RV775XT and PRO have 840 SP?

[onethreehill]

http://www.expreview.com/img/news/2008/12/12/rv775.png

RV775XT to be launched in Jan 2009
RV775PRO to be launched in Mar 2009
Source:Expreview (http://www.expreview.com/news/hard/2008-12-12/1229072147d10842.html)

Translator (http://babelfish.yahoo.com/)

[Stukov]

so, they added a cluster and plan to jack up the clocks?

[Shintai]

More SP, more TMU etc. And less transistors? And massive core clock changes.

Sounds very wishful...

Personally I would just guess its a respin with abit higher clocks.

RV775XT sounds like a nuclear furnance. And dualslot cooling for RV775Pro and with GDDR4?

[Stukov]

More SP, more TMU etc. And less transistors? And massive core clock changes.

Sounds very wishful...

Personally I would just guess its a respin with abit higher clocks.

RV775XT sounds like a nuclear furnance. And dualslot cooling for RV775Pro and with GDDR4?

Maybe they removed some stuff that didn't work or didnt increase speed? Made it more efficient?

[RPGWiZaRD]

I think 1GB vram default might have been better for these cards which cheaper 512MB versions offered slightly later or simililarly in smaller quanitity. I don't doubt there will be 1GB but I think that should become the default version, and 512MB would be treated as HD4870 1GB version as far as availability goes. If they wanna have something to compete with GTX 285 in sales that is, as consumers like the bigger capacity usually if cost is around the same or near. But I doubt RV775XT will cost more than $299 though and GTX 285 will prolly end up $399(+) so there will be some difference in price probably but I'd still have liked to see 1GB default.

[SNiiPE_DoGG]

the fact it says the xt is 512mb makes me think BS on these numbers... why would ATI release a card meant to compete with the 285 with this mem...? especially with the advantages 1gb gave to the 4870 in the ever increasing sized market for high res performance

[Shintai]

March 2009 for the RV770Pro ponders me. Because either the info is fake. Or 40nm parts are farther away than we thought.


the fact it says the xt is 512mb makes me think BS on these numbers... why would ATI release a card meant to compete with the 285 with this mem...? especially with the advantages 1gb gave to the 4870 in the ever increasing sized market for high res performance

Ye that too.

[tajoh111]

How the heck did they add shaders, texture units, drop transister count, reduce size(without switching to a new manufacturing process) and vastly increase coreclocks to liquid cooling levels. Kind of find this hard to believe considering this is supposed to be a modest revision. I also find the bump in texture units off considering the increase in shaders.

Also why have such a powerful card that could kill anything below 1600*1200 and castrate it where it's power will not be overkill by giving it 512mb of memory?

I know AMD has been showing magic lately but this seems a little impossible. These sound like the specs of little dragon almost and rv870 like.

Sounds like a fake leaked rumor to me by AMD to deflate the NV launch of the gtx 285 and 295.

[Stukov]

How the heck did they add shaders, texture units, drop transister count, reduce size(without switching to a new manufacturing process) and vastly increase coreclocks to liquid cooling levels. Kind of find this hard to believe considering this is supposed to be a modest revision. I also find the bump in texture units off considering the increase in shaders.

Also why have such a powerful card that could kill anything below 1600*1200 and castrate it where it's power will not be overkill by giving it 512mb of memory?

I know AMD has been showing magic lately but this seems a little impossible. These sound like the specs of little dragon almost and rv870 like.

Sounds like a fake leaked rumor to me by AMD to deflate the NV launch of the gtx 285 and 295.
They are probably just speculation. I bet it is a mix of some bits of information from a 40nm respin and a 55nm respin.

[clonez]

me wants xt for new rig in jan, but PLEASE, better idle power consumption

[donitsi]

This would make things very interesting ;)
March for PRO doesn't like too good though

[Mats]

me wants xt for new rig in jan, but PLEASE, better idle power consumption
I'm just plain confused about idle power consumption for the 4850, have the power consumption dropped with newer SKU's or drivers or not?

Anandtech (http://www.anandtech.com/video/showdoc.aspx?i=3437&p=10): The 4850 system uses 43 W more in idle than the GTX 260 system.
Bit-Tech (http://www.bit-tech.net/hardware/2008/10/14/palit-xpertvision-radeon-hd-4850-sonic/11): The 4850 and the GTX 260 system uses the same amount of power in idle.
:confused:

[RPGWiZaRD]

Wasn't the issue that 2D clocks didn't work correctly at first and was fixed later?

[Shintai]

I'm just plain confused about idle power consumption for the 4850, have the power consumption dropped with newer SKU's or drivers or not?

Anandtech (http://www.anandtech.com/video/showdoc.aspx?i=3437&p=10): The 4850 system uses 43 W more in idle than the GTX 260 system.
Bit-Tech (http://www.bit-tech.net/hardware/2008/10/14/palit-xpertvision-radeon-hd-4850-sonic/11): The 4850 and the GTX 260 system uses the same amount of power in idle.
:confused:

i780 SLI/x38(Bit-tech) vs i790 SLI (Anandtech) maybe ;)

x38 is far superiour to i780 in power consumption. (Anything is hoho)

[Mats]

Wasn't the issue that 2D clocks didn't work correctly at first and was fixed later?
Sounds possible, just look at the Palit card reviewed at Bit-tech, looks sweet, tho it's not ref design so it's not really comparable.

[Miss Banana]

I'm just plain confused about idle power consumption for the 4850, have the power consumption dropped with newer SKU's or drivers or not?

Anandtech (http://www.anandtech.com/video/showdoc.aspx?i=3437&p=10): The 4850 system uses 43 W more in idle than the GTX 260 system.
Bit-Tech (http://www.bit-tech.net/hardware/2008/10/14/palit-xpertvision-radeon-hd-4850-sonic/11): The 4850 and the GTX 260 system uses the same amount of power in idle.
:confused:

If you update the bios of a 4850, the idle clocks are set way way lower than what is the case with the stock bios. This makes the 4850 consume less power than the gtx 260 in idle.

[Mats]

i780 SLI/x38(Bit-tech) vs i790 SLI (Anandtech) maybe ;)

x38 is far superiour to i780 in power consumption. (Anything is hoho)
Yeah but the differences Im' talking about are measured on the same mobo.

Anand uses EVGA 790i SLI: huge difference in power consumption.
Bit-tech uses Asus Maximus Formula: no difference.

Edit: thx Ms Banana!

[donitsi]

If this rv775 info hold truth, Nvidia better change GTX 265 and GTX 285 specifications or make partners OC these alot. Otherwise GTX 265 will be eaten for breakfast by RV775XT. Even the GTX 285 would be in danger. :eek:
Nvidia can't only change GTX 265 because, it would damage GTX 285 sales with being so close to it's performance.

But for now, let's take this info with grain of salt

[Shintai]

Yeah but the differences Im' talking about are measured on the same mobo.

Anand uses EVGA 790i SLI: huge difference in power consumption.
Bit-tech uses Asus Maximus Formula: no difference.

Edit: thx Ms Banana!

Bittech uses i780 for nVidia cards, x38 for AMD cards.
Anandtech uses i790 for both.

Here is another one with x38 for both.
http://techreport.com/r.x/radeon-hd-4830/power-idle.gif

Anandtech also uses a 1200W PSU. And thats obviously not helping either due to its massive inefficiency.

[roadie]

This is rubbish. Effective 2800Mhz GDDR3 on a high, yet still midrange part? There are so many uneducated guesses in this table it's ridiculous!

[Mats]

Shintai: Yeah you're right, didnt read the whole test setup.

[NapalmV5]

nice boost.. 1GHz core oc easy

but when 1GB memory?

i want 1GHz/1GHz/1GB im sure you guys aswell? :)

[Stukov]

Shintai: Yeah you're right, didnt read the whole test setup.
Oh please tell me you didnt say those words! :rofl:

[bro20000]

Looks like 40nm process is on course to come out soon!
http://www.fudzilla.com/index.php?option=com_content&task=view&id=10911&Itemid=1

[Nedjo]

someone @ experview was really creative... pfff

[Shintai]

Looks like 40nm process is on course to come out soon!
http://www.fudzilla.com/index.php?option=com_content&task=view&id=10911&Itemid=1

I wouldnt expect 40nm before Q2. And fudzilla and sources doesnt mix well :p:

[Stukov]

I wouldnt expect 40nm before Q2. And fudzilla and sources doesnt mix well :p:

So TSMC was ready for volume production of 40nm November 17th why would it take until Q2?
http://techreport.com/discussions.x/15907

[Shintai]

So TSMC was ready for volume production of 40nm November 17th why would it take until Q2?
http://techreport.com/discussions.x/15907

Edit**

LP or GP? ;)

Also from your own link:


So, when will 40nm GPUs come out? That's tough to say, but AMD's 55nm Radeon HD 3800 graphics cards came out roughly 7.5 months after TSMC announced the "readiness" of its 55nm process last year. Extrapolating from that, you can probably expect AMD and Nvidia to launch their first 40nm products in the first half of next year.

[Stukov]

LP or GP? ;)

Light power or Green people? I don't know what you are asking.

[Shintai]

Light power or Green people? I don't know what you are asking.

Usually they ship low power first. Used for cellphones etc.

Plus the ACTUAL headline from TSMC press is:

TSMC Ramps 40nm Volume Production
Most advanced cellular, wireless, and consumer electronic innovations targeted for Foundry’s first 40 nanometer (nm) logic process

http://www.tsmc.com/tsmcdotcom/PRListingNewsAction.do#

[Mats]

Oh please tell me you didnt say those words! :rofl:
Well at least Shintai's post helped me.:rolleyes:

[madcho]

expressview got right most of times, not this one

[Mats]

The OP's link mentions 40 nm in October, that sounds pretty realistic (for a change).

[Stukov]

LP or GP? ;)

Also from your own link:

I see, http://www.tsmc.com/english/b_technology/b01_platform/b010101_45nm.htm would have been more helpful if you included what you were talking about.

As for the link (which wasn't there when I first responded) I am looking into what sort of announcement was made regarding 55nm "volume production".

Edit* Reading a bunch of old news, it seems that the dates you presented arent necessarily correct as there was a big gap in where the GPU generations where sitting. Though 55nm was ready for production in March, neither Nvidia or AMD had a GPU ready for production. AMD hadn't even got the R600 out the door yet on 65nm. So saying that it took AMD x amount of months for 55nm from announcement to release is not exactly an accurate gage.

We could still 40nm chips in Q1 of 09. I still however still expect 40nm RV870 to be around the March/Feb timeframe until I see alot of reports (and conclusive) otherwise.

[Stukov]

Well at least Shintai's post helped me.:rolleyes:

I was being sarcastic.

[Mats]

Hallå, I'm from Sweden, I don't understand sarcasm.:shrug::rofl:

[xsbb]

You say Q2 for retail 40nm? That makes sense.

[Macadamia]

Fake. There is no way to make 840SPs out of RV770 architecture bases (10X8).
It's always increasing in SPs of 80, aka 320 -> 400, 800 -> 880 etc.

And for a new chip this sounds pathetically little. Fake.

[Mats]

March, AMD will release more of the low-level RV775 Pro, based on 40nm As for the RV870 the time to market in October of 2009, AMD did not disclose any information that would once again cooked up one of the new dual-core card, but why Not take action?
:shrug:

[Stukov]

:shrug:

I wouldn't believe it, thats 1.5 years after RV770. I say disinformation.

[003]

Holy crap. I doubt those specs are true, but if they are, this may be my new card. Then again, the proposed specs for the GTX350 are equally if not more impressive. Tough decision if both of them make it to market. I am looking for the fastest single GPU configuration possible.

[w0mbat]

SPs: 800 -> 840
TMUs: 40 -> 48

:shakes:

[Dami3n]

Is rv775 codename "The phantom transistors"? :D

[Mats]

840 doesn't sound like enough, I'd expect 880 or 960.

[Mats]

I wouldn't believe it, thats 1.5 years after RV770. I say disinformation.
It's 23 months, nov 07 to oct 09.

[w0mbat]

Its possible, but i´d expect 960SPs & 48TMUs. But w/ 12*14*5 and 12*4 u´ll get that chip.

[Mr.BSEL]

Correct me if I'm wrong but I don't think its even possible to reach the rv775xt's clocks on 55nm on stock air. I could understand 850 on the core with some tweaks but 950 is way out there.:p:

[cky2k6]

It's 23 months, nov 07 to oct 09.

rv770 came out in june or july of this year...

[Mats]

rv770 came out in june or july of this year...I meant manufacturing process, not model.

[Junos]

You're forgetting that RV770 actually has a spare unit to disable (too lazy to pick up the link, but I'm right), if the core slightly fails (840SPs, 40SPs are disabled = 800SPs). So they may have been collecting succesfull cores for RV775 and use them later on.

[w0mbat]

But RV770 has only 800SPs (10*16*5) and not 840SP. U can see this on the Die-shot.

[Mats]

http://forums.vr-zone.com/showthread.php?t=289806

[Macadamia]

You're forgetting that RV770 actually has a spare unit to disable (too lazy to pick up the link, but I'm right), if the core slightly fails (840SPs, 40SPs are disabled = 800SPs). So they may have been collecting succesfull cores for RV775 and use them later on.

There are redundant (spare) SPs on RV770. There are no redundant TMUs.


Also, the die size would not change if that were the case.

[Junos]

There are redundant (spare) SPs on RV770. There are no redundant TMUs.


Also, the die size would not change if that were the case. Yeah you're right. I knew that there weren't spare TMUs but didn't know how to call the rest 40 SPs. Thanks for fixing my post. :up:

The die size change is very odd indeed, makes me want to doubt the numbers...

[tenebre]

that's because it's not rv770 but rv770 optimized. -read new design

[SNiiPE_DoGG]

Yeah you're right. I knew that there weren't spare TMUs but didn't know how to call the rest 40 SPs. Thanks for fixing my post. :up:

The die size change is very odd indeed, makes me want to doubt the numbers...


that's because it's not rv770 but rv770 optimized. -read new design

you guys wanna know why there are spare TMUs where there shouldnt be?

because these specs are
F
A
K
E

:up:

[Monkeywoman]

hey they are fake, anyone can get a news site to believe them enough to post fud.

[ToTTenTranz]

The extra SPs and less transistors could easily be explained by ATI putting the redundancy down a notch (55nm should be fully mature by now).
The additional TMUs could already be there.. or could be a result of a late redesign.


The ammount of "this is fake" posts I see in this thread reminds me of when the first leaks came out about a 800SPs/40TMUs design for RV770.
Everyone likes to be the first to say "I knew from the beggining this would be a fake". Oddly, when the same people end out being wrong, I never once saw a post saying "I guess I was wrong from the beggining".



Don't forget that ATI/AMD has to actively counterattack the "GTX260 core 216 + Big Bang II" phenomena. Even more if the GT200b 55nm cards end up with higher clocks.

[Ozzfest05]

Well obviously ATI is going to release their new cards with over 512 I mean even the 4850 has a 1gb version

[LordEC911]

Fake. There is no way to make 840SPs out of RV770 architecture bases (10X8).
It's always increasing in SPs of 80, aka 320 -> 400, 800 -> 880 etc.

And for a new chip this sounds pathetically little. Fake.
Ummm... RV770 march is 10 clusters of 16 v5 shaders.
Your statement is correct, the next step would be 880 but I'm not sure how you got there with 10x8...


There are redundant (spare) SPs on RV770. There are no redundant TMUs.

Also, the die size would not change if that were the case.
Please see Wombat's post.

[SNiiPE_DoGG]

The extra SPs and less transistors could easily be explained by ATI putting the redundancy down a notch (55nm should be fully mature by now).
The additional TMUs could already be there.. or could be a result of a late redesign.


The ammount of "this is fake" posts I see in this thread reminds me of when the first leaks came out about a 800SPs/40TMUs design for RV770.
Everyone likes to be the first to say "I knew from the beggining this would be a fake". Oddly, when the same people end out being wrong, I never once saw a post saying "I guess I was wrong from the beggining".



Don't forget that ATI/AMD has to actively counterattack the "GTX260 core 216 + Big Bang II" phenomena. Even more if the GT200b 55nm cards end up with higher clocks.

yeah but more TMUs and SPs with less transistors makes no freakin sense without a process shrink.....

[LordEC911]

yeah but more TMUs and SPs with less transistors makes no freakin sense without a process shrink.....

Process shrink wouldn't decrease transistor count, only density/size of the transistor creating a smaller die.
There are quite a few things they could have removed that would cause a decrease in trannies.
UVD was one brought up in the other thread and the sideport would be the other.

[ToTTenTranz]

yeah but more TMUs and SPs with less transistors makes no freakin sense without a process shrink.....

It's happened before.
In the transition from G70 to G71, nVidia managed to put about 9% less transistors in G71, maintaining the same general design and active units.


Here we would see only a 3% less transistors from RV770 to RV775. As told above, the increase in TMUs and SPs could also be the replace for the UVD, which is apparently no longer needed.

Cutting down the sideport could compromise the performance of a dual-RV775 (R720/R780?) for computing applications.

[HKPolice]

The SP spec is fake, an extra 40 isn't worth the trouble and doesn't fit mathematically anyways. The extra TMUs are possible and the size shrink is possible too.

[zerazax]

As others ahve stated, the SP count doesn't work. However, if they went to 960 SP's, the 48 TMUs would match up properly.

As for clocks... keep in mind the 55nm process is constantly improving, and they've been on 55nm for a year now. A Rev A12 respin could probably reach clocks close to that (which, however, seems too high on 55nm - 40 nm seems possibly). Still, if they pull this off (or heck, a 950 mhz rv770) it would be one pimpin card

And as others have pointed out, die size and transistor count can decrease by taking sideport and redundancy pieces out. Also, no one believed RV770 could fit 800 SP's in that die size with < 1 billion transistors, and ATI pulled it off just fine last I checked

[Luka_Aveiro]

I think this whole RV775XT and RV775Pro rumors are nothing but bluff...

[spursindonesia]

One thing is pretty obvious to me, do we all think ATi is gonna sit tight doing nothing with their d!ck hanging while nVidia running circle around them with 55nm parts of GT200 mArch ? ATi can see it coming from miles away, that their performance crown, which is a strong HALO beacon after R600 debacle, will be gone with GT200b, will they be thrifty & stupid enough not to create/develop a worthy competitor for these nVida new aces ? After all the financial success and efficiency of their RV770 mArch ? But well, if they do have something in their sleeves, a new ace generation with 40 nm node coming SOON -early Q2 2009 soon, perhaps they will swallow this loss and compete solely on price for the next few months.

[tajoh111]

How could anyone explain how they ramp up the clocks to good liquid cooling levels without a new process?

Unless they are crazy cherry picked with the very best after market cooler, it will be impossible to reach 900 mhz(unless they are pumping incredible volts).

If these are cherry picked chips, expect quantities to be limited.

The pro could be rebadged 4870 obviously.

[Luka_Aveiro]

But well, if they do have something in their sleeves, a new ace generation with 40 nm node coming SOON -early Q2 2009 soon, perhaps they will swallow this loss and compete solely on price for the next few months.

They have done it with HD38xx series, and if they did it with HD48xx series, they wouldn't be losing much money, because AMD graphics cards have done very well since HD48xx's series debut.

And, all in all, i think everyone would like to have a cheap 1Gb HD4870 ;)

[Shintai]

40nm and RV870 is simply too close. Or even in the case. 40nm Rv770 in Q2 if we assume no RV870. It just doesnt make sense to do any addons whatever. Plus the specs shown doesnt match anything. It looks to be some random forum posters wish as usual.

If AMD gonna do anything until 40nm. It would be a simple respin only for higher clocks. But even that...for a few months on the market?

[Boissez]

This is rubbish. Effective 2800Mhz GDDR3 on a high, yet still midrange part? There are so many uneducated guesses in this table it's ridiculous!
... must be GDDR4 then. ;)

[Voodoo²]

RV670

4 SIMD cores * 16 modules * 5 Alus (4+1) = 320 sp
4 TMU per SIMD unit = 16 TMUs

RV770

10 SIMD cores * 16 modules * 5 Alus (4+1) = 800 sp
4 TMU per SIMD unit = 40 TMUs

RV775 ???

12 SIMD cores * 14 modules * 5 Alus (4+1) = 840 sp
4 TMU per SIMD unit = 48 TMUs

or

6 SIMD cores * 28 modules * 5 five Alus (4+1) = 840 sp
8 TMU per SIMD unit = 48 TMUs

I think is fake cause the 512MB doesn´t look good but the math is doable. I don´t know much about GPU´s and their internal structures and how they work but maybe if RV775 is true is just a rearrengement of this structures too make it easier to program and to make it more efficient.

[Helmore]

RV670

4 SIMD cores * 16 modules * 5 Alus (4+1) = 320 sp
4 TMU per SIMD unit = 16 TMUs

RV770

10 SIMD cores * 16 modules * 5 Alus (4+1) = 800 sp
4 TMU per SIMD unit = 40 TMUs

RV775 ???

12 SIMD cores * 14 modules * 5 Alus (4+1) = 840 sp
4 TMU per SIMD unit = 48 TMUs

or

6 SIMD cores * 28 modules * 5 five Alus (4+1) = 840 sp
8 TMU per SIMD unit = 48 TMUs

I think is fake cause the 512MB doesn´t look good but the math is doable. I don´t know much about GPU´s and their internal structures and how they work but maybe if RV775 is true is just a rearrengement of this structures too make it easier to program and to make it more efficient.

RV670 and RV770 can't be compared as they are quite different in the way their internal structure is build up. On RV770 the shader SIMD cores and the TMU units (quads) are tied to each other, but on RV670 this was not the case and the TMU units where shared between shader unit clusters from all 4 SIMDs. Also the option with 12 SIMD units and 14 modules is not really feasible as far as I know, because you should be able to divide it by 4 as the SIMD unit has to work on pixel quads. I'm not really sure about though, but all their current design only have 8 or 16 modules per SIMD and I don't think they'll deviate from that (so we can also scrap the 6 SIMD version). Unless this is a completely new chip and that doesn't seem to be the case.

To me it all looks to be fake, although I don't think AMD won't release anything in Q1 of 2009. My guess is that we might see RV740 before the end of Q1.

[Boissez]

RV670

4 SIMD cores * 16 modules * 5 Alus (4+1) = 320 sp
4 TMU per SIMD unit = 16 TMUs

RV770

10 SIMD cores * 16 modules * 5 Alus (4+1) = 800 sp
4 TMU per SIMD unit = 40 TMUs

RV775 ???

12 SIMD cores * 14 modules * 5 Alus (4+1) = 840 sp
4 TMU per SIMD unit = 48 TMUs

or

6 SIMD cores * 28 modules * 5 five Alus (4+1) = 840 sp
8 TMU per SIMD unit = 48 TMUs

I think is fake cause the 512MB doesn´t look good but the math is doable. I don´t know much about GPU´s and their internal structures and how they work but maybe if RV775 is true is just a rearrengement of this structures too make it easier to program and to make it more efficient.

If (big if) those rumours are true I would guess that ATI would choose to add more regular ALU's to each stream processing unit (let's say 6 instead of 4) the math would then be something like:
12 SIMD cores*10 modules*7 ALUs(6+1)= 840SP

We'd a get larger but simpler SPU's giving a smaller footprint and clocking higher - makes sense no? :shrug:

[zanzabar]

they wouldent add more scalers (or alu) thats their week point, and changing the module count would mean that they would need to change the memory buss

[ColonelCain]

As much as I would like for this to be true, I don't see this as being true. Reiterating what others have mentioned:
The extra 40 shaders don't fit with their current cluster size. And, the Shader/TMU ratio would be of-set.
In order to get those kinds of core clocks, they would have needed to go back, pick apart the architecture and try and optimize current leaks.
512mb? If they really don't think that 1GB isn't very important, then why would they even bother creating an official 4870 1GB version?

Just a few quick points, but unfortunatly, I doubt this as truth.

[informal]

As much as I would like for this to be true, I don't see this as being true. Reiterating what others have mentioned:
The extra 40 shaders don't fit with their current cluster size. And, the Shader/TMU ratio would be of-set.
In order to get those kinds of core clocks, they would have needed to go back, pick apart the architecture and try and optimize current leaks.
512mb? If they really don't think that 1GB isn't very important, then why would they even bother creating an official 4870 1GB version?

Just a few quick points, but unfortunatly, I doubt this as truth.

What about 960 stream/48TMU ?:) Ratio is fine :D.It could be an option for a 40nm shrink.
The above is just one very far fetched possibility since the die area didn't increase that much from the report we saw in OP.

[ColonelCain]

What about 960 stream/48TMU ?:) Ratio is fine :D.It could be an option for a 40nm shrink.
The above is just one very far fetched possibility since the die area didn't increase that much from the report we saw in OP.

True, but this would mean that RV870 is delayed, which wouldn't be a logical gamble for ATI. Risk GT200b beating a 40nm refresh, and have a longer wait to RV870 than if they forgo a refresh?
It WOULD however follow their "new strategy", that they announced awhile back. I think that it was around every 6 months alternating new arch/new refresh?

One thing that I found kind of interesting, concerning the redundancy debate, is that when you look at a die shot of Rv770 detailing the chips internals, there is possibility for their argument, at least as far as shaders go. If you look to the left of the "SIMD group", it looks almost like there are 4 extra shaders per group. X10, and we would get the extra 40. If you don't see it, look at the pattern.

[informal]

True, but this would mean that RV870 is delayed, which wouldn't be a logical gamble for ATI. Risk GT200b beating a 40nm refresh, and have a longer wait to RV870 than if they forgo a refresh?
It WOULD however follow their "new strategy", that they announced awhile back. I think that it was around every 6 months alternating new arch/new refresh?

One thing that I found kind of interesting, concerning the redundancy debate, is that when you look at a die shot of Rv770 detailing the chips internals, there is possibility for their argument, at least as far as shaders go. If you look to the left of the "SIMD group", it looks almost like there are 4 extra shaders per group. X10, and we would get the extra 40. If you don't see it, look at the pattern.
I think you are right,there seems to be 4 extra shaders per group.Now,if they can be active with the rest or can be used just for redundancy purposes is unknown atm.It would be nice if 840SP rumor is true,but honestly how much more perf. these 40 stream units can bring?20% more TMUs and higher core clocks on the other hand can be more effective.

[Voodoo²]

What about if they get rid of the tessellator? That could explain the decrease in transistors count too. Right?:shrug:

[Stukov]

40nm and RV870 is simply too close. Or even in the case. 40nm Rv770 in Q2 if we assume no RV870. It just doesnt make sense to do any addons whatever. Plus the specs shown doesnt match anything. It looks to be some random forum posters wish as usual.

If AMD gonna do anything until 40nm. It would be a simple respin only for higher clocks. But even that...for a few months on the market?

Dear god, you agree with me. I am convinced the world is coming to an end.

[LordEC911]

I think you are right,there seems to be 4 extra shaders per group.Now,if they can be active with the rest or can be used just for redundancy purposes is unknown atm.It would be nice if 840SP rumor is true,but honestly how much more perf. these 40 stream units can bring?20% more TMUs and higher core clocks on the other hand can be more effective.
It's not... RV770 has only 800SP, just like G80 only had 128.


What about if they get rid of the tessellator? That could explain the decrease in transistors count too. Right?:shrug:
Can't...

[Stukov]

True, but this would mean that RV870 is delayed, which wouldn't be a logical gamble for ATI. Risk GT200b beating a 40nm refresh, and have a longer wait to RV870 than if they forgo a refresh?
It WOULD however follow their "new strategy", that they announced awhile back. I think that it was around every 6 months alternating new arch/new refresh?

One thing that I found kind of interesting, concerning the redundancy debate, is that when you look at a die shot of Rv770 detailing the chips internals, there is possibility for their argument, at least as far as shaders go. If you look to the left of the "SIMD group", it looks almost like there are 4 extra shaders per group. X10, and we would get the extra 40. If you don't see it, look at the pattern.

The problem is, even if the RV870 is delayed and they are going to the 'tick/tock' model of releasing products there is still an issue with inventory. I think there are still plenty of RV7xx generation boards out on the market with a good amount of supply. To release a new chip would certainly be canniblelistic on their already created chips. Then, by the time you get a good supply for this 'new' chip, the RV870 should be out, again running into problems with inventory and the like.

I don't see any reason AMD would NEED to have something until the RV870. Their chips are smaller and cards are cheaper to make. If Nvidia comes out of the blocs, AMD will just throw a price war at them. Then just make a splash for the RV870.

[hurleybird]

What about if they get rid of the tessellator? That could explain the decrease in transistors count too. Right?:shrug:

It could, but I hope they didn't, since there might be some use for it in DX11 titles. If you ask me, enabling disabled "repair" units, getting rid of sideport, and some general tweaking should be enough to make the transistor count plausible. Not nearly as drastic as G70 -> G71 at least. 512MB looks suspicious, but if anything means that ATI is thinking of launching 512MB cards along with the 1024MB cards. Changing the ALU:TEX ratio makes sense since R600/R700 is a little lopsided towards shading power anyway. I don't have a problem reconciling the clock speed increase seeing as NVIDIA's 1.4 Billion Transistor chip built on a larger process gets better clockspeed/watt so there are probably some architectural problems that would be a priority for ATI to fix in both RV775 and RV880. Overall, I think the specs look credible, but as we know with ATI rumors whether or not something looks credible doesn't always mean a whole lot.

[Shintai]

What about if they get rid of the tessellator? That could explain the decrease in transistors count too. Right?:shrug:

Please dont try and curve this faked news into something so it fits some hopes. We already have enough hype.

Shader amount doesnt mix, transistor count doesnt mix, clock and process node doesnt mix. Shader/TMU ratio doesnt mix. Even memory doesnt mix.

Is there even anything that do mix?

[Macadamia]

LOL @ RV870 Q4 09.


I love ATI PR. :ROTF:

[spursindonesia]

Well, 800 SP and 40 TMU chip with 260 mm^2 die size didn't exactly mix quite well too, but look who had to eat his hat saying nVidia would rule the market unchallenged in 2008. :rofl:

[NH|Delph1]

There is talk about one card before lil' dragon but it's not this one from what I can tell.

Fudzilla is not far from the truth with its 40nm story.

//Andreas

[marten_larsson]

There is talk about one card before lil' dragon but it's not this one from what I can tell.

Fudzilla is not far from the truth with its 40nm story.

//Andreas

What else have you heard? I've been watching NH and it seems like you guys have quite good intel (not the semiconductor company) on AMD/ATI products. Is the RV740 (that's what you guys called it back in July) going to be big upgrade from RV770? By looking at the name it shouldn't be? Is it going to be a mainstream product (around 4670 performance)?

[clonez]

afaik mainstream, between hd4670 and hd4830

pure volume product and small die size
ideal to test 40nm process on

[NH|Delph1]

What else have you heard? I've been watching NH and it seems like you guys have quite good intel (not the semiconductor company) on AMD/ATI products. Is the RV740 (that's what you guys called it back in July) going to be big upgrade from RV770? By looking at the name it shouldn't be? Is it going to be a mainstream product (around 4670 performance)?

Update coming, I just want to confirm things. If I posted every little tip I get I would end up with a bruised reputation and have about as much credibility as ... well some other sites posting information on pre-released products :)

//Andreas

[Helmore]

Then maybe you should create a NordicHardware rumor mill where all sorts of rumors end up. Would be fun to see what the imagination of some can be like :p:.

[NH|Delph1]

Then maybe you should create a NordicHardware rumor mill where all sorts of rumors end up. Would be fun to see what the imagination of some can be like :p:.

No Rumor mill coming but I have a different solution that should do about the same thing :p:

//Andreas

[Mr.BSEL]

According to ATIforum, there won't be a 55nm refresh of the rv770. Fudzilla also has this news posted as well.

http://translate.google.com/translate?u=http://news.ati-forum.de/index.php/news/34-amdati-grafikkarten/173-rv770-refresh-kommt-nicht&sl=de&tl=en&hl=en&ie=UTF-8

[informal]

Yeah Fudo picked it up too.It is quite logical that the next refresh of RV770 will be done on 40nm process.

[Face]

.. http://www.theinquirer.net/inquirer/news/020/1050020/ati-s-hd-49x0-coming-soon-ish

[Slovnaft]

ugh, i was hoping make a nice christmas present of RV870.

[RPGWiZaRD]

ugh, i was hoping make a nice christmas present of RV870.

Umm, RV870 is the next gen series with DX11 support not the refresh they're talking about. Don't hope it's true that they're aiming RV870 for as late as Q4 2009, that seems like too far away but it wouldn't be very suprising with ATI/AMD. Nvidia's cards aimed for "back-to-the-school-timeframe" I'm pretty sure of though so they'll prolly have some month's advantage. It's the INQ tho so I won't put too much trust in that article but if Q4 is the real target for next gen then it would make sense with a RV775 refresh in like February-April timeframe. Oh well let the cards arrive when they do, no need to speculate.

[Helmore]

Umm, RV870 is the next gen series with DX11 support not the refresh they're talking about. Don't hope it's true that they're aiming RV870 for as late as Q4 2009, that seems like too far away but it wouldn't be very suprising with ATI/AMD. With Nvidia's cards aimed for "back-to-the-school-timeframe" I'm pretty sure of. It's the INQ tho so I won't put too much trust in that article but if Q4 is the real target for next gen then it would make sense with a RV775 refresh in like February-April timeframe. Oh well let the card

I think we will be getting a refresh, but probably the refresh will be a little later and will be made on TSMC's 40 nm. process. Just my guess though, but I think it's pretty accurate.

[zanzabar]

Umm, RV870 is the next gen series with DX11 support not the refresh they're talking about. Don't hope it's true that they're aiming RV870 for as late as Q4 2009, that seems like too far away but it wouldn't be very suprising with ATI/AMD. Nvidia's cards aimed for "back-to-the-school-timeframe" I'm pretty sure of though so they'll prolly have some month's advantage. It's the INQ tho so I won't put too much trust in that article but if Q4 is the real target for next gen then it would make sense with a RV775 refresh in like February-April timeframe. Oh well let the cards arrive when they do, no need to speculate.

dx11 is working on the 3xxx and 4xxx with win7 right now, u just need a fully programmable shader (or driver or bios, forgot how NV dose theirs something with drivers and cuda but ati is software fully programmable), and it has support for an options teslation engine that ati has as they are fully dx10.1

then the other part the r600, rv670 and rv770 were all scheduled for late q3/q4 and depending on how the dev cycle went they were out early or late so its up for grabs what the real time frame is

[Helmore]

Nope DX11 does not fully work on any available card right now. Some of DX11's features can be used with current cards though.

[ColonelCain]

I think we will be getting a refresh, but probably the refresh will be a little later and will be made on TSMC's 40 nm. process. Just my guess though, but I think it's pretty accurate.

Yeah, I am thinking this aswell.

Makes sense now, because a 40nm refresh wouldn't have made sense if the RV870 was coming out in 1H 09.

One question though: I thought that before the RV775 rumor came about, everyone was saying RV870 in Q2 2009?

[Helmore]

Yeah, I am thinking this aswell.

Makes sense now, because a 40nm refresh wouldn't have made sense if the RV870 was coming out in 1H 09.

One question though: I thought that before the RV775 rumor came about, everyone was saying RV870 in Q2 2009?

Pretty much everyone was saying that, but where did it come from is the question. I can't remember any proper source for that date at the moment. The date could still be valid, nobody except ATI knows what they plan on doing next year. Pretty much the only certainty is RV740 which will be their first 40 nm. chip. It's not sure what kind of chip it is though, w0mbat seems to be hinting that the chip is something more than just a RV730 done on 40 nm. and that it will actually use GDDR5.
http://forum.beyond3d.com/showpost.php?p=1248915&postcount=114

[LordEC911]

Pretty much everyone was saying that, but where did it come from is the question. I can't remember any proper source for that date at the moment. The date could still be valid, nobody except ATI knows what they plan on doing next year. Pretty much the only certainty is RV740 which will be their first 40 nm. chip. It's not sure what kind of chip it is though, w0mbat seems to be hinting that the chip is something more than just a RV730 done on 40 nm. and that it will actually use GDDR5.
http://forum.beyond3d.com/showpost.php?p=1248915&postcount=114

You might want to go back and read through the fourth page carefully, lot more information there than you might have realized.
My speculation for RV740 seems to be right on and reaffirms the rumors we heard of tapeouts a few months ago, RV870 in Sept/Oct, RV740 in July.

[Jor3lBR]

So can we expect 4870x2 o bellow $350 on January/09?

[ColonelCain]

So can we expect 4870x2 o bellow $350 on January/09?

Who knows. It really depends how much GTX295 beats it by. AMD will price the respective competitors of GTX285/295 aggressively to get a decent price/performance ratio. This will happen for a little bit, until ATI has a chance to respond with new products.

[LiquidReactor]

So can we expect 4870x2 o bellow $350 on January/09?

:ROTF: doubtfully. It will be a miracle that a gtx 280 regular price will be below $350 not counting in special deals or MIR much less a 4870x2 under $350.

[Macadamia]

You might want to go back and read through the fourth page carefully, lot more information there than you might have realized.
My speculation for RV740 seems to be right on and reaffirms the rumors we heard of tapeouts a few months ago, RV870 in Sept/Oct, RV740 in July.

I'm sorry, your speculation and current rumours are just like the "R700 = 2009" rumours (Kyle Benett's tutu D: )... :p:


RV740 = 1Q 09, displaces 4830 and maybe 4850 with smaller chip that uses GDDR5.

RV870 = 2Q+ 09, new mofo.

R800 = 3-4Q 09, might be skipped if RV870 is not DX11 compliant (but most probably is)

[thebluemeanie1]

Would there be any chance that the die size and transistor reduction may have some truth if ATI removed the sideport?

[LordEC911]

I'm sorry, your speculation and current rumours are just like the "R700 = 2009" rumours (Kyle Benett's tutu D: )... :p:
Not really...

[Macadamia]

Err... RV740 is at ES Stage already. A11 is running fine, but retail should be using A12 to enhance power-saving functions.
If it was a July product, we will see something like RV770 progress this year- ES popping up at March or April.
This is more like a Feb/March launch IMO.

[gumballguy]

I cant believe you guys are discussing this.

READ THE PIC! "all 775 data is to confirmed"

As in, 'is to be confirmed'
As in, is pure speculation.

Now we're semi OT just making our own estimates about which chip will exist and when it will happen... time to lock the thread :p

[largon]

Here's my guess for "RV775" (= cherry picked RV770).
10 SIMDs
900 ALUs
40 TMUs

Yep.
90 ALUs per SIMD.
Cause that's what I see when I look at the die of RV770. There's nomore than 10 SIMDs but each of the ten SIMDs has infact 9, and not 8 clusters. No redundant TMU exists on RV770.

[cadaveca]

Here's my guess for "RV775" (= cherry picked RV770).
10 SIMDs
900 ALUs
40 TMUs

Yep.
90 ALUs per SIMD.
Cause that's what I see when I look at the die of RV770. There's nomore than 10 SIMDs but each of the ten SIMDs has infact 9, and not 8 clusters. No redundant TMU exists on RV770.

:up:



:clap:



I was wondering how long it would take someone to look at the die shot...:rolleyes:

It's also possible a respin removed UVD hardware and replaced with ALU clusters, but not likely.

"RV775XT" is the AMD equivalent of GTX260 216...:shrug:...TSMC's process has matured enough for such as show by nV already...

[w0mbat]

Can u post a high-res die-shot pls.

[cadaveca]

It's on wikimedia...has been for a few months now...

http://commons.wikimedia.org/wiki/File:RV770_Die_Shot.png

[w0mbat]

I can see the shaders, but i cant see where u two detect the 900sp.

[largon]

10 rows and 9 identical blocks in each. The "9th block" being the redundant part would be my guess.

91041

[spursindonesia]

At this moment, i'm affraid ATI's too financially constrained to make an anticipating move regarding nVidia GT200b cards, and all that they can do is just competing through price adjustment -if nVidia plays hardball and agressive in holding/reclaiming their current/lost market share. It's a shame though, since i do believe the RV770 mArch is quite efficient (in terms of die size vs performance) and expandable, relatively compared to nVidia's competing product.

[w0mbat]

AMD has more than enough money cause they were bought by some sandpeople.

[Shintai]

AMD has more than enough money cause they were bought by some sandpeople.

The sandpeople just took the fabs and didnt care about the rest.

[w0mbat]

Ok, could be.

@largon: Ur in the news: http://www.hardware-infos.com/news.php?news=2597 :D

[Shintai]

10 rows and 9 identical blocks in each. The "9th block" being the redundant part would be my guess.

91041

But the 9th row got a different size and shape.

[w0mbat]

No, compare it to the 4th.

[Helmore]

No, compare it to the 4th.

Shintai is right, that ninth row is different and is not some form of extra shader unit. It probably has more to do with scheduling and stuff like that.

LOL at Hardware-infos, they really do post everything they see :p:.

[w0mbat]

Maybe, maybe not. There was talk about more than 800SPs many times before this.

[NBF]

Interesting theory, not the first time I hear it. But this time there is a picture :D

Goes to the news (http://www.framebuffer.com.br/index.php)

[cadaveca]

:rofl:


Are you serious?

[Calmatory]

:rofl:


Are you serious?

Well, everyone could just as well shutup and wait till the thing gets launched, if you prefer that. :rolleyes:

[Voodoo²]

From Xbitlabs:
http://www.xbitlabs.com/images/video/ati-radeon-hd4850/rv770_crystal.gif

This is from june.

[ColonelCain]

Here's my guess for "RV775" (= cherry picked RV770).
10 SIMDs
900 ALUs
40 TMUs

Yep.
90 ALUs per SIMD.
Cause that's what I see when I look at the die of RV770. There's nomore than 10 SIMDs but each of the ten SIMDs has infact 9, and not 8 clusters. No redundant TMU exists on RV770.


:up:



:clap:



I was wondering how long it would take someone to look at the die shot...:rolleyes:

It's also possible a respin removed UVD hardware and replaced with ALU clusters, but not likely.

"RV775XT" is the AMD equivalent of GTX260 216...:shrug:...TSMC's process has matured enough for such as show by nV already...
Hey! I was the first one to see that! No fair! :p::rofl: (Don't mind)

10 rows and 9 identical blocks in each. The "9th block" being the redundant part would be my guess.

91041
Thanks for the picture largon. I actually spent a good chunk of time researching this rumor to see if it had any substance. And, when I looked at the die shot, I noticed the extra group of shaders. Was thinking of doing a quick paint edit and posting it, but had to resume studying for finals.

From Xbitlabs:
http://www.xbitlabs.com/images/video/ati-radeon-hd4850/rv770_crystal.gif

This is from june.
This is the picture that I saw.

Anyhow, for you guys whom can't see it:
http://www.elitebastards.com/hanners/ati/rv770/die.jpg
Look closely. If you think of it like this: Each "corner" structure is part of an SIMD block, you get 4. X5 (for the 5 shaders in each unit), then X4 for the 4 blocks composing each SIMD core. Then X10 for all 10 SIMD cores. You get 800.
And, for those that need an even bigger view, here's an 861 x 848 image that I found (but chose to leave it as a url, so people wouldn't complain about it being too big).
http://upload.wikimedia.org/wikipedia/commons/a/a4/RV770_Die_Shot.png


Anyhow, if you compare the 1st and 9th rows they look nearly identical. Perhaps, they are different because there needed to be a slight design alteration to put in a half "block"? Do I know that that is the exact reason? No. But one thing that I am fairly sure of is that the "structure", is the same as the shaders to its right.

Lastly, for anyone who don't know where largon got the 900 number from, it looks like each of the 10 "half-blocks" in the 9th row have 2 units each. 2x5x10=100.

I don't want people to get the idea that I am trying to give a math lecture on simple adding and multiplication. I just am laying it out there for anyone that is confused. :cool:

[cadaveca]

:up:

BTW, that block diagram from XBIT isn't exactly correct, but whateva.:rofl:

[Helmore]

That block that you refer to ColonelCain is not as similar as the other blocks as you think.
Here is another bigger die shot:
http://picasaweb.google.nl/lh/photo/h1CSnr-icTvr9zQQrg1_Xg
Hope the link works like that for everyone. You can download the picture from this link if you want, that may make zooming in easier although the conversion from png to jpg has decreased it's picture quality :(.
This extra block might have something to do with scheduling or something, but I don't think it's some sort of extra shader unit.

[Monkeywoman]

:rofl::rofl::rofl::rofl::rofl::rofl::rofl::rofl::r ofl: Xtremesystems has done it again!! look over here; http://forums.vr-zone.com/showthread.php?t=368145 !!!!!

and then look over here http://www.xtremesystems.org/forums/showpost.php?p=3509705&postcount=120 !!!!!!

[tenebre]

That block that you refer to ColonelCain is not as similar as the other blocks as you think.
Here is another bigger die shot:
http://picasaweb.google.nl/lh/photo/h1CSnr-icTvr9zQQrg1_Xg
Hope the link works like that for everyone. You can download the picture from this link if you want, that may make zooming in easier although the conversion from png to jpg has decreased it's picture quality :(.
This extra block might have something to do with scheduling or something, but I don't think it's some sort of extra shader unit.

the 9'th column is different

[STaRGaZeR]

Man largon what have you done...

[largon]

Whoa...
And all I said was "Here's my guess..." and "...would be my guess".
:shrug: :D

[Tiridum]

zoom into that picture, the 9th row of "shaders" looks different from a regular row, the light pink things are higher and there is a dark pink column missing between the light pinks things and the texture units

[enteon]

I think they're different, too.

and it wouldn't make any sense for AMD not to activate all SIMDs on their very expensive and presumably high-quality FirePro V8700 which, in fact, only has 800. If even the most qualitative chips don't come with more than 800 usable SMIDs then there simply aren't more. :shrug:

[STEvil]

maybe those are the parts for the currently disabled sideport?

edit - or power controllers to enable/disable shader arrays for power savings.. just didnt work out on 4870 maybe?. woo theories.

[cadaveca]

ONly half of the claimed parts are SP cores...the bottom half. It's not 900SP, hence me laughing about this being news earlier.

[RejZoR]

It wasn't really a news and it wasn't related to RV775 series. I found it somewhere else and just asked here for confirmation if that is true. It turned out it is, partially.
But nothing end users can actually use.

[renzokuken]

At the risk of sounding noobish (but in the spirit of learning!)...how important are SP's? I never understood how, for example, the GTX 280 has 240 SP's while the 4870 has 800 (!) and yet the 4870 isn't leaps and bounds better than the GTX 280 (in fact it's worse despite having more SP's and a higher GFLOPS output). Where does the "problem" lie in ATI's cards (the architecture?) and why can't we compare the performance between the two cards based on the # of SP's in ATI cards to the # of SP's in Nvidia cards? If anyone can answer either of these questions it would be greatly appreciated.

[hurleybird]

I never understood how, for example, the GTX 280 has 240 SP's while the 4870 has 800 (!) and yet the 4870 isn't leaps and bounds better than the GTX 280 (in fact it's worse despite having more SP's and a higher GFLOPS output)

GTX 280's SPs are clocked much higher, and, on average, operate closer to their theoretical power than the SPs in R700.

[Monkeywoman]

ATI SPs are locks to the same speed as the core (4870-750/750/900) with Nvidia they are separate (8800gtx-630/1000/1350)

one ATI SP can do 5 things per clock where as the nvidia can do less but it equals out to be more because of the shader clock AFIK:shrug:

[renzokuken]

GTX 280's SPs are clocked much higher, and, on average, operate closer to their theoretical power than the SPs in R700.

Awesome! Thanks so much for the exact answer I was looking for.

[Macadamia]

At the risk of sounding noobish (but in the spirit of learning!)...how important are SP's? I never understood how, for example, the GTX 280 has 240 SP's while the 4870 has 800 (!) and yet the 4870 isn't leaps and bounds better than the GTX 280 (in fact it's worse despite having more SP's and a higher GFLOPS output). Where does the "problem" lie in ATI's cards (the architecture?) and why can't we compare the performance between the two cards based on the # of SP's in ATI cards to the # of SP's in Nvidia cards? If anyone can answer either of these questions it would be greatly appreciated.


Let's put this way.

ATI's architecture has 160 Execution blocks. Each of the blocks can do 5 ops simultaneously if under an ideal condition in VLIW. In most cases 3-4.5 per 5 "shaders" is used.


nVidia's GT200 architecture has 240 ALUs that can do dual issue. They also run at 1+GHz clock speeds, which makes them faster in processing.

However, there are a few notes to consider:
1. GT200 has 80 TUs. RV770 has 40 (32 interpolators 40 filters).
2. GT200 has 32 ROPs. RV770 has 16 (which are on par until 8x MSAA, where they take the lead in Z-fill)
3. GT200 is almost 2x the size of RV770 with both in 55nm (480 mm^2 vs 256 mm^2)

It's not as simple as you think. Texturing power still plays quite a role in graphics performance, but I do predict that once you have ~64 TMUs (750Mhz quantified) you've pretty much reached the peak.


I'm pretty confident in ATI's next chip on 40nm. It's not a performance evolution per se, but from a silicon budget perspective it should be quite crazy, making even the RV770 look bloated (let's not even mention GT200 here).

[ColonelCain]

ONly half of the claimed parts are SP cores...the bottom half. It's not 900SP, hence me laughing about this being news earlier.

You sound like you are very well versed in the topic of GPU uarch. Could you please elaborate? I'm very curious. :)

[cadaveca]

LoL. ME? Well versed? :rofl: I may sound like I know what I'm talking about, but please take ALL MY POSTS with a grain of salt...


Really....Thanks:yawn:, but I don't know as much as my posts seem to relate. I mean, sure, I've been following gpus for years, and it's the silicon that interests me really, and what made me an ATI fanboy, as ATI has had "new" tech in every gpu, since R300 was designed for M$ to test DX9 on...but I only know what is related by the company itself, and I listen to investor meetings and such stuff, as sometimes little tidbits fall through....but I know NOTHING 100%....probably not even 50%...

Anyway, given that I know that the sideport is a pci-e port(next to the other pci-e port, STEvil), and given that although SPs are right there, so are other bits that need some redundancy as well, and understanding that a bit of non-active silicon right in the middle of the gpu would help deal with heat issues, only so many things are possible.

Plus add in the XT naming that came with this news, and knowing who used that nomenclature in the past, I think I know who sourced the info, and I know they'd be wrong.