AMD HD 6950 and 6970 released

clear_lot

Journeyman
u two go to some vacations...

:lol:
man...
you are thinking we are nerds who are spending the new year hunting the net for more info...

:lol:


EDIT: errrmm, you are actually true. :oops::oops:
 
Last edited:

Liverpool_fan

Sami Hyypiä, LFC legend
The lineup begins with "Wimbledon", an ultra high-end mGPU. It has a 256-bit wide high-speed GDDR5 memory interface, 2-4 GB of dedicated memory, and 65W TDP. The DirectX 11 GPU will be about 25% faster than "Blackcomb", the Mobility HD 6000 series flagship. This is slated for Q2-2012. Next up is the high-end "Heathrow" mGPU, which has a 192-bit or 128-bit (selectable between variants) GDDR5 memory interface, 1-3 GB of dedicated memory, up to 45W TDP, and 30% higher performance compared to "Chelsea". This is slated for Q4-2011 (this should tell you that Radeon HD 7000 series will be in existence towards the end of 2011).

Going down, there's "Chelsea" itself, with its 128-bit GDDR5 memory interface, 1-2 GB memory, 20-30W TDP, performance 30% higher than "Whistler", production starting in Q4-2011. Lastly, there's "Thames". This mainstream mGPU will have 128-bit GDDR5 with option of GDDR3, 1 GB memory, 15-20W TDP, and 100% higher performance than "Seymour", Radeon HD 6000 series' mainstream mGPU. Production for this starts in Q4-2011, as well.
lol that's weird. What's next? Millwall and Longhorn?
 

aby geek

Cyborg Agent
and socket 2011/ivy bridgge is a die shrink of sandy bridge.

the true succesor of sandy bridge is haswell which comes arnd 2013.and haswells dieshrink will be rockwell.


hers a n edit: socket 2011 is sandy bridge e ,its 22nm die shrink is ivy bridge. and haswell will be the true 22nm architecture and its 16nm die shrink wil be rockwell.

and some news on gtx560/gtx475 :

*www.fudzilla.com/graphics/item/21488-geforce-gtx-560-comes-in-late-january

amd radeon 6990:

*www.fudzilla.com/graphics/item/21487-amd-to-have-hd-6990-dual-cards-in-february
 
Last edited:

vickybat

I am the night...I am...
They won't scale in multigpu setups like their 2gb counter parts. And GTX560 is far from defeat. It will have better tricks up its sleeve. Maybe amd sensed this.
 
Last edited:

bilallucky

Broken In
6950 is taking a new place- no one to compete!!!as 5xx gtx r revised version if gtx 4xx, (their alu,simd,sps pattern is almost same), i guess driver updates will not effect that much like 6900 driver updates.becoz, 6900 r totally different in VLIWs pattern...amd has a long way to go with their driver updates.....
 

vickybat

I am the night...I am...
Gtx 5xx and 4xx are not same. There are significant differences. FP16 texture filtering happens in one clock cycle, just as it does on the GF104 found in GeForce GTX 460, and not the two cycles endured by GF100. Nvidia also made improvements to GF110’s Z-culling efficiency. This means the GPU is smarter about discarding pixels that don’t need to be rendered in a scene (because they’re behind other objects), conserving memory bandwidth and compute muscle.

In case of radeon vliw4 in 69xx series has a special function unit cutoff and not a breakthrough in architectural improvement. Its nothin like 69xx series has a long way to go in driver updates.

The fermi architecture has the upper hand technically and upcoming gtx 560 will pose a serious threat to 68xx series and maybe even 6950.
 
J

Joker

Guest
^^ err...

GTX 5xx is just an improved/optimized Fermi gtx 4xx.

Radeon HD 6xxx is a new architecture just like Fermi gtx 4xx was. still not optimized...would get better with HD 7xxx.

i dont care about technical mumbojumbo.
In case of radeon vliw4 in 69xx series has a special function unit cutoff and not a breakthrough in architectural improvement.
Gtx 5xx and 4xx are not same. There are significant differences. FP16 texture filtering happens in one clock cycle, just as it does on the GF104 found in GeForce GTX 460, and not the two cycles endured by GF100.
and u mean this is an architectural improvement? :mrgreen: stop posting like a fanboy.

They won't scale in multigpu setups like their 2gb counter parts. And GTX560 is far from defeat. It will have better tricks up its sleeve. Maybe amd sensed this.
FUD. :mrgreen:
 
Last edited by a moderator:

vickybat

I am the night...I am...
@ joker

If radeon 7xxx will get better so will the upcoming kepler from nvidia. Of course the above points i stated are improvements from 4xx series. You being an AMD FANBOY will always deny this no matter what. If you don't care about technical mumbojumbo then don't argue please. Post something relevent to prove your points.

In case you don't know, framebuffer is directly proportional to FPS(frames per second) for a same gpu. In this case 69502gb>6950 1gb and this will reflect greater in a multi gpu setup because the graphics core has lesser memory to play around whilst rendering in higher resolutions. I meant that and it doesn't mean i'm a fanboy or something.

About your FUD remark, we will see once the 560 releases. Counting out fermi prove you being a silly fanboy not me.
 
Last edited:
J

Joker

Guest
You being an AMD FANBOY will always deny this no matter what.
*img151.imageshack.us/img151/8529/46306145.th.png
*img811.imageshack.us/img811/7369/72608869.th.png
:lol:
fanboi me? naah

If you don't care about technical mumbojumbo then don't argue please. Post something relevent to prove your points.
my whole point is about cooking up of things. u cooked up CUDA without realzing what it is... :p

GTX 5xx is a refined Fermi...fact you are bigging up gtx 5xx (yup they rock) but then when AMD took a different route with radeon hd 6xxx with their VLIW4 architecutre, u think it is not a architectural breakthrough which in fact gtx 5xx is even not. :lol: nvidia got it horribly wrong with gtx 4xx except gtx 460 and improved it in gtx 5xx. AMD has got it wrong in in hd 6xxx and will improve the VLIW4 in next generation.
 

aby geek

Cyborg Agent
well fudzilla like to call gtx 560 gtx 475 instead so vicky maybe right about 560 being faster than 6950 as it will be between 470 and 570.

but vicky everyone is trying to tell you that gtx 5 series are improved 4 series nothing more whereas amd are working hard to build something new from scratch.

the amd 6 series you see was not to come so early it was preponed coz the other family of their graphic cards had to be postponed as global foundaries moved to new fab process and the plants were to be upgraded for that.

we are just applauding amd for coming in early and very wel for a preponed product.
 
Last edited:
J

Joker

Guest
well fudzilla like to call gtx 560 gtx 475 instead so vicky maybe right about 560 being faster than 6950 as it will be between 470 and 570.
i actually want THIS to be right :mrgreen: GTX 560 should be faster.

but my whole point is him bigging up gtx 5xx as a new architecutral improvement etc. etc. over Fermi and smalling the new VLIW4 architechture of AMD and saying it is mere a unit cut off when he hardly has any idea about it.

he was bigging up CUDA before & it just got raped by Stream and quicksync.
 
Last edited by a moderator:
OP
topgear

topgear

Super Moderator
Staff member
look what I've found about GTX560 :

According to the grapevine, the chip is based on the same GF104 chip used in its predecessor, the GeForce GTX 460, but using the full allocation of 384 stream processors, rather than the 336 found in the GTX 460.

The clock speeds also represent a significant boost over its predecessor, and could help the GPU hit the performance levels it needs to seriously compete with the Radeon HD 6950.

Read More
 

vickybat

I am the night...I am...
i actually want THIS to be right :mrgreen: GTX 560 should be faster.

but my whole point is him bigging up gtx 5xx as a new architecutral improvement etc. etc. over Fermi and smalling the new VLIW4 architechture of AMD and saying it is mere a unit cut off when he hardly has any idea about it.

he was bigging up CUDA before & it just got raped by Stream and quicksync.

Fermi was an architectural improvement but was plaugued by heat issues. These were sorted in 5xx & along with these, so were the compute efficiency as i mentioned earlier.

In case of AMD, they realised their folly with vliw5 as the 5th sp was never used cause as the workloads went narrower . Actually in a unified shader architecture, vliw design is unecessary and more scaler architecture is deemed fit which utilises TLP( Thread level parallelism). Amd relies more on ILP (Instruction level parallelism)

"AMD’s architecture is heavily invested in Instruction Level Parallelism, that is having instructions in a single thread that have no dependencies on each other that can be executed in parallel. With VLIW5 the best case scenario is that 5 instructions can be scheduled together on every SPU every clock, a scenario that rarely happens. We’ve already touched on how in games AMD is seeing an average of 3.4, which is actually pretty good but still is under 80% efficient. Ultimately extracting ILP from a workload is hard, leading to a wide delta between the best and worst case scenarios." ( source within quotes-ANANDTECH).


*i56.tinypic.com/1zf323l.png

So amd discarded the 5th sp which computes scalar operations and allowed three of its 4 sp's to handle it one at at time i.e vertex and pixel shading.

Now i will not call it an architectural breakthrough but rather a fine tune to utilise its resources properly and the reduced space can be used to add more simd's. This is what makes caymans faster than cypress.

Now in their next design it has to be seen whether amd goes the vliw 4 way or takes a more contemporary scaler path like nvidia or maybe have some new tricks in their sleeve.

Now about CUDA, i never cooked up anything but was backed by facts. Cuda is stiil great and transcoding is not the only thing cuda utilises unlike quicksync. There are lots of other apps acclerated by CUDA and stream though getting better is nowhere near cuda yet.

Cuda is good and everybody will agree but in case you don't remember , i was backing physx & i still consider it good but as other members pointed, its not a make or break deal when purchasing a card. Performance matters most.
 

mohiuddin

Journeyman
@vickybat ,u may not a fan of nvidia,but any one reading these thread can tell that ur post are NVIDIA biased.
tell me thinking twice,bro isn't the gtx500 are the revision of gtx400?
bro, answer it as a neutral person,'will u tell adding z-cull ,fp16 filtering and a bit tesselation improvement, a major architectural improvement for shifting to a new-generation naming-scheme?'
and about amd,
they planned that vliw4 for their planned 28nm fab.as the schedule failed ,they where bound to stick with 40nm.And it is a totally different architecture to work with.they even wanted to clock 6970 at 930~mhz.which would be possible with 28nm fab.But, they failed in 40nm for high yeild failure.For that cause amd lagged in 6900 series.

--
NVIDIA, is doing great with their 5xx gtx, though it was expected with 4xxgtx.
gtx560 and amd HD6950 1gb r going to fight well only for us ,end users.no need to support any one of them biased.:)
 
Top Bottom