NVIDIA GF100 & Friends speculation

possibly its a marketing directive, depending on which sites would have more impact on sales. I wouldn't be surprised, most companies actually have statistical data on best exposure.
 
It's safe to say that if they had an awesome product they wouldn't be cherry picking reviewers. That or volume is such a joke they can't even supply all the review sites.
 
It's safe to say that if they had an awesome product they wouldn't be cherry picking reviewers. That or volume is such a joke they can't even supply all the review sites.
Exactly, with an awesome product they would rings the bells to anyone who wants to hear, and even to those who dont :LOL: I really hoped Charlie was wrong about paper launch...

Seriously, how company can have only few boards of a refresh GPU? In Fermi or R600 case it was somewhat understandable, new generation of a huge chip, *** happens. Yet AMD had a fixed refresh half a year later, in massive quantities. Nvidia cant do that even after a year? If NV would be about to launch Kepler, I would understand, but its very surprising in GF100b case.
 
lol@newsflash :D

Not being able to reach significant volume is one thing, not being able to send 5~10 cards per country to the press is quite another!

But also no briefing or NDA either? Looks like Nvidia is cherry-picking sites to get good reviews, as tight supply would be no reason not to brief sites.

I'm not sure, but I don't think NVIDIA briefs websites that don't get samples.
 
Video from PDX LAN: http://www.youtube.com/watch?v=oCPMHIDpT88

1) GTX580 is having a vapor chamber cooler
2) multiple displacement mapping and morphing is very interesting
3) Call of Duty: Black Ops

And, all the recent slides are real. :)

Nice video. So it looks GTX580 is a litte faster than a GTX480 with a couple of new uses for tesselation... I wonder if that thing with the laser or the changes in the body use any sort of Cuda, or are possible only with DX11. Wonder if they would work on ATi hardware...
 
I wonder if that thing with the laser or the changes in the body use any sort of Cuda, or are possible only with DX11.
It shouldn't require anything special, that laser probably just draws onto the tesselation map, thus deforming the model...

Vapor chamber cooler looks radically different to the leaked images we've seen so far. Probably just prototype hardware we've seen so far, with the vapor chamber being production version. I hope that NV dude wasn't exaggerating when bragging about how much quieter the thing is, that'd be a huge win for them if true. I'm very excited! :D

The endless city demo looked much better in motion than static shots, but still doesn't really sell those half-million lights etc. It could just as well be "only" 5000 lights and I wouldn't know the difference. ;) Still, the raw specs is very impressive, the ability to process that many lights in realtime just shows how far we've come in a short period of time. I saw some Silicon Graphics Onyx workstations with multiple raster managers and geometry engine boards installed back in '99 or '00, basically any budget GPU today could beat those machines, and that was around 1.5-2 square meters of PCB space's worth of chips... Then compared to this monster GPU... Dayum.

Like I said, I'm very excited. The chance this product might be extremely low in availability is somewhat worrysome, but that's still just forum speculation right now. I'm hoping there will be decent availability and that price won't be astronomical.

That presentation was kinda lame with people cheering about a damn videocard (can you say "fanservice"?), but...yeah. That super secret video card that can't be mentioned does look Real Good I must say! Did I mention I'm very excited? :)
 
That presentation was kinda lame with people cheering about a damn videocard (can you say "fanservice"?), but...yeah. That super secret video card that can't be mentioned does look Real Good I must say! Did I mention I'm very excited? :)
Hopefully, this time it's almost not compared to its predecessor at all... not like with Tesla 2 and the intriguing efficiency drop with respect to Tesla 1 (DP wasn't nearly as efficient - ~75% iirc - but SP wasn't either, so that's not only related to increased DP throughput). Strangely enough, GF100 is way faster at what a GPU isn't fast, but slower in almost all non-pathological cases.

If the slides seen so far are real, then this "GTX580" is nothing more than a complete GF100 at slightly higher clocks, not really exciting to say the least. With such specs/performance, I'd really be surprised if AMD released a slower Cayman, although I've already been surprised in the past by R600 and Barts (which, IMO, is too much of an economical move).
 
>10 in stock, too.


It's marked as "usually ships in 10 days" so they obviously don't have stock and they've given themselves a get out with the use of "usually". If they don't have stock, they don't know what numbers they will be able to get.

It not unusual for the sort of place that breaks the launch date in order to get pre-order sales to indicate they have stock that they don't really have.
 
Back
Top