GeForce 500 series

The GeForce 500 Series is a family of graphics processing units developed by Nvidia, based on a refresh of the Fermi (microarchitecture) (GF-codenamed chips) used in the previous 400 series. Nvidia officially announced the GeForce 500 series on 9 November 2010 with the launch of the GeForce GTX 580.

Overview

The Nvidia Geforce 500 Series graphics cards are significantly modified versions of the Nvidia GeForce 400 Series graphics cards, in terms of performance and power management. Like the Nvidia GeForce 400 Series graphics cards, the Nvidia Geforce 500 Series graphics cards support DirectX 12.0, OpenGL 4.6, and OpenCL 1.1.

The refreshed Fermi chip is large: it includes 512 stream processors, grouped in 16 stream multiprocessors clusters (each with 32 CUDA cores), and is manufactured by TSMC in a 40 nm process.

The Nvidia GeForce GTX 580 graphics card is the first in the Nvidia GeForce 500 Series to use a fully enabled chip based on the refreshed Fermi architecture, with all 16 stream multiprocessors clusters and all six 64-bit memory controllers active. The new GF110 GPU was enhanced with full speed FP16 filtering (the previous generation GF100 GPU could only do half-speed FP16 filtering) and improved z-culling units.

On 25 January 2011, Nvidia launched the GeForce GTX 560 Ti, to target the "sweet spot" segment where price/performance ratio is considered important. With its more than 30% improvement over the GTX 460, and performance in between the Radeon HD 6870 and 6950 1GB, the GTX 560 Ti directly replaced the GeForce GTX 470.

On 17 February 2011, it was reported that the GeForce GTX 550 Ti would be launching on 15 March 2011. Although the GTX 550 Ti is a GF116 mainstream chip, Nvidia chose to name its new card the GTX 550 Ti, and not the GTS 550. Performance was shown to be at least comparable and up to 12% faster than the current Radeon HD 5770. Price-wise, the new card trod into the range occupied by the GeForce GTX 460 (768 MB) and the Radeon HD 6790.^[2]

On 24 March 2011, the GTX 590 was launched as the flagship graphics card for Nvidia. The GTX 590 is a dual-GPU card, similar to past releases such as the GTX 295, and boasted the potential to handle Nvidia's 3D Vision technology by itself.

On 13 April 2011, the GT 520 was launched as the bottom-end card in the range, with lower performance than the equivalent number cards in the two previous generations, the GT 220 and the GT 420. However, it supported DirectX 11 and was more powerful than the GeForce 210, the GeForce 310, and the integrated graphics options on Intel CPUs.

On 17 May 2011, Nvidia launched a less expensive (non-Ti) version of the GeForce GTX 560 to strengthen Nvidia's price-performance in the $200 range. Like the faster GTX 560 Ti that came before it, this video card was also faster than the GeForce GTX 460. Standard versions of this card performed comparably to the AMD Radeon HD 6870, and would eventually replace the GeForce GTX 460. Premium versions of this card operate at higher speed (factory overclocked), and are slightly faster than the Radeon 6870, approaching the performance of basic versions of the Radeon HD 6950 and the GeForce GTX 560 Ti.

On 28 November 2011, Nvidia launched the "GTX560Ti With 448 Cores".^[3] However, it does not use the silicon of the GTX560 series: it is a GF110 chip with two shader blocks disabled. The most powerful version of the 560 series, this card was widely known to be a "limited production" card and was used as a marketing tool making use of the popularity of the GTX560 brand for the 2011 Holiday season. The performance of the card resides between the regular 560Ti and 570.

Products

GeForce 500 (5xx) series

• ¹ Unified shaders : Texture mapping units : Render output units
• ² Each Streaming Multiprocessor (SM) in the GPU of GF110 architecture contains 32 SPs and 4 SFUs. Each Streaming Multiprocessor (SM) in the GPU of GF114/116/118 architecture contains 48 SPs and 8 SFUs. Each SP can fulfil up to two single precision operations FMA per clock. Each SFU can fulfil up to four operations SF per clock. The approximate ratio of operations FMA to operations SF is equal 4:1. The theoretical shader performance in single-precision floating point operations (FMA)[FLOPS_sp, GFLOPS] of the graphics card with shader count [n] and shader frequency [f, GHz], is estimated by the following: FLOPS_sp ˜ f × n × 2. Alternative formula: FLOPS _sp ˜ f × m × (32 SPs × 2(FMA)). [m] – SM count. Total Processing Power: FLOPS_sp ˜ f × m × (32 SPs × 2(FMA) + 4 × 4 SFUs) or FLOPS_sp ˜ f × n × 2.5.
• ³ Each SM in the GF110 contains 4 texture filtering units for every texture address unit. The complete GF110 die contains 64 texture address units and 256 texture filtering units.^[4] Each SM in the GF114/116/118 architecture contains 8 texture filtering units for every texture address unit but has doubled both addressing and filtering units.

All products are produced using a 40 nm fabrication process. All products support Direct X 12.0, OpenGL 4.6 and OpenCL 1.1.

GeForce 500M (5xxM) series

The GeForce 500M series for notebook architecture.

• ¹ Unified shaders : Texture mapping units : Render output units

Chipset table

See also

• GeForce 200 series
• GeForce 400 series
• GeForce 600 series
• GeForce 700 series
• GeForce 800M series
• GeForce 900 series
• GeForce 1000 series
• Nvidia Quadro
• Nvidia Tesla

Notes

• David Kanter (30 September 2009). "Inside Fermi: Nvidia's HPC Push". realworldtech.com. Retrieved 16 December 2010. 

References

External links