Anand has the details about r520,rv530,rv515

Maintank said:
Judging by ATi the last couple of rounds. You will be lucky to see the 600Mhz variant in quantity by Christmas. 700Mhz would be a miracle.
I agree. It would be nice if the opposite was true, though.

Also hasn't TSMC got about 3x the capacity of 90nm versus 130nm?

Jawed
 
Maintank said:
Judging by ATi the last couple of rounds. You will be lucky to see the 600Mhz variant in quantity by Christmas. 700Mhz would be a miracle.

Seriously, I can't see ATI pulling another "Phantom Edition", with availability 8 months after the launch. If they can't deliever X1800XT's to us this time..well, the 7800's are right there. ;) Btw, this time I believe NVIDIA has most of the OEM contracts, so it should be raining R520's. :LOL:
 
IbaneZ said:
Seriously, I can't see ATI pulling another "Phantom Edition", with availability 8 months after the launch. If they can't deliever X1800XT's to us this time..well, the 7800's are right there. ;) Btw, this time I believe NVIDIA has most of the OEM contracts, so it should be raining R520's. :LOL:

Raining? I don't see any clouds on my sky ;)
 
IbaneZ said:
Btw, this time I believe NVIDIA has most of the OEM contracts, so it should be raining R520's.

I was wondering roughly what the window timeframe was for hitting the OEM refreshes for spring and fall? That said, it's pretty safe to say ATI has missed the fall refresh so hopefully availability is there.
 
With the huge price difference between the XL and the XT, I'm assuming the XT will be a limited product. In addition, there seems to be only about a 10% difference in performance between the two of them. IMO, the XL would cannabalize the XT sales. ATI would only be willing to do that IF they were not worried about high volumes of XTs selling. The XT will be there for bragging rights, while the XL and Pro will probably be the real competition with Nvidia. Honestly, I'd get rid of the XL if I was ATI and focus on the Pro and XT.

Also, why are they going to "flood" the market with high end skus. Seems weird.
 
It has a footnote that states estimated by the INQ. What's worse is that the INQ doesn't even mention amount of quads in that link hahahaha ;)
 
Ailuros said:
It has a footnote that states estimated by the INQ. What's worse is that the INQ doesn't even mention amount of quads in that link hahahaha ;)

The footnote shows they are only linking to the Inq for the prices of the X1600 and X1300.
 
John Reynolds said:
I was wondering roughly what the window timeframe was for hitting the OEM refreshes for spring and fall? That said, it's pretty safe to say ATI has missed the fall refresh so hopefully availability is there.
If memory serves me right its something like may-june. There was a press release or something like that saying NV had won X number of OEM orders, and it was right around that time frame. So Im guessing thats their deadline for fall (christmas/school opening) refreshes.

epic
 
Rur0ni said:
I can understand the whole "but its 512MB" argument, but as someone stated, nVidia will surely drop their prices a bit once ATI's cards come into play, and the better deal is apparent when the performance is identical but there is a $100 difference.

Yea very well could play out like that which is a good thing for us consumers. But what if that extra memory made a difference? We have seen Wavey's 800xl 512 vrs 256 mb in HL2 under very high settings where it the extra memory helps. So when know for a fact it will help HL2 prefromance. If other new games (FEAR, UT2007, ect) start to show something simular then it would help "warrent" the extra cost. Of course that is a lot of "if"s :)
 
I wonder how much Lost Coast will appreciate >256MB hanging off the GPU.
 
Hmmm... Nothing to explain the "3" for RV530.
I'm beginning to wonder about the credibility of those numbers.

Dave's comments from my memory:
1) the "3" relates to "texturing capabilities" (still hold to this Dave?)
2) Dave questions how to reconcile "4-1-3-2" with "12-pipelines", with following comments suggesting "it's not really 12 pipelines"; but Dave replies with "pipelines with the modern/evolved definition" to paraphrase (probably inaccurately).

Eh, I give up. :(
 
We all gave up.

I think it's:
  • pipes per array
  • TMUs per array pipe
  • arrays
  • ROPs per array pipe
which translates as:
  • 4 pipes per array
  • 4 TMUs
  • 3 arrays (therefore 12 pipes total)
  • 8 ROPs
Jawed
 
Back
Top