The Official RV630/RV610 Rumours & Speculation Thread
- Thread starter Arun
- Start date
- Status
Sound_Card
Regular
Could it be a sound chip?
It's supposedly, based on slides, integrated to the die itself, too.
Sound_Card
Regular
http://it-review.net/index.php?option=com_content&task=view&id=1288&Itemid=1&limit=1&limitstart=6
I hope by different memory configurations, they mean 256bit is a option too. Not just GDDR3/4 or 256mb/512mb.
I very much doubt that is there were both 128 and 256-bit RV630 boards, both would be named HD 2600 XT. Sounds more like just what you said 256 Vs 512 mb.
They all feauture identical UVD based on AMD slides, which can do 1080p @ 40Mbit/s
I think its high time you dropped the idea that there will be a 256bit RV630..
caboosemoose
Regular
Do we have a handle on the shader count and pixels and textures per clock of RV630?
Had a scan of the thread and can't see anything. Various sources indicate 160 shaders (I guess 32 unfied vect5 etc) and therefore 0.5x R600. But have heard a whisper from darkest Africa that it might be less. Same goes for the ROPs and TMUs. Any insight anyone?
Had a scan of the thread and can't see anything. Various sources indicate 160 shaders (I guess 32 unfied vect5 etc) and therefore 0.5x R600. But have heard a whisper from darkest Africa that it might be less. Same goes for the ROPs and TMUs. Any insight anyone?
ATi HD 2600XT will come in fanless design
Even if they manage a fanless design on the Pro versions, then I'd say that is a one nice incentive which could not have been possible on 80nm.
Even if they manage a fanless design on the Pro versions, then I'd say that is a one nice incentive which could not have been possible on 80nm.
I believe 160 (32) is correct. That is just an inference based on the fact G84 is 1/4 (or slightly greater in some specs) of G80. If the R600 is slower than the GTX, or rather on-par with the GTS (~3/4 spec GTX), at least at this point driver-wise, and the 2600 is supposedly slightly faster than the 86GTS, it would seem to make the most sense.
That being said, I initially believed it was 1/4 R600, therefore 80 (16), but that was under the assumption there will be a "mid-life kicker" similar to RV560. If that part doesn't exist, which it may not, it would make sense X2600 would have those specs from the start.
Quite sure actually because you are comparing apples to oranges here. RV630 was always rumored to be clocked higher (compared to G84) >700MHz, so I think AMD would have been hard pressed to get a passively cooled RV630 on 80nm.
Besides the card you linked to seems to be using quite an extensive passive cooler like the Silent X1950Pro. These limited-run silent cards are fine but if either IHV manages to get a ref mainstream board that is passively cooled, color me impressed.
There are other, simpler passive cards with the 8600 GTS, GT and 8500 GT GPU's, too.
That 8500 GT is even using the reference Nvidia single-slot passive heatsink, first seen in the 7600 GS and 7300 GT.
The Asus GTS also quite extensive, just look at the name; "Powerful Heatpipe". Plus its dual-slot to boot.
And isnt 8500 entry-level? I can recall 7300, X1300 entry level cards which also had passive ref cooling designs.
There are dual-slot active coolers for the RV630 XT already.
What makes you think that most passive solutions for the RV630 (if any) will be single-slot ?
Where?
One of the designs, not sure where the final design will end up.
Thats an assumption you are making. I said that if AMD had RV630 on 80nm, they wouldnt have a passive card at all, single or dual slot.
If AMD can manage to not gimp the Pro version and clock it down enough to get a passive ref design, then thats a big win; for their integrators, AIB partners and what not.
- Status
