コンテンツにスキップ

英文维基 | 中文维基 | 日文维基 | 草榴社区

利用者:Tommy6/新規作成/GeForce 8 シリーズ

Template:NvidiaGPU

GeForce 8 シリーズ(ジーフォース エイト -)は、米NVIDIA社が開発したコンシューマ向けグラフィックチップ(GPU)GeForceの第八代目にあたるシリーズである。このシリーズでは、NVIDIA社として初めて統合シェーダアーキテクチャを採用しており、同社のGPU開発における三度目の抜本的改革の象徴となっている。

The GeForce 8 Series is the eighth generation of NVIDIA's GeForce graphics cards. The series also represents the third fundamentally new GPU design developed at NVIDIA as well as the company's first unified shader architecture.[1][2]

全体像,3Dレンダリング

GeForce 8 Series Overview

[編集]

3D rendering

[編集]

GeForce 8 シリーズ ではNVIDIAで初めて、統合シェーダ・Direct3D 10・シェーダモデル 4.0及びOpenGL 2.1を採用している。

The GeForce 8 series arrives with NVIDIA's first unified shader Direct3D 10 Shader Model 4.0 / OpenGL 2.1 architecture. The design is a major shift for NVIDIA in GPU functionality and capability, the most obvious change being the move from the separate functional units (pixel shaders, vertex shaders) within previous GPUs to a homogeneous collection of universal floating point processors (called "stream processors") that can perform a more universal set of tasks.


GeForce 8's unified shader architecture consists of a number of stream processors (SPs). Unlike the vector processing approach taken with older shader units, each SP is scalar and thus can operate only on one component at a time. This makes them less complex to build while still being quite flexible and universal. Scalar shader units also have the advantage of being more efficient in a number of cases as compared to previous generation vector shader units that rely on ideal instruction mixture and ordering to reach peak throughput. The lower maximum throughput of these scalar processors is compensated for by efficiency and by running them at a high clock speed (made possible by their simplicity). GeForce 8 runs the various parts of its core at differing clock speeds (clock domains), similar to the operation of the previous GeForce 7 Series GPUs. For example, the stream processors of GeForce 8800 GTX operate at a 1.35 GHz clock rate while the rest of the chip is operating at 575 MHz.[2]

GeForce 8 performs significantly better texture filtering than its predecessors that used various optimizations and visual tricks to speed up rendering while impairing filtering quality. The GeForce 8 line correctly renders an angle-independent anisotropic filtering algorithm along with full trilinear texture filtering. G80, though not its smaller brethren, is equipped with much more texture filtering arithmetic ability than the GeForce 7 series. This allows high-quality filtering with a much smaller performance hit than previously.[2]

NVIDIA has also introduced new polygon edge anti-aliasing methods, including the ability of the GPU's ROPs to perform both Multisample anti-aliasing (MSAA) and HDR lighting at the same time, correcting various limitations of previous generations. GeForce 8 can perform MSAA with both FP16 and FP32 texture formats. GeForce 8 supports 128-bit HDR rendering, an increase from prior cards' 64-bit support. The chip's new anti-aliasing technology, called coverage sampling AA (CSAA), uses Z, color, and coverage information to determine final pixel color. This technique of color optimization allows 16X CSAA to look crisp and sharp.[3]

The claimed theoretical processing power for the 8 Series cards given in FLOPS may not be correct at all times. For example the GeForce 8800 GTX has 518.4 GigaFLOPs theoretical performance given the fact that there are 128 stream processors at 1.35GHz with each SP being able to run 1 Multiple-Add and 1 Multiply instruction per clock [(MADD (2 FLOPs) + MUL (1 FLOP))×1350MHz×128 SPs = 518.4 GigaFLOPs][4]. This figure may not be correct because the Multiply operation is not always available[5] giving a possibly more accurate performance figure of (2×1350×128) = 345.6 GigaFLOPs.

Display capabilities

[編集]

The GeForce 8 series supports 10-bit per channel display output, up from 8-bit on previous NVIDIA cards. This potentially allows higher fidelity color representation and separation on capable displays. The GeForce 8 series, like its recent predecessors, also supports Scalable Link Interface (SLI) for multi-card rendering.

NVIDIA's PureVideo HD video rendering technology is an improved version of the original PureVideo introduced with GeForce 6. It now includes GPU-based hardware acceleration for decoding HD movie formats, post-processing of HD video for enhanced images, and optional High-bandwidth Digital Content Protection (HDCP) support at the card level.[6]

Here is a list of how selected GeForce 8 GPUs compare to NVIDIA's previous single-card flagship GeForce 7900 GTX and ATi's flagship Radeon HD 2900 XT.

GeForce 7900 GTX GeForce 8500 GT GeForce 8600 GT GeForce 8600 GTS Geforce 8800GT GeForce 8800 GTX GeForce 8800 Ultra Radeon HD 2900 XT
Transistor count 278 million 210 million 289 million 289 million 754 million 681 (~686) million 681 (~686) million 700 million
Manufacturing process 90 nm 80 nm 80 nm 80 nm 65 nm 90 nm 90 nm 80 nm
Die Area 196 mm² 132 mm² 169 mm² 169 mm² 315mm² 480 mm² 480 mm² 425 mm²
Core clock speed 650 MHz 450 MHz 540 MHz 675 MHz 600 MHz 575 MHz 612 MHz 743 MHz
Shader clock speed 650 (700) MHz 900 MHz 1.18 GHz 1.45 GHz 1.5 GHz 1.35 GHz 1.5 GHz 743 MHz
Number of
Shader Processing units
24 + (8) 16 32 32 112 128 128 320
Number of ROPs 16 4 8 8 16 24 24 16
Number of TMUs 24 8 16 16 28 32 32 16
Peak pixel fill rate
(theoretical)
10.4 Gigapixel/s 3.6 Gigapixel/s 4.3 Gigapixel/s 5.4 Gigapixel/s 9.6 Gigapixel/s 13.8 Gigapixel/s 14.7 Gigapixel/s 11.9 Gigapixel/s
Peak texture fill-rate
(theoretical)
15.6 Gigatexel/s 3.6 Gigatexel/s 8.6 Gigatexel/s 10.8 Gigatexel/s 33.6 Gigatexel/s 36.8 Gigatexel/s 39.2 Gigatexel/s 11.9 Gigatexel/s
Video-block
PureVideo 1 PureVideo 2 PureVideo 2 PureVideo 2 PureVideo 2 PureVideo 1 PureVideo 1 AVIVO HD
On-board memory interface 256 (4*64-bit) 128 (2*64-bit) 128 (2*64-bit) 128 (2*64-bit) 256 (4*64-bit) 384 (6*64-bit) 384 (6*64-bit) 512 (8*64-bit)
Memory clock speed 1.6 GHz GDDR3 800 MHz GDDR2 1.4 GHz GDDR3 2.0 GHz GDDR3 1.8 GHz GDDR3 1.8 GHz GDDR3 2.16 GHz GDDR3 1.65 GHz GDDR3
2.2 GHz GDDR4
Peak memory bandwidth 51.2 GB/s 12.8 GB/s 22.4 GB/s 32.0 GB/s 57.6 GB/s 86.4 GB/s 103.68 GB/s 105.6 GB/s (GDDR3)
140.8 GB/s (GDDR4)

GeForce 8800

[編集]
EVGA GeForce 8800 GTX
Underside

The 8800 series, codenamed G80, was launched on November 8, 2006 with the release of the GeForce 8800 GTX and GTS. A 320 MB GTS was released on February 12th and the Ultra was released on May 2nd 2007. The cards are larger than their predecessors, with the 8800 GTX measuring 10.6 in (~26.9 cm) in length and the 8800 GTS measuring 9 in (~23 cm). Both cards have two dual-link DVI connectors and a HDTV/S-Video out connector. The 8800 GTX requires 2 PCIe power inputs to keep within the PCIe standard, while the GTS requires just one.

8800 GTX / Ultra

[編集]

The GTX is equipped with 768 MB GDDR3 RAM. The 8800 series replaced the GeForce 79x0 series as NVIDIA's top-performing consumer video card. GeForce 8800 GTX and GTS use identical GPU cores, but the GTS model disables parts of the GPU and reduces RAM size and bus width to lower production cost.

As of September 2007, the G80 was the largest commercial GPU ever constructed. It consists of 681 million transistors covering a 480 mm² die surface area built on a 90 nm process. (In fact the G80's total transistor count is ~686 million, but since the chip was made on a 90nm process and due to process limitations and yield feasibility, NVIDIA had to break the main design into two chips: Main shader core at 681 million transistors and NV I/O core of about ~5 million transistors making the entire G80 design standing at ~686 million transistors).

A minor manufacturing defect related to a resistor of improper value caused a recall of the 8800 GTX models just two days before the product launch, though the launch itself was unaffected.[7]

The 8800 Ultra, retailing at a higher price is identical to the GTX architecturally, but features higher clocked shaders, core and memory.

8800 GT

[編集]

The 8800 GT, codenamed G92, was released on October 29, 2007. The card is the first to transition to 65nm process, and supports PCI-Express 2.0.[8] It has a single-slot cooler as opposed to the double slot cooler on the 8800 GTS and GTX. The 8800 GT, unlike other 8800 cards, is equipped with the PureVideo 2 engine for GPU assisted decoding of the H.264 and VC-1 codecs. Performance benchmarks at stock speeds place it above the 8800GTS (640MB and 320MB versions) and slightly below the 8800GTX. Presently, cards utilizing the chip are retailing from USD $269 (reference models) to USD $299 (for overclocked models) MSRP(512MB) in the US, and £165 for reference models and up to £190 for overclocked versions in the UK.

The release of this card presents an odd dynamic to the graphics processing industry. At a NVIDIA projected street price of around $200, this card outperforms ATI's flagship HD2900XT and HD3870 in most situations, and even NVIDIA's own 8800GTS (previously priced at an MSRP of $400). The card, only marginally worse in synthetic and gaming benchmarks than the 8800GTX, also takes much of the value away from NVIDIA's own high end card. This release was shortly followed by the 8800GTS SSC (the original 8800GTS re-released with 112 shader processor units), and ATI's counter, the HD3800 series.

8800 GTS

[編集]

The first release of the GTS line came in 640 MB and 320 MB configurations of GDDR3 RAM.[9] The 8800 GTX has 8 clusters of 16 stream processors, for a total of 128 stream processors. The first 8800 GTS, in comparison, features 96 stream processors.

The 320 MB version was released in February to tap into a more mainstream market. Aside from the decreased amount of video memory, all other aspects of the 8800 GTS remained unchanged. Despite this, the 320 MB version performs near identically to the 640 MB version in games at resolutions up to 1680 x 1050- the standard set by monitors up to 22 in wide. The unit retailed at US$299.[10].

Nvidia will also release a new 8800GTS 512MB based on G92 and 65nm process on December 11. The new 8800GTS will have 128 stream processors.

Technical Summary

[編集]
Model Release Date Codename Fabrication process (nm) Core clock max (MHz) Fillrate max (billion texel/s) Shaders Memory Power Consumption (Watts) Transistor Count (Millions) Shader Processing Power (Gigaflops)
Stream Processors Clock (MHz) Bandwidth max (GB/s) Bus type Bus width (bit) Megabytes Clock (MHz)
GeForce 8800 GTS (G80-100)[11][12][13] 8 November 2006 G80 90 500 24.00 96 1200 64.00 GDDR3 320 320 1600 108 681 (~690) 345.60
640
GeForce 8800 GT[14] 29 October 2007 G92 65 600 33.60 112 1500 44.8 GDDR3 256 256 1400 754 504
57.6 512 1800
GeForce 8800 GTS rev 2 [15][16][17][18] 11 December 2007 G92 65 650 41.6 128 1625 62.1 GDDR3 256 512 1940 754 624
GeForce 8800 GTX

[11][12][13]

8 November 2006 G80 90 575 36.80 128 1350 86.40 GDDR3 384 768 1800 145 681 (~690) 518.40
GeForce 8800 Ultra 2 May 2007 G80 90 612 39.17 128 1500 103.68 GDDR3 384 768 2160 175 681 (~690) 576.00

GeForce 8500 and 8600

[編集]

On April 17th, 2007, NVIDIA released 3 new members of the GeForce 8 product family: the GeForce 8500 GT; 8600 GT; and 8600 GTS. These products are based upon the G84 and G86 cores which are much smaller than the G80 core used for GeForce 8800, for mid-end users that want low cost DirectX 10-capable graphics cards.

With regards to performance, the 8600 GT performs slightly worse than X1950 GT and the 8600 GTS is similar to an ATi Radeon X1950 Pro.[19] The GeForce 8 series midrange cards seem to take a larger hit on performance than same price competitors when AA is enabled.[20] Some graphics card manufacturers, such as BFG Technologies, Micro-Star International and XFX, are releasing factory overclocked versions of the 8600 series. Gigabyte Technology has released an 8500 GT "Turbo Force" models which has a core clock of 600MHz GPU and 700MHz GDDR3 memory (1400MHz effective).

8500 GT cards may use GDDR3 or "DDR2" memory. Cards which claim to use "DDR2" actually use standard DDR2 SDRAM chips designed for use as main system memory, and should not be confused with GDDR2.

The 8500/8600 family introduces the PureVideo2 engine. PureVideo2 improves upon PureVideo by adding more decoding-assistance for VC-1 and H264. With the 8500/8600, NVIDIA claims PCs with slow CPUs can play HD-DVD and Blu-ray without skipping frames. The functionality of PureVideo2 is similar to ATI's Universal Video Decoder minus bitstream processing/entropy hardware support. As of the latest beta drivers, PureVideo 2 support is available in both Windows XP and Windows Vista.

Technical summary

[編集]
Model Release Date Codename Fabrication process (nm) Core clock max (MHz) Fillrate max (billion texel/s) Shaders Memory Power Consumption (Watts) Transistor Count (Millions) Shader Processing Power (Gigaflops)
Stream Processors Clock (MHz) Bandwidth (GB/s) Bus type Bus width (bits) Size (MB) Clock (MHz)
GeForce 8300 GS (OEM)[21] July 2007 G86 80 450 1.80 8 900 6.40 GDDR2 64 128/256 800 ? 210 21.60
GeForce 8400 GS[21] 15 June 2007 G86 80 450 3.60 16 900 6.40 GDDR2 64 256/128 800 38 210 43.20
GeForce 8500 GT[22][21] 17 April 2007 G86 80 450 3.60 16 900 12.80 GDDR2 128 256/512 800 40 210 43.20
GeForce 8600 GT[21] 17 April 2007 G84 80 540 8.64 32 1190 22.40 GDDR3 128 256/512/1024 1400 43 289 114.24
GeForce 8600 GTS[21] 17 April 2007 G84 80 675 10.80 32 1450 32.00 GDDR3 128 256/512/1024 2000 71 289 139.20

GeForce 8M mobile GPUs

[編集]

On May 10, 2007, NVIDIA announced the availability for their first notebook GPUs through select OEMs. So far the lineup consists of the 8400M, 8600M and 8700M series chips.[23]

GeForce 8700M Series

[編集]

Announced chip is the GeForce 8700M GT version currently featured in the Toshiba X205 model in the United States while it is also available in the Toshiba x200 series in the European Union and Australia. It's also available in Clevo M570RU/M571RU and D900C/D901C laptops. Base platform for this chipset is MXM III module.

Dell has implemented the 8700M GT in the XPS M1730 notebook. This is Dell's gaming notebook system. It offers the M8700 GT in SLI setup only.

GeForce 8400M Series

[編集]

Announced chips are the GeForce 8400M G, the GeForce 8400M GS and the GeForce 8400M GT.

Technical summary

[編集]
Model Release Date Codename Fabrication process (nm) Core clock max (MHz) Fillrate max (billion texel/s) Shaders Memory Power Consumption (Watts) Transistor Count (Millions) Shader Processing Power (Gigaflops)
Stream Processors Clock (MHz) Bandwidth max (GB/s) Bus type Bus width (bit) Size (Megabytes) Clock (MHz)
GeForce 8400M G 9th May 2007 G86M 80 400 3.2 8 800 9.6 GDDR3 64 128/256 1200 - - 19.2
GeForce 8400M GS 9th May 2007 G86M 80 400 3.2 16 800 9.6 GDDR3 64 64/128/256 1200 - - 38.4
GeForce 8400M GT 9th May 2007 G86M 80 450 3.6 16 900 19.2 GDDR3 128 256/512 1200 - - 43.2
GeForce 8600M GS 9th May 2007 G86M 80 600 4.8 16 1200 22.4 GDDR3 128 256/512 1400 - - 57.6
GeForce 8600M GT 9th May 2007 G84M 80 475 7.6 32 950 12.8/22.4 GDDR2/GDDR3 128 128/256/512 800/1400 22 - 91.2
GeForce 8700M GT 12th June 2007 G84M 80 625 10 32 1250 25.6 GDDR3 128 128/256/512 1600 29 289 120.0
GeForce 8800M GTS[24] 19th November 2007 G92M 65 500 16 64 1250 51.2 GDDR3 256 512 1600 35 - -
GeForce 8800M GTX[25] 19th November 2007 G92M 65 500 24 96 1250 51.2 GDDR3 256 512 1600 37 - -

Note that the GeForce 8700M GT is simply a higher clocked version of the GeForce 8600M GT, except that it also comes with "dual-ranked" memory architecture.

Future developments

[編集]

Template:Future chip

  • The series will be succeeded by a yet-to-be-named GeForce series. The top-end product of the series, according to Michael Hara (VP of Investor Relations), will be capable of 1 TFLOPs per chip, based on a 65 nm fabrication process.

See also

[編集]

References

[編集]
  1. ^ Q3 2007 NVIDIA Corporation Earnings Conference. NVIDIA.com. November 9, 2006.
  2. ^ a b c Wasson, Scott. Nvidia's GeForce 8800 graphics processor, Tech Report, November 8, 2007.
  3. ^ Sommefeldt, Rys.NVIDIA G80: Image Quality Analysis, Beyond3D, December 12, 2006.
  4. ^ http://forum.beyond3d.com/showthread.php?t=33576&page=83
  5. ^ Sommefeldt, Rys. NVIDIA G80: Architecture and GPU Analysis - Page 11, Beyond3D, November 8, 2006.
  6. ^ Shrout, Ryan. NVIDIA GeForce 8800 GTX Review - DX10 and Unified Architecture, PC Perspective, November 8, 2006.
  7. ^ "Visionary". All 8800 GTX Cards Being Recalled, VR-Zone.com, November 6, 2006.
  8. ^ GeForce 8800GT 65nm and PCI-E 2.0 support, VR-Zone.com, accessed October 7, 2007.
  9. ^ GeForce 8800 Press Release, NVIDIA.com, accessed November 9, 2006.
  10. ^ Shilov, Anton. Nvidia Prepares GeForce 8800 GTS 320MB, xbitlabs.com, January 10, 2007.
  11. ^ a b GeForce 8800 specifications, NVIDIA.com, accessed November 9, 2006.
  12. ^ a b NVIDIA GeForce 8800 GTX/GTS Tech Report, TechARP.com, accessed April 10, 2007.
  13. ^ a b NVIDIA GeForce 8800 GTX/GTS Performance Preview, FiringSquad.com, accessed April 10, 2007.
  14. ^ [1]
  15. ^ [2], theINQUIRER.net, accessed October 8, 2007.
  16. ^ http://www.fudzilla.com/index.php?option=com_content&task=view&id=4300&Itemid=34
  17. ^ http://www.techarp.com/showarticle.aspx?artno=474
  18. ^ http://www.reghardware.co.uk/2007/11/26/nvidia_g92_geforce_8800_gts_512/
  19. ^ [3]
  20. ^ [4]
  21. ^ a b c d e [5], Theinquirer.net, accessed April 12, 2007.
  22. ^ "Mid-range GeForce 8000 series Launch Dates, Prices". [6], DailyTech.com, accessed April 8, 2007.
  23. ^ NVIDIA GeForce 8M Series, nvidia.com, May 10, 2007.
  24. ^ NVIDIA GeForce 8800M, NVIDIA.com, Nov 19, 2007.
  25. ^ NVIDIA GeForce 8800GTX Next week, tt-hardware.com, Nov 14, 2007.
[編集]

[[Category:Graphics cards]] [[Category:NVIDIA]]