WHEN YOU DON’T have any product, you spin and spin and spin. Nvidia is doing just that with the upcoming paper ‘launch’ of the GTX580.
Yes, you heard that right, the GF110 is officially being touted to partners as the GTX580. While there are a lot of pretty outrageous rumors going around, this mystery part looks to be a mildly updated, basically just fixed, GF100.
The plan is simple, in November when DAAMIT launches Cayman, the top single chip Northern Islands family, Nvidia will ‘launch’ the GTX580 as a ‘spoiler’. Tame sites will tout the wonderfullness of this paper part, and a few loyal Nvidia ‘partner sites’ may even get a card.
If you are a buyer, you will be waiting until 2011. GF110/GTX580 is not expected to hit the shelves until the very end of January, likely early February. Given that Nvidia has yet to release a single GF1xx part on time, and hasn’t released a single desktop Fermi chip that is fully working, don’t hold your breath for a timely launch.
As for the rumors of a huge chip with massive shader counts, or more laughably a much more svelte GF100b, lets do a little math. The GF100 is 529mm^2 and has 512 shaders, that works out to 1.033mm^2 per shader. GF104 is 367mm^2 and has 384 shaders, .956mm^2 each, and GF106 is 240mm^2 for 192 shaders, 1.25mm^2 each, worse than the GF100 for areal efficiency.
GF104 is slightly better than GF100 in area per shader, GF106 substantially worse. Both are vastly better than GF100 in performance per watt however, but both trail DAAMIT equivalents by a lot. The ‘new’ GF110 will be built on the same process, TSMC’s 40nm, as well.
What is the point? Easy, the GF104/6/8 line is not substantially more efficient than their big brother, just a little bit more efficient in some specific parts. If the GF100 uses the same ‘new’ architecture with all the ‘updates’ that the GF104 has, it will more likely than not get bigger, but it will use less power. At idle.
It will not however get smaller, and will not be a 10 billion shader part that blows ATI back to the stone age. If Nvidia can squeeze 512 shaders that actually work out of a piece of silicon, and also up the clock by 10%, you are looking at about a 17% increase in speed over a GTX480.
That simply is not enough. When you have nothing, spin. By now, there is a pile of yarn behind Nvidia HQ big enough to be seen from orbit.S|A
Latest posts by Charlie Demerjian (see all)
- Globalfoundries 7nm process isn’t even close to the name - Sep 26, 2016
- ARM upgrades realtime offerings to v8-R and adds Cortex-R52 - Sep 21, 2016
- Everspin and Globalfoundries team up for embedded ST-MRAM - Sep 15, 2016
- Intel’s Xpoint is pretty much broken - Sep 12, 2016
- ARM adds 2048-bit vectors to v8A with SVE - Sep 7, 2016