Bookmark and Share


When Nvidia Corp. unveiled its Fermi architecture first in September, 2009, the firm said that the top-of-the-range offering will include 512 stream processors. However, it released no actual products on the fully-fledged chip. On Wednesday a web-site published a picture of a graphics card that it claims to be powered by the code-named GF100 graphics chip with all 512 stream processors activated.

The graphics solution resembles the original GeForce GTX 480, which comes with 480 stream processors activated, but has tangibly improved power supply circuitry and consumes much more power as it is equipped with two PCIe 8-pin power connectors (together with PCIe x16 slot they can provide up to 375W of power). Based on the image published by Expreview web-site, the board also looks slightly longer than the predecessor, possibly because of higher power consumption and greater heat dissipation.

Nvidia GeForce GTX graphics card. Image by Expreview web-site.

The board carries a chip marked as "4", twelve GDDR5 memory chips and a regular set of connectors, including two dual-link DVI, one HDMI, two MIO ports for 4-way SLI multi-GPU operation support and so on. Considering that all the formal marks of the chip are blurred, just like the revision of the graphics processing unit (GPU) on a screenshot from GPU-Z software, there are some chances that Nvidia plans to use a new revision of the GF100 graphics chip.

Based on the screenshot from GPU-Z program, the board has clock-speeds similar to those of the original GeForce GTX 480: 650MHz for the graphics chip, 1400MHz for stream processors and 3700MHz for 1536MB of GDDR5 memory. The card surprisingly carries the same Device ID as the original GeForce GTX 480 and the software indicated March 26, 2010, release date. Usually, different graphics boards carry different Device IDs, based on which programs like GPU-Z determine the amount of stream processors and other peculiarities. Performance of the board obtained using an Intel Core i7-920-based system (3.80GHz) along with Futuremark 3DMark Vantage also cause more questions than answers since the board only achieves P23904 points, whereas a similar system without overclocked CPU with GeForce GTX 480 gets P21427 points with older drivers.

Even though Nvidia said earlier this year that there were chances to release a graphics card with fully-functional GF100 GPU, the company has never set any dates or concrete plans for it. Provided that Nvidia will need a more powerful offering to compete against ATI's forthcoming Radeon HD 6000 graphics processors due in October or November, it is likely that the company will indeed try to bin certain amount of GF100 chips with 512 stream processors and sell them as limited edition products. The firm may also completely refresh the lineup by introducing GeForce GTX 485 and redesign the remaining high-end offerings.

Nvidia did not comment on the news-story.

Tags: Nvidia, Geforce, Fermi, GF100, 40nm


Comments currently: 2
Discussion started: 08/04/10 06:20:06 PM
Latest comment: 08/05/10 03:59:34 AM
Expand all threads | Collapse all threads


"Based on the screenshot from GPU-Z program, the board has clock-speeds similar to those of the original GeForce GTX 480: 650MHz for the graphics chip, 1400MHz for stream processors and 3700MHz for 1536MB of GDDR5 memory."

This can't be right since the shader processors run at 2:1 ratio relative to GPU clock speed. If the stream processors are operating at 1400Mhz, the GPU speed would be 700 Mhz not 650.
0 0 [Posted by: BestJinjo  | Date: 08/04/10 06:20:06 PM]
- collapse thread

It is 700MHz. The screen shoot on the GPU-Z on the Expreview reads 701/1401/924MHz.
0 0 [Posted by: jonup  | Date: 08/05/10 03:59:34 AM]


Add your Comment

Related news

Latest News

Monday, July 28, 2014

6:02 pm | Microsoft’s Mobile Strategy Seem to Fail: Sales of Lumia and Surface Remain Low. Microsoft Still Cannot Make Windows a Popular Mobile Platform

12:11 pm | Intel Core i7-5960X “Haswell-E” De-Lidded: Twelve Cores and Alloy-Based Thermal Interface. Intel Core i7-5960X Uses “Haswell-EP” Die, Promises Good Overclocking Potential

Tuesday, July 22, 2014

10:40 pm | ARM Preps Second-Generation “Artemis” and “Maya” 64-Bit ARMv8-A Offerings. ARM Readies 64-Bit Cores for Non-Traditional Applications

7:38 pm | AMD Vows to Introduce 20nm Products Next Year. AMD’s 20nm APUs, GPUs and Embedded Chips to Arrive in 2015

4:08 am | Microsoft to Unify All Windows Operating Systems for Client PCs. One Windows OS will Power PCs, Tablets and Smartphones