• 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.

R600 die-reduction

Associate
Joined
24 Feb 2004
Posts
1,083
Location
Leeds/Cyprus
When is the die-reduction of the X2900s to 65nm expected to come out? I remember a few months ago some sites put it as expected in September, and theinq among others were speculating that they'll be the most overclockable graphics chips since the 9700Pros, but AMD stuff has had a habit of missing their release dates recently... Anyone read anything on the news about it recently?
 
There isn't going to be a 65nm 2900, the next card for ATi will be on a 55nm process, and its due around January.

The 65nm 2900 was just a rumour ATi started to try and find out who was leaking their information out. :)
 
Last edited:
I find it hard to believe that ati wouldn't refresh the r600 series. Theres been no real confirmation or denial (just the almighty INQ) so it spossible a 65nm version could be coming in time for christmas.
 
Well there's not a whole lot of reason to at the moment. Granted the core could run a bit cooler with the lower die size but it's really not as bad as people make out. They're probably concentrating on getting a high end 55nm out as soon as possible. They might however come out with a budget range 55 or 65nm 2600XT or some such like nVidia do to test the production and real world performance and of course to get the "We did it first" award.
 
I thought it was ATi themselves who said there wasn't going to be a 65nm 2900 in Sept, and that it was only started to find the leak, they are moving straight to the 55nm. :confused:
 
65nm 2900 AFAIK, was only annouced internally to see who would leak the information after the R600 leak fiasco's.
 
Darg said:
Well there's not a whole lot of reason to at the moment. Granted the core could run a bit cooler with the lower die size but it's really not as bad as people make out. They're probably concentrating on getting a high end 55nm out as soon as possible. They might however come out with a budget range 55 or 65nm 2600XT or some such like nVidia do to test the production and real world performance and of course to get the "We did it first" award.
The 2400 and 2600 series are already 65nm afaik but I agree that producing another version of the 2900 with a smaller die size wouldn't be of great benefit.
 
El Jimben said:
The 2400 and 2600 series are already 65nm afaik but I agree that producing another version of the 2900 with a smaller die size wouldn't be of great benefit.

Ah, I stand corrected :)
 
the 90-80nm shifts, bought about by tsmc/tcsm/tsmc or whatever that damn plant is called, was pretty sweet in that it was so close to 90nm that it required very little work to reduce the size. but 90 to 65nm is a pretty large amount of work. it can literally mean, worst case scenario, realligning almost every ickle transistor, not likely to be that bad, but theres a huge amount of work. considering, as with basically every single company thats big at the moment, they will have multiple teams working on multiple projects. they'll have had a R700 team going for a long time. you can either lets be honest, waste time on getting a little more speed out of the 2900xt, or get going on the next card. remember, even the next gen will have been mostly done without amd's involvement i would think, these cores are worked on for a couple years.

now i'm not saying the 2900xt isn't a good card, but so far, the 2900/8800's aren't exactly looking fantastic with dx10. maybe its just drivers but, the games i've tried aren't exactly flying along with any card i have. the sooner the better for both truely new gens rather than refreshes really.

the only possibly big problem is. ok ati may have jumped the gun on the software AA, a lot of the big wigs in the gaming industry are saying IQ accurate, high IQ AA properly done with HDR now being used massively really WILL require software AA. so nvidia might be going 1teraflop with the next card, but if the game designers are asking them for software AA aswell, how much of the extra power might be used instead of hardware AA?
 
Yup ditto pointless spending any extra effort in 2900 series, they will just fall back revamp and redo a whole new series tweeked/refined but with a smaller/better chipset. I dont see them coming in December though with it, Spring-Summer 08 proberly.
 
drunkenmaster said:
the only possibly big problem is. ok ati may have jumped the gun on the software AA, a lot of the big wigs in the gaming industry are saying IQ accurate, high IQ AA properly done with HDR now being used massively really WILL require software AA. so nvidia might be going 1teraflop with the next card, but if the game designers are asking them for software AA aswell, how much of the extra power might be used instead of hardware AA?
That's not the only thing, they also (apparently) spent a lot of effort implementing pageable VRAM as well, and that feature ended up being cut from DX10. They're a forward-looking company, and as usually happens they end up shooting themselves in the foot by spending time and money researching features the market isn't ready for, while their competitors are simply refining their current tricks, ekeing ahead on performance, publicising their success better and eating up market share. Same thing with Intel sticking 2 dual-core chips on a single die and stealing AMD's thunder while they were busy researching a natively-quad core chip.
 
Both companies are forward looking, i seem to recall shader model 3 being implemented WAY ahead of Ati by Nvidia.

You pick your cards, take your chances. Basically, Nvidia got the drop on ATi because of the AA performance. At present ATi gets killed with AA on by the 8800 series because Nvidia either couldn't get it working or chose not to due to costs/timescales :( Ati should have done better but looks like they got stuck in actually trying to code cards for mainly for DX10, instead of for a DX9/DX10 cross-over.

Would be nice to have a refresh, but by a small team. The majority of it work/research should be going into it's next gen cards right now so they don't get wiped out when Nvidia releases it's next line of cards, and ATi/AMD can become competitive again.

All I can say is.. I hope there refresh (if it happens) some how includes MORE ROP's.

Matthew
 
manveruppd said:
That's not the only thing, they also (apparently) spent a lot of effort implementing pageable VRAM as well, and that feature ended up being cut from DX10. They're a forward-looking company, and as usually happens they end up shooting themselves in the foot by spending time and money researching features the market isn't ready for, while their competitors are simply refining their current tricks, ekeing ahead on performance, publicising their success better and eating up market share. Same thing with Intel sticking 2 dual-core chips on a single die and stealing AMD's thunder while they were busy researching a natively-quad core chip.


well lets be honest, i've had a gts, got a gtx, which i'm selling, was selling ages ago and just not got round to actually doing it. on vista i still had/have problems with the 8800's. the vram was dropped, basically because nvidia simply could not get their cards to work with it or more to the point, simply could not write drivers that worked. they still can't get their vista drivers completely right.

MS really did cave very quickly to agree to cut something that nvidia simply failed to do right. the ways of doing AA, its entirely a choice and maybe, maybe ati did jump the gun. maybe we'll see higher IQ in ut3/crysis with fine performance, who knows, though i personally think, right thing slightly to early as ATi very often seem to be. but the vram was something nvidia tried to do, failed miserably and cried till MS cut the it from dx10 specs.

giving into companies that are holding us up is not helping anything.

It was always a doubling pipes on new gen setup up until nvidia went 6800 to small increase on the 7800. in the last 2 gens, 7800-8800 we haven't seen doubling of performance as before, or at least relatively close. we all seem to be fine with smaller performance boosts, higher new prices. most definately nvidia are the ones with the freaking stupid prices and trying to push them higher every gen.

the main problem is, much as cpu's are now literally able to double performance(as good as) with extra ores, gpu's really can just pump everything up and give us more performance but we're not being given the same increases as before, even though, its fairly easy to do.
drop size, increase pipes, increase shaders in the right ratio, bump up mem speed as new mem's available. but when software/design specs are dropped because someone can't get them to work, we're just slowing the push forwards.
 
drunkenmaster said:
7800-8800 we haven't seen doubling of performance as before, or at least relatively close.

What? The 8800GTX wasn't a small boost, killed the 7800GTX.

13538.png

fearep_8x.gif

image008.gif

coh1600.gif
 
Last edited:
fornowagain said:
What? The 8800GTX wasn't a small boost, killed the 7800GTX.
Yea, infact from a few benchmarks I've seen the 8800GTX is around double the performance of a 7900GTX.
 
No wonder hes called drunkenmaster, as he must have been ****** to have said that. :D :D

I went from an x1800 XT 512mb to a 320mb GTS, and even that was double or more performance. :p
 
I already knew it but this pretty much confirms there will be NO 2900 (R600) refresh. Straight to R700 this time.

AMD Roadmap from their analyst day yesterday.

5488_large_performanceplatforms.jpg
 
Last edited:
I have to say ATI line up is pretty ****. Theres no card covering the £100-200 pound area, plus what happened to the xl,gt,pro????
Same for nvidia but not as bad, i wish they made a 8800gs or gt to cover up the performance gap between the 8600gts and 8800gts.
 
Back
Top Bottom