GT200 is Geforce GTX 280

And Geforce GTX 260


Nvidia has dared to make a drastic move in its naming convention. According to several sources close to the company, the new flagship chip codenamed GT200 and D10U-30 and D10U-20 will end up with Geforce GTX 280 brand for the faster card.

There will be one more branded as Geforce GTX 260 and this would be the one that many believed will be branded as Geforce 9900GT.

The launch is scheduled for the 25th week of the year and it should take place between the 16th and 20th of June. Roy Tailor the Vice President of "The Way It's Meant to Be Played" said that Nvidia’s names got a bit too complicated.

We remember that John Byrne, Vice president of Graphics, said the same as ATI lost its XT and PRO suffixes in its names. We believe that Nvidia learned from ATI that this actually did help and therefore they are doing the same; the bigger number means better, at least in their three digit brands, but we are still stuck with the suffixes for the time beeing.

NVIDIA declara la guerra visual a INTEL

Una misiva de un tipo llamado Roy Taylor dirigida al prestigioso sitio Web theinquirer.net, les resultó sumamente intrigante no tanto por su contenido sino por las ideas de lo que depára el futuro para los ya en pugna NVIDIA e INTEL, en la Guerra por dar forma a la próxima era en computación Visual

Éste autentico Nvidian boy, Roy Taylor retrucó declarando que el CPU es en realidad el que está muerto, y los chips de Nvidia son los que van a hacer todo el trabajo por nosotros en nuestras PCs.

Para respaldar sus afirmaciones, citó casi la totalidad de un artículo publicado en TGDaily, escrito por Theo Valich, el cual se puede encontar aquí.

Parte de éste polémico artículo dice:

"...En caso de que se pregunten, no, el CEO de Nvidia no ha emitido explicaciones sobre los resultados de la encuesta de hardware de Windows que culpó al driver nv4_displ.dll casi una tercera parte de BSODs en Windows Vista (Google revela alrededor de 613,000 resultados de ´Nvidia BSOD´... "


La siguiente carta se reproduce exactamente:

From: Roy Taylor [mailto:RTaylor@nvidia.com] De: Roy Taylor [mailto: RTaylor@nvidia.com]
Sent: 10 April 2008 23:36 Enviado: 10 de abril de 2008 23:36
Subject: The best job in the world. Asunto: El mejor trabajo en el mundo.

Chicos tengo el mejor trabajo del mundo. Es Oficial. No puedo contarles cuanto me divierto en mi trabajo ahora mismo.

No sé cuánto esto significa para ti lo que te voy a decir, pero para aquellos que no saben, acaba de empezar una guerra, sobre la que se escribirá muchísimo durante los próximos años y que afectará a todos el que sea propietario de una PC. A Todos.


Básicamente, el CPU está muerto. Sí, el procesador que Intel anuncia por todas partes. Está agotado. El hecho es que ya no se lo puede hacer correr más rápido. Ya no necesitas uno más rápido. Es por ello que AMD está en problemas y por qué Intel está en pánico. Tienen tanto miedo que han comenzado a atacarnos. Esto es porque todavía necesitas un chip que vaya más y más rápido, y ése es el GPU. El chip GeForce. Sí honestamente. Y no digo esto sólo para venderles la idea. Lo que estoy diciendo es muy importante.

Hoy en día el PC reproduce vídeo (tu GPU lo hace) Juegas? (el GPU lo hace), ripear películas (tu GPU lo hace) entendés el punto?

Hoy le devolvimos el golpe de nuevo a Intel, y esto es lo que la prensa está diciendo, pensé que les interesaría…

-------------------------------------------------------------------------------------------------------------
The visual computing clash: Nvidia CEO opens a can of whoop-ass for Intel
El choque en computación visual: El CEO de Nvidia abre una caja de Pandora

Negocios y Derecho
Por Theo Valich

Jueves, 10 de abril de 2008 17:12

--------------------------------------------------------------------------------------------------------------
Santa Clara (CA) - Nvidia e Intel se encuentran en un acelerado curso de colisión: Con Nvidia moviendo sus GPUs y todo su potencial en el territorio de las CPU; e Intel modificando sus CPUs para tomar el territorio de las GPUs, tienes entonces el clásico escenario de enfrentamiento entre dos gigantes de la industria que tienen el mismo objetivo - dar forma a la era de la informática visual.

Nvidia’s chief executive officer Jen-Hsun Huang today lashed out at recent Intel announcements and claims that indicated how the company wants to build up its graphics front line.


Nvidia's CEO Jen-Hsun Huang arremetió hoy contra los recientes anuncios de Intel respecto a l aforma en que la empresa quiere construir su primera línea de gráficos. Huang eligió palabras bastante fuertes, para golpear bajo a Intel, respecto a la oferta en graficos discretos y lo llamó "Laughabee".

Huang, conocido por su gran pasión por la compañía que fundó, al parecer, se ha visto afectado por est tema. En el momento en que debía abrir el Dia de Análisis financiero de la compañía, expresó "Nvidia is a Visual Computing company, not a semiconductor corporation" "Nvidia es una empresa de Cómputo Visual, no una empresa de semiconductores" y que su meta no es otra cosa que "mejorar GPUs y entregar una mejor experiencia a sus consumidores". Pero dicha apertura en su contenido rápidamente se trasladó cambiando el tema, cuando comparó el rendimiento de Intel dentro de su plan de trabajo para la primavera de 2008, con los productos actuales de Nvidia.

"Intel es una farsa. Han cruzado la línea, están diciendo cosas falsas". "Ellos dicen", declaró Huang, "Que Nvidia está muerto. Que sus gráficos son buenos, y que estarán en cada CPU, cuando eso en realidad no tiene sentido".

Él siguió comparando los actuales Intel Core 2, con la próxima plataforma en generación de procesadores y dijo que sería "nada más que poner más transistores en lugar de pensar de una solución."

"La gente no compra los productos Nvidia porque pueden hacerlo. Compran nuestras cosas porque quieren hacerlo, y los desean, y los dejamos abrumados por el valor de nuestros productos y el beneficio que aportamos", señaló Huang.

"Nuestro equipo [Nvidia] es como Ferrari. Sabemos cómo traer a la realidad tecnología visualm, en un 20-30-40x de desempeño, y una ventaja 27x en relación precio/rendimiento". Incluso si Intel es capaz de entregar 10 veces un aumento en rendimiento, aún no serían capaces de llegar a ponerse al día con AMD y Nvidia dentro del campo de gráficos discretos", dijo Huang.

Jen-Hsun también comentó sobre el artículo de Jon Peddie mostrando los últimos diez años del mercado de gráficos, publicado recientemente en TG Daily, afirmando que Nvidia pasó por una gran cantidad de competidores y considera que Intel sólo uno más de ellos.

Intel Larrabee llamado por Huang como "Laughabee". Gran parte del rendimiento que ofrecerá esa tarjeta, de hecho, dependerá de la calidad de los drivers para DirectX y el API OpenGL.

Huang puso abiertamente en duda que Intel pueda cumplir con el desarrollo de esos drivers, sobretodo por su estado actual de incompatibilidad, así como hay que tener en cuenta que los gráficos integrados de Intel no han dado grandes resultados en pruebas de Microsoft DCT, y la mayoría de las cuestiones fueran declinadas a los laboratorios WHQL debido a la falta de soporte para el hardware, lo cual no quiere decir que sean pura basura (en referencia a que Nvidia tienen los mismos problemas con la serie de productos GeForce FX y 6 / 7).

Durante las últimas semanas, numerosos representantes de Intel han estado hablando de las ideas para la informática visual de Intel - comenzando por Paul Otellini en la presentación de la firma en su Día de Análisis, y Pat Gelsinger tuvo en la presentación la exposición más agresiva denro de la FDI. Todo en relación a la integración de gráficos en el CPU Nehalem lo cual está creando mucho alboroto.

Estábamos dispuestos a dar Intel el beneficio de la duda sobre su futuro papel, pero el hecho es que que sus actuales sistemas de gráficos integrados probablemente acabarán costándole a Microsoft miles de millones de dólares y una plataforma integrada de PC, que se cree empujará lentamente la corriente principal del Mercado de PC hacia el mercado de consolas.

Dada la cantidad de problemas que los gráficos integrados de Intel enfrentan hoy en día, incluyendo las críticas procedentes de los gurús de la industria como Tim Sweeney y John Carmack, puedes esperar que Nvidia tome como objetivo a Intel.

A pesar de que ciertamente se ve que Intel y Nvidia se dirigen a una confrontación bélica, talvez alguna dato podamos extraer de todo esto. Por ejemplo, fuimos contactados por Intel acerca de un reciente artículo en el que un ingeniero de Intel afirmó que la gente "probablemente" no necesitará tarjetas gráficas en el futuro.

En una declaración enviada a nosotros por correo electrónico, la compañía dijo: "Intel no está prediciendo que es el final de los gráficos discretos. La Ley de Moore ha permitido a Intel innovar e integrar. Como resultado, esperamos que nosotros y otros podamos integrar en el futuro capacidad de gráficos visuales directamente en nuestras CPUs, como así fue con los coprocesadores de que punto flotante, y otras otras funciones multimedia en el pasado. Sin embargo, no esperamos que esta integración vaya a eliminar el mercado para tarjetas high-end en gráficos discretos y el valor al consumidor que proporcionan."

Roy Taylor
VP Content Business Development (CBD) Relations
NVIDIA Corp,. NVIDIA Corp,. Cell +1 408 XXX XXXX Celular +1 408 XXX XXXX

GT200, G100, D10E, NV60: New chip from Nvidia

We learned that Nvidia's big 65nm chip is going to launch in Q3 or possibly even later. Nvidia is about to launch the Geforce 9800 GX2, its dual chip G92GTS card and after that it plans to launch a product codenamed GT200.

Someone calls this card the D10E (Desktop 10th generation Enthusiast) and it is likely that Nvidia might brand this product as Geforce 10.

The earliest we believe that Nvidia can launch its GT200 / G100 / D10E is in late Q3, but we wouldn't be surprised if it slips to early Q4 2008.

Ugliest video card ever!

Diamond dual slot 3850 512MB Ruby Edition

Look this pictures "guacala!!!":


Review: Fudzilla

9600GT delayed yet again

NVIDIA SEEMS TO be hitting on all cylinders recently, and the streak continues.

The delay to the 9600GT we told you about two weeks ago has been added to. It is now a week after the week delay, so the 21st is now the 28th.

Nvidia GeForce 9800GTX, 9800GT and 9600GT availability and specs

Current reports from the web are claiming that we will in fact see a similar move from Nvidia that we already saw with ATI and it's HD3800 series.
That is saying they will rebrand the 8800GTS as the 9800GTX and the 8800GT as the 9800GT. What is unknown is if they will beef up the clocks or memory speed and size in the stock versions of the cards, compared to the 8800 series. This stems from the expected hardware capabilities of the 9800GX2, that doesn't allow them to place the G100 as the 9800GTX even if they could deliver it in time for the supposed launch date for these cards.

Conveniently, from a marketing viewpoint, they also place some extra motivation for unaware buyers, since they now have also something "new" from Nvidia, that competes more favorably with AMD/ATI. They have recently launched the HD3800 series of cards, which are, performance-wise, no more than softly tweaked HD2900s with DX10.1 and PCIe 2.0; the G92 chip doesn't even pack the DX10.1 capabilities though.

Expected availability is February or March for the 9800GTX and March or April for the 9800GT. The 9600GT will come in the previously disclosed date of February 14th and will fit between the 8600GTS and 8800GT in performance and price.
The 9800GX2 is expected to arrive by the same time the 9800GTX does.

9800GTX Known Specs:
- 65nm process technology at TSMC.
- Over one billion transistors.
- Second Generation Unified Shader Architecture.
- Double precsion support (FP64).
- GPGPU native.
- Over one TeraFLOPS of shader processing power.
- MADD+ADD configuration for the shader untis (2+1 FLOPS=3 FLOPS per ALU)
- Fully Scalar design.
- 512-bit memory interface.
- 1024MB GDDR4 graphics memory.
- DirectX 10.1 support.
- OpenGL 3.0 Support.
- eDRAM die for "FREE 4xAA".
- built in Audio Chip.
- built in tesselation unit (in the graphics core"
- Improved AA and AF quality levels

NVIDIA 9800 GX2 & 9600 GT

HardOCP today leaked an NVIDIA GPU Roadmap Outline for 2008, how true or how false this is we'll learn soon. But it's really weird to see HardOCP doing a thing like this. They are not the usual suspects when it comes to leaking out NDA information.

Details aren't quite hard and fast, but this is NVIDIA's top-of-the-line to-be, so it's never too early to start nerding out. HardOCP says the best way to think of the GeForce 9800 GX2 is a 8800 that's been shrunk down to 65nm and SLI'd onto a "single" card. The card is supposed to be at least 30% faster than a 8800 Ultra, and will apparently support Quad SLI. So:

* 1GB frame buffer
* Two PCBs
* Two 65nm GPUs
* 256 Stream Processors

That's the way HardOCP tells it, and who are we to argue?

In other news NVIDIA's newest mid-range processor, codenamed D9M, will make its official debut as the GeForce 9600 GT. Corporate guidance from NVIDIA lists the initial GeForce 9600 GT shipments come stock with a 650 MHz core clock and a 1625 MHz unified shader clock. Unlike the G84 core found on GeForce 8600 GT, D9M will feature a 256-bit memory bus interface. Coupled with a 900 MHz memory clock, NVIDIA calculates the memory bandwidth at 57.6 GB/s.

The texture fill rate is estimated at 20.8 billion pixels per second. The company would not indicate how many shaders or stream processors reside on the D9M core. Here's HardOCPs take:



We serve up a glimpse into the NVIDIA GeForce GPU cycle for early 2008. Some will laugh, some will cry, and some will wonder “WTF?”

NVIDIA Roadmap Outline for 1H08

While this is in no way complete, it should help our readers understand the direction NVIDIA is taking their video cards in early 2008. Some of this is sketchy and we will surely have follow up articles on the subject.

While a continued GPU die shrink to 65nm is certainly welcome, many of our readers are going to be disappointed by the fact that no next-gen technology is on the outlook as of yet. It has been over a year since the 8800 family of GPU was launched.

* GeForce 8800 Ultra will be replaced by the GeForce 9800 GX2 in February / March timeframe. More information and pictures are here.

* The GeForce 8800 GTX will be replaced by the GeForce 9800 GTX in February / March timeframe. This card will support Tri-SLI.

* The GeForce 9800 GT should appear in the March / April timeframe. We have limited information on this card currently.



* The recently “released” GeForce 8800 GS will be a limited GPU in terms of production. Do not expect more than 100,000 GPUs to be shipped worldwide, but soon. ASUS will supply Asia, Palit will supply China, EVGA will supply North America, and XFX will supply Europe. The GS is an “inventory solution.” The 8800 GS is 192-bit bus and will ship in 320MB, 512MB, and 640MB versions with a trimmed down number of stream processors as well. Will fall in line below 8800 GT but above the 9600 GT.

* The GeForce 9600 GT will fall in line below the 8800 GT, but give better performance than the GeForce 8600 GTS. The 9600 GT will be a whole new card not based on the 8800 GT PCB. Price point should be at $169 in retail/etail and plans are to carry this GPU throughout 2008.

So as it looks right now, we should not expect anything out of NVIDIA in terms of next-gen technology at least until mid-2008. Don’t be confused by the new “98XX” model numbers as they don’t signify much more than the die shrink to 65nm. You might agree or disagree with this naming scheme, but the entire NVIDIA card market is getting confusing and this might at least help things be a bit not-as-confusing to consumers looking for a newer product, but most likely is being done for the big system builders needing “new” specs for new system builds.

My feeling is that NVIDIA is holding back their true next-gen technology (if they actually have it working now) for the AMD R700 release that we could see around mid-2008, if not sooner.

As for the recent rumors on Intel purchasing or merging with NVIDIA, well, we think that is a bunch of BS.

We do realize this “roadmap” is far from complete. When we have new information we will be sharing it.

GeForce 9600 GT Specifications Surface

The rumor mill is is spinning hard at expreview today, as it seems NVIDIA will debut it's Geforce 9 series on valentine’s day, 2008. According to sources the first product of Geforce 9 series will be GeForce 9600 GT.

9600GT’s GPU is a 65nm processing chip. We already seem this fabrication process used on on G92/D8P (8800 GT / GTS 512MB) . The nex-gen mid-range product GPU should have core clocks at 500MHz with it's memory at 2000MHz. The GPU will have 64 Shader Processors (a 8800GS will be 96SP 192bit 384MB). It is a main-stream product and only have one SLI connector so it can not support 3-Way SLI.

GeForce 9600 GT uses a P545 PCB design and the memory will be 256-bit 512MB.

Judging from these early specs it's a pretty save bet to say that the product’s proformance will be lower than the 8800GT, and likely even its rival Radeon HD 3850/3870.

8800GT 1GB STFU Edition emerges

It´s code name is: "SF-PX88GT1024D3-HP Cool-pipe 3"

Taipei, Taiwan – December 21, 2007 - SPARKLE Computer Co., Ltd., the professional VGA card manufacturer and supplier, today announced the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card, the World's first graphic cards based on NVIDIA GeForce 8800 GT GPU with 1GB video memory and passive cooling solution, The SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card will bring Ultra HD gaming experience to high-end gamers.



Based on NVIDIA’s next generation GeForce 8800 GT architecture, the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card adopts innovative unified architecture, dynamically allocates processing power to geometry, vertex, physics, or pixel shading operations, delivering up to times the gaming performance of prior generation GPUs. Built upon technologies such as NVIDIA Lumenex Engine, providing support for DirectX 10 Shader Model 4.0, NVIDIA Quantum Effects technology for physics computation and GigaThread Technology for extreme processing efficiency in advanced, next generation shader programs.

Specifically developed to answer enthusiast gamers’ demand for Ultra HD high-performing graphics, the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card is not only powered by most advanced 65nm GeForce 8800 series GPU from NVIDIA, but also combined with formidable R&D strength from SPARKLE, which brings inconceivable 1GB GDDR3 video memory on independent developed solution. Now the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card can provide unbelievable video memory bandwidth for latest DirectX 10 games such as Crysis and so on, to reveal gorgeous gaming texture at 1920x1200 Ultra HD resolution with fluent gaming speed.

The SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card possesses 600MHz core speed, 112 stream processors with 1500MHz shader clock and 1GB 1800MHz 256-bit GDDR3 video memory. GeForce 8800 series rendering architecture, DirectX 10 shader model 4.0 support, Ultra High speed stream processors and super capacity 1GB video memory, all these luxuriant features endow the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card with brutal 3D horsepower, to incisively and vividly present most cinematic gaming scenes to the gamers. In addition to all these incredible features, the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card is designed for the new PCI Express 2.0 bus standard and are backwards compatible with the original PCI Express standard. So they are full ready for future games and applications which have rigorous requirements on bandwidth this video hard



65nm process lets SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 graphics card to bring less heat volatilization than 90nm and 80nm products on market. Further more, SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 graphics card is is equipped with heat-pipe-based passive cooling system using high technology. Compared with conventional HSF, this noiseless passive cooling solution features higher throughput thanks to high efficiency heat-pipe optimized for mass heat volatilization. It has high performance thermal compound ensures optimal thermal dissipation even after years of use.

The SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card also marries the biggest video memory with the best-in-HD. They leverage NVIDIA’s revolutionary second-generation video processing engine to deliver high-quality playback of HD DVD and Blu-ray movies on high-end PCs. The new programmable video processing engine takes on all of the high definition H.264 video decoding, freeing the CPU to perform other tasks and significantly reducing power consumption, heat, and noise.

“With powerful 65nm GeForce 8800 series GPU, the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card offers a hyper-reactive, intensely robust gaming experience for the latest next-generation DirectX 10 games, such as Crysis, Gears of War, Bioshock, as well as full support for current DirectX 9 games." Said Precilla Wu, sales manager of SPARKLE Computer Co., Ltd. ” SPARKLE decided to raise the bar and deliver GeForce 8800 series graphics cards with 1GB video memory, giving our customers incredible confidence to handle gorgeous DX10 gaming scenes which have rigorous demand on video bandwidth. Also with high efficient heat pipe passive cooling system, the SPARKLE GeForce 8800 GT 1GB Cool-pipe 3 Graphics Card best choices for today's Ultra HD gaming."

New GeForce 8800 GTS 512

NVIDIA’s G80 GPU has been sitting atop the 3D graphics food chain for well over a year now. It was way back in November of ’06 that the GeForce 8800 GTX and original 640MB GTS arrived – the first two graphics cards based on the G80 GPU. Sometime later, the more affordable 320MB GeForce 8800 GTS and current flagship GeForce 8800 Ultra arrived, but these two cards are based on the same GPU that powers NVIDIA’s initial 8800-series offerings.

Obviously, even though NVIDIA’s products still clearly outperform ATI’s at the high-end, it’s time for a refresh, if just to keep things exciting for consumers this holiday season. The first glimpse of what NVIDIA had in store came a little over a month ago in the form of the GeForce 8800 GT. The GeForce 8800 GT was based on a brand new GPU, internally codenamed the G92. G80 and G92 are fundamentally very similar, but NVIDIA hadn’t unleashed the full potential of the GPU due to the odd number of stream processor partitions enabled in the GT - seven.

And today, NVIDIA is launching yet another addition to the GeForce 8800 series, a new GTS card featuring 512MB of frame buffer memory and a G92 at its heart, with get this, 128 stream processors. There was an eighth stream processor partition lurking within the G92 all along...

Fabrication: 65nm
Number of Transistors: 754 Million
Core Clock (Includes dispatch, texture units and ROPs): 650MHz
Shader Clock (Stream Processors): 1.625GHz
Stream Processors: 128
Memory Clock: 970MHz (1940 DDR)
Memory Interface: 256-bit
Memory Bandwidth: 62 GB/ps
Frame Buffer Size: 512 MB
ROPs: 16
HDCP Support: Yes
HDMI Support: Yes
Connectors:
2xDual-Link DVI-I
7-Pin TV Out
RAMDACS: 400MHz
Bus Technology: PCI Express 2.0
Max Board Power: 150 Watts