ATI RV870 no início de 2009 em 40nm e com 2000 shaders!?

por **RuiAndrés** Qui Dez 11, 2008 11:48 am

O rumor mais quente que está circulando pela a web dá conta de que a ATi já estaria acelerando os passos para a produção da geração seguinte de GPUs, com a adição da RV870!
O chip supostamente seria lançado no primeiro trimestre de 2009 e contaria com litografia em 40nm! Exatamente o mesmo processo de fabricação que a futura RV740 (versão atualizada da RV730 que sequer fora lançada).
A informação que é motivo de fortes discussões em diversos fóruns refere-se a quantidade de shaders que o RV870 terá, num total de 1000! Dando assim à R870 (2xRV870), nada mais nada menos do que 2000 shaders!

Fonte: http://www.adrenaline.com.br/noticia...000_shaders!?/

R800 e DX11 nos planos da AMD?

Mais rumores

Citaçao : NH

Enquanto NVIDIA fala em GPUs de 55nm para final de 2008 e começo de 2009, AMD já começa a se preparar para o ano que vem com o próximo salto evolucionário de fabricação, GPUs de 40/45nm, sendo o RV870 usará o mesmo nodo de uma versão RV740 também em 40nm que será lançada antes. Se alguém duvidou que 2000 Stream Processors seriam impossíveis para a solução da AMD, agora creio que não existem mais dúvidas para tal proeza.

Fonte: http://www.framebuffer.com.br/

MAIS AINDA:

http://en.hardspell.com/doc/showcont.asp?news_id=3768

960 SP-laden RV870 in the Works, and more

Sources at TSMC told Hardspell that the next graphics processor (GPU) by ATI/AMD, the RV870 in the works have specifications such as:

* 40nm or 45nm fab process
* 140 sq. mm die size
* 192 ALUs, 960 stream processors
* 256-bit GDDR5 memory controller

It is believed that this GPU could perform 1.2 times better than RV770, purely based on the paper-specs.

Another interesting news is that R800, unlike its predecessors the R700 and R680, might not be a dual-GPU card but the world's first dual-core GPU.

Fonte

It's been known for some time that RV870 would be a shrink of RV770. The node following 65/55nm would be 45/40nm, but we weren't sure if AMD would go to 45nm or directly to half-node 40nm. It's now clear that AMD will go directly to 40nm and aim for a performance per watt ratio twice that of RV770. RV870 has been code-named "Lil Dragon", which does indicate that this will be something small, yet something to fear.

The overall performance has been suggested to be somewhere around 20% better than RV770 with a significant drop in power consumption. This points to 960 shaders (192x5), 48 texture units, 40 ROPs and so forth. If this is the case, TDP of the RV870 should be around 150W at most with today's frequencies. This is just an estimate on our part as technical specifications are still hard to come by. We also presume AMD to stay at DirectX 10.1 for now.

http://www.nordichardware.com/news,7999.html

NOTICIAS MAIS RECENTES:

Afinal...https://2img.net/h/i114.photobucket.com/albums/n...870Concept.png

e
Concept -> 1 master chip and up to 4 slave chips.

The master chip is like a normal GPU... but with 4 SidePorts.
These SidePorts use the Rambus FlexIO Interface:
· Rambus FlexIO is capable of running from 400 MHz to 8 GHz.
· Contains only 12 lanes (5 lanes are inbound, 7 outbound)
· Theoretical peak I/O bandwidth of 76.8 GB @ 8 GHz (44.8GB out, 32GB in)
· Total bandwidth: 76.8GB/s x 5 SIMD Core (1 internal + 4 "slave chips") = 384GB/s

Q. Why use Rambus FlexIO Interface ?
A. Because AMD can use it (Rambus Signs Patent License Agreement with AMD) and because it's realy fast :

Quote:
"While Cell’s XDR interface offers over 2x the memory bandwidth of any PC-based microprocsesor, Cell’s FlexIO interface weighs in at 76.8GB/s - almost 10x the chip-to-chip bandwidth of AMD’s Athlon 64." - Anandtech

A picture is worth a thousand words -> HD 5870 Concept
And this is the cards for the other market segments -> HD 5000 Series

A slave chip, or SIMD Core, is made of 8 blocs, for a total of 640 SP (128 SP 4D+1) and 16 TMUs -> picture

HD 5450 -> 640 SP + 16 TMUs (1 master chip)
HD 5650 -> 1280 SP + 32 TMUs (1 master chip + 1 slave)
HD 5670 -> 1920 SP + 48 TMUs (1 master chip + 2 slaves)
HD 5850 -> 2560 SP + 64 TMUs (1 master chip + 3 slaves)
HD 5870 -> 3200 SP + 80 TMUs (1 master chip + 4 slaves)

· Need only 2 chips (master/slave) to create a complete series of graphic cards.
· The slave chip is an exact copy of the "SP/TMUs" part of the master chip.
So when you create the master chip, you also create 95% of the slave chip.
· Less time/money needed to create a new series -> 9 month product cycle instead of 12 months.
· Very small chips -> very good yields -> very good price.
· Shared memory architecture !

· The highest memory bandwidth hogs are the ROPs and the Level 2 texture caches, so ATI has physically placed these units next to the memory controllers and used a thousand or so traces to connect them. So, for now, it's impossible to put the RBEs inside the slave chips. Thus, the master chip has to contain enough RBEs for the HD 5870... but this high number of RBEs will be overkill for the HD 54xx/56xx. This is perhaps the major problem with this design.

Links:

http://www.rambus.com/us/news/press_releases/2006/060103.html
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2379&p=4
https://2img.net/h/i114.photobucket.com/albums/n273/Wirmish/ATi5870Concept.png
https://2img.net/h/i114.photobucket.com/albums/n273/Wirmish/ATi5000Series.png
https://2img.net/h/i114.photobucket.com/albums/n273/Wirmish/ATi5000SIMDCore.png

Noticias Frescas..

AMD has been offering 55nm graphics processors since the Radeon HD 3800 series launch in November 2007, and Nvidia hopped on that bandwagon this summer with cards like the GeForce 9800 GTX+. What's next? The two companies probably have a few more 55nm GPUs in the pipeline, but Fudzilla reports that they'll both begin the switch to 40nm in the second quarter of next year.

Fudzilla says it doesn't know the code name of the Nvidia part, but it knows that GPU will use TSMC'S 40nm process technology and show up in late May or early June. AMD will supposedly introduce a 40nm RV870 graphics processor in that same time frame, which could put the two companies in an interesting position. As Fudzilla points out, a near-simultaneous launch would give neither party much time to respond to the competition and tweak their GPUs accordingly. That could leave either AMD or Nvidia with a clearly superior product, and Nvidia may have a rough time if it ends up facing another card like the Radeon HD 4870.

As you might remember, the 4870 pretty much forced Nvidia had to slash prices and launch a new-and-improved version of its GeForce GTX 260. AMD does have the process technology edge this time, though, so the two firms may end up on a more even playing field next year.

Fonte:http://techreport.com/discussions.x/15611

-------

Qualquer update ou noticia que algum utilizador saiba por favor informar para proceder á actualizaçao do post principal..

Cumps.