Really??? Cause he wasn't the first one to be reporting +1000mhz core speeds...Quote:
Originally Posted by NH|Delph1
Printable View
Really??? Cause he wasn't the first one to be reporting +1000mhz core speeds...Quote:
Originally Posted by NH|Delph1
I know I did it months ago, when the chips was in an early stage, and since then the projected final frequency have been rounded down several times.Quote:
Originally Posted by LordEC911
If you have lab chips runnings at 1050MHz, do you really find it feasible that the retail product will do the same?
It's quite feasible we will see partner products with GPU frequencies in the 1GHz range, but the reference card? I doubt it.
I would be very glad if ATI did manage to do it though, but it seems like risking a bit much.
R770 is using the same manufacturing process as RV670, it's even more complex than RV670, the fact that we're seeing higher frequencies coming should be enough to please most people really.
Bsides, 4850 looks like the only card available at launch anyway...
//Andreas
Yes Im sure the 4850 will be the only card at launch because GDDR5 isnt ready just yet.
So we will see the GDDR3 4850 first, followed up by the GDDR5 models a month later.
And by the way, the core frequency of the 4870 is definately going to be 1337 Mhz. You read that right, now spread the info :p
Whats this? If this is right, theyll be undercutting the 9800gtx AND 9800gx2 with those specs.
Release the damn thing already!!
If 800 SP paired with 32 TMU is true, then those 9800GX2 benches may be possible, but I doubt it.
How recent was that pic posted anyways?
mmm seem pretty unbelievable scores there
How many times now?...
With all the BS hooplah going around, I find it hard to believe those scores, but at the same time, I want to...
Didnt toms say itd be 32 TMUs and 16 ROPs.
We need final specs before we can start estimating scores, as my attempts to do so were in vain as well as news of 32 ROPs starting springing up.
Though I somehow doubt ATI is SO spooked that they believe they need to go 480/32/32. That would limit the R6xx and R7xx-type architecture`s longevity to one year after Rv770 launch at most i.e. RV870 would be 640/32/32 and then they would have to release something ENTIRELY new, and I doubt Fusion is going to answer that.
Perkam
isn't the r800 supposed to be fusion?
But I agree, if the card is supposed to debut in late may, we should at least have some legit info on it beyond that its 55nm and is supposed to bring around 50% over the rv670
I want these benchies to be true, but somehow I doubt. But maybe it has got 800 SP and the GDDR5 ups the performance a lot?Quote:
Originally Posted by awhir
And think of how the HD4870X2 would perform compared to this, would easily be 30k+ probably 35-40 in 3dm 06 :D
News article title: FAKE Radeon HD4870...Quote:
Originally Posted by Caubecari
Definitely fake! According to the 3dmark2006 calculator the score should be 15017...
Please stop reposting fakes... it only spreads it around and around
seriously, enough with the fakes, and besides, you can obviously tell that image is fake because there's no way the rv770 would go down from 170mm*2 to 153*2 and gain 50% more shaders with the same manufacturing process
I dont know why everyone is so hyped for this release I had two 3870 in crossfire and its the worst gpu experience ive ever had Ati drivers are garbage what is suppose to change with this release? If your not a hardcore gamer I think these cards will be great but if your a gamer i think Nvidia is a better option unless you like to wait for months for driver releases.
That's because Crossfire sucks, not because ATI sucks ;)Quote:
Originally Posted by malik22
But I agree, if you want fps, get a NV card.
thanks for the constructive arguement in favour of nvidia in this ati thread.Quote:
Originally Posted by malik22
The "4870" used in the 3DMark 2006 GPU Identification is the font from CPU-Z, not 3dMark's font, as proven by the fact that the idiot that made the fake could have picked up the "4", the "8", the "7", and the "0" from the numbers 2 cm below, which use the correct 3dMark Font.
Perkam
http://www.guru3d.com/news/ati-radeo...hmark-results/
I think is fake, but guru3d post them so...
Super Fake-ness.
Posted and disproven before.
Perkam