Nvidia Pascal Announcement

Historically, the ultimate Geforce chip has always sported the full count of CUDA cores available: 8800GTX, 285, 580 ,680, 780Ti, 980, 980Ti .. etc, the only exception was the 480, and that was the exception due to heat and power consumption and then quickly got rectified with the 580. The 780Ti even scored more cores than the original Titan Kepler (not counting Titan Black). So I can't see a solid reason why NV would go out of it's way to announce a half assed cut down chip as the ultimate Geforce chip, when everyone in the enthusiast community knows there is a more complete "ultimate" chip. Unless they become morons of course!


Yes, but we are in an era of internet words of marketing and buzz.. we cant fight against this .... 28th is the " full event" about Ryzen and presentation of Vega .. Nvidia set his launch date of the 1080Ti the 28th.. today AMD have decide to lift some embargo about Ryzen and open Preorders... Nvidia set his countdown for the 1080ti in the same raw.. everything is marketing, public visibility ... dont try understand and read between the lines of thoses marketing plot .. Sadly peoples eat it today, and forget it tomorrow. As for News and informations, they are satured by this type of message... how do you want they analyze it and understand it ? ..
 
Last edited:
Is a "Ti" the same as an "Ultra"?
Ti has been synonymous with that very meaning since maybe Geforce 3. Top Geforces had these two letters before they got replaced by Ultra starting from the FX line, down to the 6 series.
everything is marketing ... dont try understand and read between the lines .. most consumers dont.
I suppose you are right. It's just mighty disappointing if the full GP102 is never released to consumers in the form of 1080Ti, if it was never released in a Titan form or Ti form, then chances are it will never be released, ever!
 
Ti has been synonymous with that very meaning since maybe GeForce 3. Top GeForces had these two letters before they got replaced by Ultra starting from the FX line, down to the 6 series.
Not really. First Ti was a die shrink, but Geforce3 Ti was two SKUs above and below original and recently Ti was used to mark interesting price/performance point or big chip. Ultra was used for high clock parts up to GeForce 8. So no rules for Ti.
 
Ti has been synonymous with that very meaning since maybe Geforce 3. Top Geforces had these two letters before they got replaced by Ultra starting from the FX line, down to the 6 series.

I suppose you are right. It's just mighty disappointing if the full GP102 is never released to consumers in the form of 1080Ti, if it was never released in a Titan form or Ti form, then chances are it will never be released, ever!
[pedant]
The Ultra pre dated the Ti, in the form of the TNT2 Ultra and GeForce 2 Ultra.
[/pedant]
 
So the 1080Ti is a 3584 core GPU (same as Titan Pascal), with faster clocks for both the cores and memory. This should make it faster than Titan, still I am massively disappointed it is not the full GP102 though. VRAM size is 11GB?! Price is 700$.

EDIT: Price is 700$, the normal 1080 also got a price cut, and is now 500$.
16998725_421954131480546_5337972077033491090_n.jpg


Also NV is launching upgraded versions of GTX 1080 and GTX 1060, with faster VRAM, these will be sold as factory overclocked cards through partners.

17021630_421954584813834_761844593742746687_n.jpg
 
Last edited:
Talk about a complicated product stack. Guess the people that bought Titan Pascal got shafted badly.

Do the new 1080's only get released with the better RAM? Or is it a whole new 1080+ or something?
 
The 1080 Ti isn't exactly earthshattering (we all saw the 11-12Gbps stuff show up on Micron's site).

But the thing that surprised me was the 9Gbps GDDR5. What's going on there? Who's making that?

1060-overclock.jpg


I thought GDDR5 was only supposed to get up to 8Gbps. Is this really just an overclock via aggressive binning?

Another thought, why doesn't the 1070 have that option? I'm sure some aggressively overclocked 1070s could utilize the bandwidth. And there's room for that due to the large gulf between the 1070 and 1080 pricing.
 
Nvidia told us that the 1080 Ti uses ‘next-gen’ GDDR5X VRAM from Micron that operates over a 352-bit bus – slightly less than the 384-bit bus found on the Titan X. However, the next-gen GDDR5X memory is faster than the Titan X’s VRAM, thanks to its 11Gbps (effective) transfer rate. It is this faster GDDR5X – plus the chip’s overclocking headroom – which, Nvidia says, makes the 1080 Ti faster than a Titan X.

In terms of VRAM capacity, the 1080 Ti features 11GB of memory. It may be a strange number, but it makes sense as the 1080 Ti uses the same GP102 chip as the Titan X but with a single memory controller disabled. Nvidia were keen to emphasise that more and more games are now using more and more VRAM, though, so the added capacity over the GTX 1080 should offer tangible benefits.
http://www.kitguru.net/components/graphic-cards/dominic-moass/nvidia-announces-gtx-1080-ti/
 
Last but not least, Nvidia also announced new 'OC' SKUs based upon the GTX 1080 and GTX 1060 designs. Both of these new 'OC' cards will be pepped up with the same faster GDDR5X from Micron. With the GTX 1080 price cut to $499 the GTX 1080 11Gbps OC will cost somewhere between it and the new GTX 1080 Ti. Similarly we will be seeing an Nvidia GTX 9Gbps 1060 OC with the faster GDDR5X memory installed - which should help it combat the AMD Radeon RX 480 more effectively. Pricing and availability for these OC cards is yet to be announced.
http://hexus.net/tech/news/graphics/103000-nvidia-geforce-gtx-1080-ti-worlds-fastest-gaming-gpu/
 
Back
Top