Didn't ATi have another Shader Day yesterday? Did anyone hear anything from there? (My apologies if this has already been covered, but I haven't seen it.)
ATi's Shader Day yesterday, any info?
- Thread starter digitalwanderer
- Start date
Joe DeFuria
Legend
Evildeus said:The only thing i heard from someone who attended is:
The battle between the NV40 and R420 will be harsh. Hope it helps
That's what I like to hear.
Me consumer = me win.
I have no clue if it's an NDA event or not, and my EB e-dress is down so I can't find out for a while either.
Thanks for the bit-o-info. Unrelated to the Shader Day event I have heard the same thing from a few ATi sources, they seem hungry to get the direct comparisons going...which I take as a good sign.
Let 'em try and top each other for a while! Joe is right, we'll come out the winners for it. 8)
Thanks for the bit-o-info. Unrelated to the Shader Day event I have heard the same thing from a few ATi sources, they seem hungry to get the direct comparisons going...which I take as a good sign.
Let 'em try and top each other for a while! Joe is right, we'll come out the winners for it. 8)
Joe DeFuria said:Evildeus said:The only thing i heard from someone who attended is:
The battle between the NV40 and R420 will be harsh. Hope it helps
That's what I like to hear.
Me consumer = me win.
I agree. Though i hope that the R420 is a bit faster if it only supports SM2.0. I think that'll keep the prices a bit lower then if Nvidia was both as fast and also supported SM 3.0.
jimmyjames123
Regular
When you say "harsh", what exactly do you mean? Do you mean "heated" and "hotly contested"?
Probably with the loser (whichever one it is, I don't know) resorting to NV3x-style methods to improve performance.jimmyjames123 said:
jimmyjames123
Regular
You mean hacking the drivers? Do you think they would be gutsy enough to try that after the NV3x fiasco? Wow, wouldn't that be something, hehe
This time around, I'm not so sure there will be a clear cut loser or winner. It looks like both cards may each have unique attributes that make them attractive to different consumers.
Do you think they would be gutsy enough to try that after the NV3x fiasco? Wow, wouldn't that be something, heheThis time around, I'm not so sure there will be a clear cut loser or winner. It looks like both cards may each have unique attributes that make them attractive to different consumers.
Without a doubt. Not like it's really hurt NVIDIA's reputation in the long run, it seems; ATI would survive, too.jimmyjames123 said:You mean hacking the drivers?
jimmyjames123
Regular
If NVDA in particular resorted to hacked drivers again, then IMHO that would be a very unwise business decision. This is very short-sighted business practice, and what they may gain in relative performance is outweighted by a loss of credibility! Even if they can gain a few fps here and there, many of the consumers, reviewers and developers will start to see through it. Even fanatics would be banging their heads against the wall to see NVDA do it again
PVR_Extremist
Newcomer
Well I won't benefit as a consumer until these boards are £150 Uk notes or less
Just bought a Redeon 9800 Pro so I won't see another one in my machine unless I build a new PC or 18 months pass......or Videologic send me a Series 5...... or pigs fly etc etc
Just bought a Redeon 9800 Pro so I won't see another one in my machine unless I build a new PC or 18 months pass......or Videologic send me a Series 5......
or pigs fly etc etcI don't think NVidia will need to resort to hacked drivers for awhile. The current drivers do not fully take advantage of the dual issue+co issue capability of the pipeline. I think you will see that future driver compilers will deliver atleast a 25-33% speedboost in per-clock shader performance, and that this will be interpreted as an unbelievable driver hack at first, but will eventually be proven to be true.
The difference between now and 1 year ago is that the NV3x was broken in a very serious place: shader performance, and the NV40 isn't, or atleast, isn't in the areas that matter. The NV3x was also penalized for a more expensive AF, which has been fixed in the NV40, so no need for filtering hacks either.
The difference between now and 1 year ago is that the NV3x was broken in a very serious place: shader performance, and the NV40 isn't, or atleast, isn't in the areas that matter. The NV3x was also penalized for a more expensive AF, which has been fixed in the NV40, so no need for filtering hacks either.
Hellbinder
Banned
I suppose.. in a *performance Today* or *Features tomorrow* kind of way.Evildeus said:Hotly contestedjimmyjames123 said:
jimmyjames123
Regular
Nice to see you Hypebinder
I would say it would be most rational to say that the "performance today" will be dependent on what games are tested, what AA/AF levels are tested, and what resolutions are tested. For instance, the R420 may be fastest in Half Life 2, and NV40 may be fastest in Doom 3, etc.
"Features tommorow" sounds like a PS 3.0 thing. I assume that you are implying that the R420 will not initially have full support for PS 3.0, while the NV40 will. I'd say that "features tommorow" is generally a good thing, as long as the hardware can perform ok while using these features (and the NV40 seems quite capable of handling PS 2.0 and PS 3.0 effects). It looks like several developers are including PS 3.0 effects in their new games. Whether visual quality changes is debatable, but apparently the PS 3.0 effects are supposed to be somewhat more efficient than PS 2.0 effects.
By the way, I somehow happened to stumble upon your comment at Rage3d about how "Jen Hsun Huang is going to have a coronary and keel over lifeless after he sees the R420". That is a pretty low blow, even for you HBinder. Quite a disgusting and evil thing to say really. Obviously you have some troubling personal issues if you need to resort to saying something as sick as that.
