enfrdepluk
Search find 4120  disqus socia  tg2 f2 lin2 in2 X icon 3 y2  p2 tik steam2

GeForce GTX 590

In essence, the GeForce GTX 590 is an evolutionary product, since NVIDIA did not use any new technologies or change the architecture during its production. GPU GF110, however, AMD has a similar situation with their Radeon HD 6990. Architecturally, the GF110 chips used in the GeForce GTX 590 are not very different from the GF100, which is the basis for the first model of the Fermi line - GTX 480. Both GPU uses the same multiprocessor configuration, these chips consist of four Graphics Processing Clusters.

1

The GeForce GTX 590 accelerator will generally show performance similar to the accelerated version of the Radeon HD 6990, slightly outperforming AMD's flagship product in games that actively use complex tessellation modes and lagging behind in those applications that have always been AMD's strength. Each of the two GF110s contains four GPC clusters, sixteen SM multiprocessors and six 64-bit memory controllers connected to L8 cache and having 512 ROP units each. In total, each chip includes 16 stream processors, assembled into 32 multiprocessors of 590 pieces each. And in the released dual-chip GTX XNUMX model, their number is not artificially reduced; all multiprocessors are active.

Each multiprocessor in the GF110 has 64 kilobytes of on-chip memory, which can be configured in two different options: 48 kilobytes of shared memory and 16 kilobytes of L1 cache, or vice versa - 16 KB of shared memory and 48 KB of cache. In addition, the GF110 has 768 KB of unified LXNUMX cache, which services all data load and save requests, as well as texture fetches.

Characteristics of NVIDIA GeForce GTX 590

Name GeForce GTX 590
Core GF110 x 2
Process technology (µm) 0.04
Transistors (millions) ~3000x2
Core frequency 607
Memory operating frequency (DDR) 1707 (3400)
Bus and memory type 384-bit x 2 GDDR5
Bandwidth (Gb/s) 327.7
Unified shader units 512 x 2
Unified shader unit frequency 1215
TMU on conveyor 64 x 2
ROP 48 x 2
Shader Model 5.0
Fill Rate (Mpix/s) 58272
Fill Rate (Mtex/s) 77696
DirectX 11.0
Memory Capacity 1536 x 2
Interface PCI-E 2.0

This model contains two GF110 graphics processors, which are manufactured using the 40-nanometer process technology based on the Fermi architecture. The device also has 1024 stream processors, 128 texture units, 96 raster operation units and 3 GB of GDDR5 memory with a 2x384-bit access bus. The operating frequencies of the video chip/shader domain/memory are set at 607/1214/3414 MHz, respectively. To remove heat from the components of the GeForce GTX 590 video card, a dual-slot cooler with a fan located in the center is used. At the same time, the cooler uses its own evaporation chamber for each graphics processor. The new product has a maximum power consumption of 365 W.

2

When installed in a PC, the GeForce GTX 590 accelerator occupies two expansion slots, which is currently standard. The length of the board itself is not much longer than the length of the GeForce GTX 580 reference sample and is 28 cm. The design of the GTX 590 cooling system is made in the color scheme typical of all modern NVIDIA products: a black casing with branded gray-green inserts that covers the entire printed circuit board. On the reverse side of the PCB there are pads that create rigidity and at the same time dissipate heat from the memory chips.

3

For normal operation, the GeForce GTX 590 requires the connection of two 8-pin PCI-Ex connectors. By the way, the manufacturer recommends the use of 700-W power supplies for stable operation of the system with the GTX 590 accelerator. At the moment, this video card is one of the fastest graphics solutions on the market, showing performance approximately equal to the Radeon HD 6990.