Overclock.net banner

1 - 7 of 7 Posts

·
Premium Member
Joined
·
12,992 Posts
Discussion Starter · #1 ·
Ok, I'm going to start off by saying that I don't intend to start any kind of flamewar. I'm going to state my observations and come to as unbiased a conclusion as I can to help people understand what I believe to be happening.

So nVidia is going to be unveiling their GTX 480 and GTX 470 models exactly 1 week, and we're getting all kinds of "leaked" information. The GTX 470 is fairly locked down, but the GTX 480 information varies wildly from source to source.

We have reports of it having either 480 or 512 CUDA cores and a TDP ranging from 250-300W, and we're just now starting to get clock speed information.

But which numbers are correct?

The answer is its impossible to tell. Reviewers are apparently just now getting their "finalized" cards in.

nVidia has also apparently pushed the retail launch of GF100 back a week. Why would they do that?

Well, personally, I think they are still tweaking it.

I don't think anyone will deny that having the fastest GPU is extremely important to nVidia. Since G80 the size and cost of their flagship GPU has come a distant second to the raw performance.

But if the performance numbers we're seeing are true, GF100 is surprisingly close to a gpu that is nearly half it's size. This is worrisome.

What's even worse is that the 40nm node isn't exactly friendly. Even AMD is having issues, so yields aren't that great.

So now we've established two primary variables: performance, and yields.

nVidia needs to figure out the maximum performance they can get out of the chip while still yielding enough working units to sell.

The way it works is that nVidia will order X number of dies, and when they get them back they will be tested, and ranked according to their performance, since every die they get back is a little different. This process is called binning.

Now, in order to maximise performance nVidia needs to tweak other variables, the most important being the number of active CUDA cores, the voltage being applied to the chip, and the core, shader, and memory frequencies. All of these add up to a TDP, or thermal design power. The TDP is limited by both the effectiveness of the cooler as well as an artificial cap of 300W that they need to be under for marketing reasons.

For the GTX 470 the decision was easy, as the lower binned card and not the flagship nVidia is less concerned about it. Based on all of the rumours it looks like nVidia simply disabled 4 blocks, bringing the core count down to 448 and the TDP to 225W. These numbers have been fairly constant over the past few weeks so I'm going to assume this was finalized a while ago.

The GTX 480 on the other hand is a different matter. The rumours put it's performance at anywhere from 5-20% faster than the HD 5870. 5% isn't a lot, and the early release of AMD's Catalyst 10.3 has increased performance in a number of games by roughly that amount.

This potentially puts nVidia in a situation where they don't definitively have the fastest gpu! So what do they do? They delay shipping the cards to reviewers a week and go back and start tinkering with the GTX 480's final specs.

Here's a rough pseudo mathematical forumula for what they need:

[ A*Core Clock + B*Shader Clock + C*Memory Clock + D*CUDA Cores ]/ [(Actual TDP < 300) * (Actual performance > 5870) * (Yields > Yields required to launch) ]

Those familiar with mathematics will recognize this as a system of 1 equation and 4 unknowns, with three conditions that need to be met, or the answer is undefined.

But none of that really matters, what matters is that there is no single answer, so nVidia spends a week or so re-tooling the card, and the engineers periodically leak information as they are wont to do near launch.

This is why the past week or two has been filled with all kinds of numbers. The TDP is 280W, oh wait, now its 275W, hold on, now its 295W, no scratch that, its now 250W.

Also, its going to have 512 cores, no its going to have 480 cores now, wait its going to have both 512 and 480 cores!

The only thing we can conclusively say right now is that the GTX 480 will have X Cores, run at Y MHz with 1536MB of GDDR5 on a 384bit bus running at Z MHz, its TDP will be under 300W, there will be enough stock to launch, and it should be at least 5% faster than the 5870.

Anything else you hear until the NDA drops, while it might have been true at the time, can not be confirmed.
 

·
Registered
Joined
·
1,344 Posts
great post, well said. i am one of the people that posted non-news in the news section today, and learned a valuable lesson. DON'T DO THAT! this is well said (again) i just wish it had come from a nvidia user (current) so people would take it more seriously. and there was no bash in your post, congratz
 

·
Registered
Joined
·
1,330 Posts
The numbers you threw around sounds like Fermi is going to pull alot of power. . I don't know the 5 series exacts but I know on idle they pull like less than 50 and I only have a 650 PSU with a 125watt cpu and 2 HDDs.. my 5770 sure doesn't pull anywhere near 300ish O_O
 

·
Team Red Lobbyist
Joined
·
1,600 Posts
Awesome post, and you've summed it up pretty well.

Too many variables still, and I can't remember a time where there wasn't competition like this.

I fear for performance in regards to the pricing factor. If rumors are true and a 470 is like $350ish...well I think that will be a good thing.

1 more week. Here's my call. When in 2 more weeks when they really drop for that whole 10 minutes they'll randomly be scattered around the 5 major etailers.

There will be threads, by God. There will be threads.

ATI won't touch pricing at first. They will hold off for at least 2 weeks post launch (plus the 2 weeks until they really hit - so 4 from now). They need to see what the public versions really can pull and really how available they are.

If (BIG if) the 470 can beat a 5870 (with the 480 beating it even more) - then the price war will commence if they do in fact release it at $350 USD. I want this outcome myself. I'd definitely consider the 470 at $350 if it can beat out a 5870.

If not, we're all screwed...and will continue to pay high prices for the next 18 months until another generation is generally available (with recent history dictating that this will likely be ATI).

And the original adopters who got their 5850's for $259 will have officially and finally have one and beaten the game!! Bastards!

I may just cry myself to sleep if I have to pay over $300 for a 5850 if what occurs in 2 weeks doesn't go our way.
 

·
Banned
Joined
·
1,545 Posts
Good stuff nathris (it was unbiased) I'd have to agree with 470 being ready for quite some time and the 480 struggling to really show how far it can go with what TDP as a final product but i'm also not sure an extra week could fix something like this...
 

·
Premium Member
Joined
·
12,992 Posts
Discussion Starter · #7 ·
Quote:


Originally Posted by Markisa
View Post

The numbers you threw around sounds like Fermi is going to pull alot of power. . I don't know the 5 series exacts but I know on idle they pull like less than 50 and I only have a 650 PSU with a 125watt cpu and 2 HDDs.. my 5770 sure doesn't pull anywhere near 300ish O_O

Fermi's high power consumption stems from poorly designed transistors. The 40nm node has a few quirks, so you need to build in a little extra redundancy or the transistors aren't going to function correctly.

AMD learned this with the 4770. Remember how they were in stock, then they disappeared for a few months? That was AMD learning from it's mistakes. The lessons from the 4770 were transferred over to Cypress and as a result its now March and AMD has DX11 and nVidia doesn't.

nVidia needs to apply more power to the malfunctioning transistors to get them to work right. As a result the TDP is artificially inflated, and efficiency is very low.

To fix the problem they need to move onto another stepping. That takes a fair amount of time, and Fermi is already late.

If they do everything right and fix it expect to see the yields improve and TDP drop significantly with the B1 revision, but don't expect to see that for another few months.

My best guess is that the A3 GF100s are going to be available in early April, then disappear entirely, similar to how Cypress disappeared in October.
 
1 - 7 of 7 Posts
Top