• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

GT200 GTX280/260 "official specs"

... doesn't that mean the 65nm first released cards will have a heat problem?

You mean the first 260/280 GTX? Yeah that's the whole point in wanting off 65nm cause they are too hot.

Although I think am reading you wrong, I re-read your reply a few times and didnt quite understand it lol. Could you reiterate it please?

Martyn
 
You mean the first 260/280 GTX? Yeah that's the whole point in wanting off 65nm cause they are too hot.

Although I think am reading you wrong, I re-read your reply a few times and didnt quite understand it lol. Could you reiterate it please?

Martyn

I was sorter trying to say this:

If it's just heat problems and the die shrink solves it and released within three months do you want one of the early "hot" cards?

If the heat isn't too much of a problem but enough to hold back clock speeds on the 65nm cards to what Nvidia want them to be then they are aiming for 55nm cards to reduce heat and hence up stock clocks. Same question, do you want one of the early "slower" cards?

Hope that makes more sense ;)
 
It's like buying a brand new car instead of a few yrs old one with 1 previous owner, it does not make sense! why spend so much and end up losing so much when the new cards are out every year?

I don't really agree with that analogy, as with gfx cards you aren't buying like for like.

When the 8800GTX was released, yes it carried a price premium, but it was also head and shoulders above every other GPU on the market. It may not have represented bang for buck, but equally if you wanted that level of performance, it was your only option - you can't timetravel forward a year and buy something just as fast for less money.

There is an argument to suggest that buying one uber card every 2 years is just effective as buying a midrange card every year - it costs a little more but during the first year you should get more gaming pleasure out of it than the cheaper alternative.
 
it was never about bang for buck. it was the sheer fact that it was twice as fast as the next competing gpu and then some. some people believed that justified the price:)
 
it was never about bang for buck. it was the sheer fact that it was twice as fast as the next competing gpu and then some. some people believed that justified the price:)

i would say it will be twice as fast like 8800gtx was not twice faster then 7950gx2 or 1950xtx?
 
I was sorter trying to say this:

If it's just heat problems and the die shrink solves it and released within three months do you want one of the early "hot" cards?

If the heat isn't too much of a problem but enough to hold back clock speeds on the 65nm cards to what Nvidia want them to be then they are aiming for 55nm cards to reduce heat and hence up stock clocks. Same question, do you want one of the early "slower" cards?

Hope that makes more sense ;)

Ahhh I see. Again though there is nothing to say the smaller die would mean a quicker card. And no I wouldnt want one of the hot cards, but really only under the basis I wouldnt pay £450 for a graphics card, my gf would not only kill me, she would subject me to a weekends shopping where she proves that she can spend more than I can........ :P

I think the clocks of the new card, are more of a compromise tbh. Rather than cripling the 260 further and potentially loosing out to ATi (which may still happen), they pushed the 280 further on and found 602 core clock to be the limit. I cant seem to find the other clocks atm but you get the idea. The fact it is 602 suggests to me they where trying desperetly to push it on as much as possible and every little counts, but ultimatly lost the battle to heat.

My point is that the card was more likely originally going to be 55nm and shoved back onto 65 since they cant get it right, as such getting it onto 55nm will put them where they wanted to be originally, so far as heat goes. It could be the case 65nm performs (as in fps) the way they wanted to, it just runs too hot, but rather than killing its performance, they would rather it melted iron :). This would explain why they are trying desperatly to get it onto 55nm, not so much because the card isnt going to perform, its just heat. And that is why I dont see a performance (in fps) being likely when they move onto 55nm, unless it has some other features (more sp etc).

One last thing, I think the engineers where faced with one of the following choices:

1. 55nm part not performing as top brass want it.

2. 65nm part performing as top brass want's it (just runs hot).

I think we can all see what one is easier to explain to your boss :)

Martyn
 
i would say it will be twice as fast like 8800gtx was not twice faster then 7950gx2 or 1950xtx?


Whilst it wasn't twice as fast, at release the 8800GTX utterly destroyed all previous cards, and at higher image quality too. The higher the res or quality, the better it was.

It is very easy to look back through cynic tinted spectacles, but when the 8800 series was released, it was a revelation similar to the 9700, or even more.
 
i would say it will be twice as fast like 8800gtx was not twice faster then 7950gx2 or 1950xtx?

it was twice as fast as the 1950xtx. it was faster than 1950xtx crossfire in some situations. obvously,it wasnt twice as fast as a 7950gx2. but for simplicity sake, i was refering to single gpu cards.
 
I think the clocks of the new card, are more of a compromise tbh. Rather than cripling the 260 further and potentially loosing out to ATi (which may still happen), they pushed the 280 further on and found 602 core clock to be the limit. I cant seem to find the other clocks atm but you get the idea. The fact it is 602 suggests to me they where trying desperetly to push it on as much as possible and every little counts, but ultimatly lost the battle to heat.
Its more likely the 602 comes from the 27Mhz clock generator. 575>602. You see clock steps based on 1/3 and 2/3's multiples of 27Mhz but not all.
 
well if 280 will destroy 8800gtx like 8800gtx destroyed previous cards and lasted a year i would say 450£ to spend its okey as u know the card that u have will last you long while but if it will only last u 2/3 months and even something else is released faster and cost 400+£ its a waste
 
tbh tho it should destroy the 8800gtx's as they are what 2 years old now!

Yeah but I don't think they will. Yes they will be quicker but quick enough to be worth £450-500? No. Unless you have the money spare which plenty of people do. If you have a GTX and a SLI mobo you might as pick up another GTX for £90 and chances are you will have a system as quick as a GTX280.

Personally I think ATI will sell more 4870 cards especially if they do end up been 30% quicker than the 9800GTX and sell for £200.

4870 are looking like the bang for buck cards and GTX280 if you want the quickest, albeit at a cost.

With potentially two 4870 cards been quicker than the GTX280 and cheaper, I don't think I will be buying one unless the reviews once released prove me wrong. ;)
 
tbh that's what most of us high end GPU users do on here, as soon as I found out about the GTX280 and confirmed when it was coming out I put my 8800gtx up for sale, I now have a 6200 in my rig :o, I got a £130 for it, take that away from the £479 I paid for the 8800gtx = £349, £349 for one of the best gpu's that's ever been released and one that held the crown for the longest, certainly money well spent for any GPU enthusiast.

I didn't quite get this "It's like buying a brand new car instead of a few yrs old one with 1 previous owner", did you mean that it was better to wait and buy a secoundhand one ?

To be honest the GT wasn't exactly "more powerful" than the 8800GTX.
 
high end cards do produce better fps at high resolutions, but as res goes up performance drops.

so the point of having a high end card to run at high resolutions is kinda mute as it would be too slow to play anyway
 
Yeah but I don't think they will. Yes they will be quicker but quick enough to be worth £450-500? No. Unless you have the money spare which plenty of people do. If you have a GTX and a SLI mobo you might as pick up another GTX for £90 and chances are you will have a system as quick as a GTX280.

)


I very much doubt that
 
I very much doubt that

Reasons? I thought the GTX280 was only going to be 30%-50% quicker than a 9800GTX and since, at high res, the 8800GTX is still as quick as the 9800GTX and then you get up to 50% gain in SLI, I thought two 8800GTX in SLI would be around GTX280 performance?

Yes/No * please delete as appropriate :D
 
Yeah but I don't think they will. Yes they will be quicker but quick enough to be worth £450-500? No. Unless you have the money spare which plenty of people do. If you have a GTX and a SLI mobo you might as pick up another GTX for £90 and chances are you will have a system as quick as a GTX280.

whereabouts can you get a GTX for £90? I've been looking around but they're not that common anymore... and I'm not up for getting one off EBay.
 
whereabouts can you get a GTX for £90? I've been looking around but they're not that common anymore... and I'm not up for getting one off EBay.

Well although I agree with you about Ebay although I got lucky when I bought my 8800GTS G92 for £110 about 3 weeks after release when they were £170, I understand your weariness at buying a GTX on the fleabay but the point is, you can get them cheaper although I was wrong about the £90, £110 seems to be the current cheapest prices.

Okay, if you stick to MM it will cost you £135 for one including postage but my point is still valid and they will be £90 once GTX280 cards come out.

There are probably going to be a lot of people upgrading their GTX cards to the new ones so the market be flooded.

Plus the 8800GTX will finally stop been king of the hill for high res/lots of aa so will drop in value.
 
Last edited:
Back
Top Bottom