses at best (a critical component at the time due to the superior sound quality produced by competing products). The Sega Saturn console was a market failure compared to Sony'sPlayStation or Sega's earlierSega Genesis, and so the unique and somewhat limiting support of these gamepads was of limited benefit.Nvidia, by integrating all of these usually separate components, raised their costs considerably above what they would have been if the card had been designed solely for 3D acceleration.
During the NV1's release timeframe, the transition fromVLB/ISA (486s) to PCI (Pentiums and late model 486 boards) was taking place, and games often usedMIDI for music because PCs were still generally incapable of large-scale digital audio playback due to storage and processing power limitations. Reaching for the best music and sound quality, and flexibility withMS-DOS audio standards, often required 2 sound cards be used, or a sound card with a MIDIdaughtercard connector. Additionally, NV1's 2D speed and quality were not competitive with many of the high-end systems available at the time, especially the critical-for-games DOS graphics speed. Many customers were simply not interested in replacing their often-elaborate system setups with an expensive all-in-one board and so the heavy integration of NV1 hurt sales simply through inconvenience.
Market interest in the product quickly ended whenMicrosoft announced theDirectX specifications, based upon triangle polygon rendering. This release by Microsoft of a major industry-backedAPI that was generally incompatible with NV1 ended Nvidia's hopes of market leadership immediately. While demos of quadratic rendered round spheres looked good, experience had proved working with quadratic texture maps was extremely difficult. Even calculating simple routines was problematic. Nvidia did manage to put together limited Direct3D support, but it was slow and buggy (software-based), and no match for the native triangle polygon hardware on the market.[5]
Subsequent NV1 quadratic-related development continued internally as the NV2.
.jpg)
.jpg)
NV2 was to be NVIDIA's second PC3D accelerator graphics chip, but it was cancelled before completion. It was planned to succeed the NV1.
NV2 built upon its predecessor's unusual quadratic 3D-rendering architecture. It was initially considered for use in Sega's Dreamcast console, due to the relationship cultivated between NVidia and Sega over the porting of Sega arcade and Saturn console titles over to the PC platform, where the similarity in NV1's and Saturn's 3D-rendering architecture aided in the porting process. (The NV1 graphics cards had 2 Sega Saturn gamepad ports integrated so that Saturn titles could be easily ported over to the NV1 cards and have an equal gameplay experience.) However, experience with both Saturn and NV1's 3D-rendering architecture in the Saturn ultimately led the company to abandon quadratic 3D-rendering architecture altogether, in favor of a more traditional architecture that operated ontriangle primitives.[6]
NVIDIA's strong desire to stick with their maturing quadraticforward texture mapping technology was a great cause of friction between Sega and NVIDIA. One part of the equation was undoubtedly that Sega's PC games division. A quadratic 3Dgame engine would be very difficult to port over to just about any other contemporary 3D graphics hardware, all of which used triangle primitives andinverse-texture mapping. More importantly, although consumer 3D hardware was still in its infancy, there was general consensus within the industry that triangle primitives with inverse-texture mapping would be standard going forward. Sega ultimately selectedNEC/VideoLogic'sPowerVR2 to power the 3d-graphics in its Dreamcast console.[7]
Because the demand was not there from Sega, and the PC market had drastically changed direction away from QTM due to the popularity of the triangle polygon-basedOpenGL andDirectX, NVIDIA abandoned further development of the NV2 and started on a new architecture, a.k.a. "NV3" orRIVA 128.
Model | Launch | Transistors (million) | Die size (mm2) | Core clock (MHz) | Memory clock (MHz) | Core config[a] | Fillrate | Memory | TDP (Watts) | LatestAPI support | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
MOperations/s | MPixels/s | MTexels/s | MVertices/s | Size (MB) | Bandwidth (GB/s) | Bus type | Bus width (bit) | |||||||||||||
| STG-2000 | May 22, 1995 | NV1 | SGS 500 nm | 1[9] | 90 | PCI | 12 75 | 50 60 | 1:1:1 | 12 | 12 | 12 | 0 | 1 2 4 | 0.4 0.48 | FPM EDO VRAM | 64 | ? | 1.0 | n/a |