The Inquirer-Home

G92 gunning for 800 MHz+ clock

Much improved hardware specs, speed
Tue Aug 21 2007, 17:25
NVIDIA WANTS TO keep details about G92 and G98 under-wraps, but as the product is getting ready some details are slowly, but certainly slipping out of Graphzilla's cave.

Even though the company is trying to keep everyone in the dark, we managed to dig some information out for you. For starters, Nvidia is strongly on-board with DisplayPort and it is committed onto delivering DisplayPort in time for Dell's launch of the 2x08 series of displays, all of which should support DisplayPort connections. The launch should coincide with CES 2008 in Las Vegas. However, the company still has to sort out the problems that Charlie reported last week.

When it comes to performance, even if this chip was just a shrunk-down G80, the clock would give significant benefits in performance thanks to increased clock of the GPU itself. Luckily, for future owners, that is not the case - Nvidia reworked quite a lot of things from the original marchitecture to make this thing work faster.

We managed to learn that GPU clock has surpassed 800 MHz and that company plans to deliver the boards depending on yields they will get. So far, launch clock could be anywhere between 800 and 850 MHz, with 812/3, 824/5 and 830/5 being usual suspects. This time around, there is no hurry and Nvidia knows they can afford the launch of GTS/GTX and wait for R680 to show up and launch the Ultra version as a counter attack.

There is quite a lot Nvidia has to support with the G92 series of chips, and what yours truly wonders the most is will the G92 support just Open GL 3 (Long Peak), or will the chip support Mount Evans as well (upgrade of Long's Peak). Of course, DirectX 10.1 is also a moot point at this time, and worst case scenario would be support for DX10.0+, but somehow we don't feel that source was right on target.

Overall, one thing is certain - G92 will deliver significant performance boost at 8800 price standpoint. µ


Share this:

blog comments powered by Disqus
Subscribe to INQ newsletters

Sign up for INQbot – a weekly roundup of the best from the INQ

INQ Poll

Heartbleed bug discovered in OpenSSL

Have you reacted to Heartbleed?