The AF hit can be more significant on R600 at this time. It should show higher quality in exchange (not on the LOD isotropy front, but on the mipmap transition front). I believe some of that will be addressed, in part, in future drivers. I can't promise a % improvement at this time.
Sir Eric Demers on AMD R600
- Thread starter B3D News
- Start date
- Status
The AF hit can be more significant on R600 at this time. It should show higher quality in exchange (not on the LOD isotropy front, but on the mipmap transition front). I believe some of that will be addressed, in part, in future drivers. I can't promise a % improvement at this time.
So, can we conclude, that resolving AA samples in INT type (non-HDR type) of surfaces is actually under-utilizing overall throughput capacity of moving data, inside the chip?
I see that as a way to say TMUs were designed to shine at filtering FP64 textures as filtering rate for them is the same as it is for RGBA8 textures
leoneazzurro
Regular
I saw that in tech specs for HD2900 there is listed "bicubic fitlering" but this seems not yet being reviewed anywhere, could you please give some detail on this feature and if it will be really implemented in future drivers?
Hopefully by the time the R650 (or whatever the revision is) gets released, the drivers will have caught up. About the time I'll be in need of an upgrade. But Ati REALLY needs to make the cards quieter. I'm sick of the loudness of my x1900xtx.
Sireric,
You have referenced being "late" as reason why perhaps some of the forward thinking tech in the R5xx series wasn't exploited by games but would it be just as important to recoginze nvidia well recognized developer relations through their TWIMTBP vs. Ati's non existant GITG program? Being late is only part of the issue, I believe.
Sireric,
You have referenced being "late" as reason why perhaps some of the forward thinking tech in the R5xx series wasn't exploited by games but would it be just as important to recoginze nvidia well recognized developer relations through their TWIMTBP vs. Ati's non existant GITG program? Being late is only part of the issue, I believe.
Last edited by a moderator:
Unknown Soldier
Veteran
No we didn't, but it also doesn't explain why AMD delayed the chip. The explanation that they wanted to release the family is bull. I mean, are there RV630/610 reviews yet(and I mean actual reviews)?
The delay has cost AMD much.
US
I think there is an irony in the fact most of the R600 problems are being attributed to the inability to get driver quality up there in time, while a few months back ATI representatives bombastically chided Nvidia for the lack of planning and preperation when it came to dirvers.
I recall the same thing. Remember ATI and NVIDIA are treated differently for some reason when it comes to the public. Odd, but true.
My only question is when is the power output going to be even looked at? It's getting completely out of hand that the latest cards are requiring so much power to run them. We have 1000 and even 2000 watt power supplies now. It's unnecessary.
When is ATI ever going to look at reducing power? Why can't high-end chips be looked at the same way as mobility chips? There is no reason for the high watt output of today's electronics.
It`s not there, it was simply a forum post trying to stir stuff up. I may be wrong,even though I`m not certain bicubic filtering would be that useful for anything.
Good question. A13 was fabricated in January, but released in May. Maybe another revision was on the way, but didn't came right?
/edit: Morgoth the Dark Enemy: and what's this?
Bicubic is supported, and was intended for some OS font features (I think). It's not exposed at this time. I don't believe that DX allows it to be exposed. It does not operate at the same rate as bilinear. Might show up in OGL.
In a way, yes. But it's not simple -- Some busses are under utilized, but that doesn't mean you could easily go faster. For example, we only filter in float; but having int doesn't make it go faster, since the number of units doesn't change (we actually promote most things to float).
Good question. A13 was fabricated in January, but released in May. Maybe another revision was on the way, but didn't came right?
/edit: Morgoth the Dark Enemy: and what's this?
Most of the work was on the drivers -- 2 OS's, new arch, DX9, DX10 and OGL. Work is still ongoing. I'm not sure on the silicon dates, off hand. But at some point, it just made more sense to launch all at the same time.
Thanks -- a slight resemblance of the PentiumPro era and the 32-vs-16-bit performance hype.
Good question. A13 was fabricated in January, but released in May. Maybe another revision was on the way, but didn't came right?
/edit: Morgoth the Dark Enemy: and what's this?
Proof that I should document my claims more carefully?
And yet one of his answers talks about the derivatives being "very popular". If there haven't been any sales, how could that be true? I do wonder if they did launch together, just not in both channels (OEM and retail) simultaneously in order to preference supplies for one over the other in the fat part (volume-wise) of the market. The "sucks to be you" part of such deals is you don't announce for the customer, so you can't beat your chest until he's ready for you to do so.
Of course not. Only actually playing games can give you the full picture.
How are you counting the ALU part of these ratios? Each arithmetic op, scalar or vec4 MAD, clocks on R600?
Ah, things used to be so much simpler in the old days...
Ah yes, back in the old days the only tough decision was what brand of 3dfx Voodoo video card to buy, an Orchid Righteous or a Diamond Monster.
- Status
