Nvidia is ready for its next-generation GPU launch. According to sources at graphics card makers, the company plans to launch its GeForce 9 series GPU after the Lunar New Year in February.
The first chip to rollout of in GeForce 9 family will be the D9E, a high-end product that adopts 65nm manufacturing. The new product will also support DirectX 10.1 and Shader Model 4.1, revealed the sources.
In addition to the D9E, Nvidia will roll out a mid-range GeForce 9 family product named D9P in June 2008. The new GPU will adopt 55nm processing, the sources pointed out.
Original Story: Monica Chen, Taipei; Emily Chuang, DIGITIMES [Friday 30 November 2007]
NVIDIA IS PLANNING to make lots of hay this Christmas, rolling out revisions to its mid-range GeForce 8 series and raking in the cash on the technology it has spent the last few years developing.
But one thing is noticeably absent from the firm's Christmas lineup - any signs of GeForce 9.
Earlier on this year, Nvidia's senior management were telling press in behind-the-scenes briefings that the plan was to stick to the GeForce 7 and 8 launch schedule - a new high-end part for Christmas, and mainstream spin-outs in Q1 and 2 the following year.
This hasn't happened for GeForce 9. Why? Well, because it hasn't had to. The various re-spins of GeForce 8 are plenty powerful enough to keep Nvidia at cruising speed through the holiday season, without pulling out the big guns - after all, the competition this year has been barely worth looking at so the green team has had the market to itself.
Well, now Digitimes reckons that GeForce 9 is on track for a February 2008 release date. It seems feasible, but there is nobody at Nvidia who will confirm this either way, not least because the company is notoriously wishy-washy when it comes to rumours.
As for gamers - there's no doubting that they'd love to see a GeForce 9 this year, if only to get Crysis running at a decent speed. But when there's not much competition, there's no pressure to get parts out of the door - so perhaps ATI better get its act together for all your sakes, lest it's 2009 before we see the next major upgrade to the GeForce series
Friday, November 30, 2007
Been fired over a negative game review
A SENIOR HACK working for the C|Net affiliated gamer's website Gamespot apparently has been fired for panning a game marketed by a major advertiser.
Reportedly Gamespot Editorial Director Jeff Gerstmann was summarily fired because game publisher Eidos didn't like his negative review of its first-person shooter game Kane & Lynch: Dead Men.
Gerstmann's text review of the game gave it a score of only 6 out of 10, but it seems his video review was even more direct, criticising it for, among other failings, "impossible to like" characters, a "lazy script" and excessive profanity.
Gamespot's review guidelines state it has never "altered our verdict about any game due to advertiser pressure." Well, that rings rather hollow if they're going fire reviewers who don't always write nice things about every game, doesn't it?
Kyle Orland at Joystiq fairly well nails the issue, and we couldn't agree more. He writes:
"Readers should fairly expect there to be an inviolable firewall between advertising and editorial in journalism, and game journalism (yes, that includes "just reviews") is no different. ... Giving a [software] publisher the power to fire a senior editor is a line no outlet should be willing to cross."
If Gerstmann was really fired due to advertiser pressure, why should gamers trust the integrity of Gamespot's game reviews ever again?
Reportedly Gamespot Editorial Director Jeff Gerstmann was summarily fired because game publisher Eidos didn't like his negative review of its first-person shooter game Kane & Lynch: Dead Men.
Gerstmann's text review of the game gave it a score of only 6 out of 10, but it seems his video review was even more direct, criticising it for, among other failings, "impossible to like" characters, a "lazy script" and excessive profanity.
Gamespot's review guidelines state it has never "altered our verdict about any game due to advertiser pressure." Well, that rings rather hollow if they're going fire reviewers who don't always write nice things about every game, doesn't it?
Kyle Orland at Joystiq fairly well nails the issue, and we couldn't agree more. He writes:
"Readers should fairly expect there to be an inviolable firewall between advertising and editorial in journalism, and game journalism (yes, that includes "just reviews") is no different. ... Giving a [software] publisher the power to fire a senior editor is a line no outlet should be willing to cross."
If Gerstmann was really fired due to advertiser pressure, why should gamers trust the integrity of Gamespot's game reviews ever again?
Thursday, November 29, 2007
Next-gen Radeons to be multi-core?
Modern graphics processors rely on extensive amounts of parallelism to get calculations done as quickly as possible, but those GPUs are still monolithic chips designed with a given number of stream processors, texture units, render back-ends, and the like, each depending on their performance grade. However, the folks at Fudzilla say they have it on good authority that AMD will challenge that paradigm with its next-generation graphics processor, code-named R700. According to Fudzilla, low-end, mid-range, and high-end R700 cards will all have GPUs with varying numbers of identical R700 cores. To determine speed grades, AMD will simply outfit higher-end cards with more R700 cores.
AMD's top-of-the-line R700 product, for instance, will supposedly have four or more R700 cores in one die and will be able to crunch nearly two trillion floating point operations per second, or teraFLOPS. By contrast, Fudzilla explains that the existing Radeon HD 3870 is in the 500 gigaFLOPS range. FLOPS don't tell the whole story, of course, but high-end R700 cards will be an order of magnitude faster than existing products if those numbers are even remotely accurate.
Interestingly, this rumor sounds similar to information that recently seeped out about Intel's Larrabee project. Larrabee is expected to be a discrete, game-worthy Intel graphics processor scheduled for the not-too-distant future, and an Intel presentation nabbed by Beyond3D in April suggests Larrabee products will be based on multiple, small "throughput cores." A diagram showed a chip based on ten of those cores with a shared pool of 4MB of cache.
AMD's top-of-the-line R700 product, for instance, will supposedly have four or more R700 cores in one die and will be able to crunch nearly two trillion floating point operations per second, or teraFLOPS. By contrast, Fudzilla explains that the existing Radeon HD 3870 is in the 500 gigaFLOPS range. FLOPS don't tell the whole story, of course, but high-end R700 cards will be an order of magnitude faster than existing products if those numbers are even remotely accurate.
Interestingly, this rumor sounds similar to information that recently seeped out about Intel's Larrabee project. Larrabee is expected to be a discrete, game-worthy Intel graphics processor scheduled for the not-too-distant future, and an Intel presentation nabbed by Beyond3D in April suggests Larrabee products will be based on multiple, small "throughput cores." A diagram showed a chip based on ten of those cores with a shared pool of 4MB of cache.
Tuesday, November 27, 2007
Nvidia delays GeForce 8800GTS
NVIDIA has delayed the release of the GeForce 8800GTS until December 11.
The new GeForce 8800GTS 512MB cards will replace the previous generation GeForce 8800GTS 640MB cards with a 65nm process GPU, supporting PCI Express 2.0, HDMI and PureVideo Gen2. The GPU will have 128 streaming processors, 970MHz GDDR3 memory frequency, 650MHz core frequency and a maximum TDP of 140W, according to the sources.
Furthermore; considering the G92's success as an overclocker we can expect manufacturers to release overclocked versions in a bid to boost demand.
The 8800GTS 512MB cards are expected to retail for around US$299-349
The new GeForce 8800GTS 512MB cards will replace the previous generation GeForce 8800GTS 640MB cards with a 65nm process GPU, supporting PCI Express 2.0, HDMI and PureVideo Gen2. The GPU will have 128 streaming processors, 970MHz GDDR3 memory frequency, 650MHz core frequency and a maximum TDP of 140W, according to the sources.
Furthermore; considering the G92's success as an overclocker we can expect manufacturers to release overclocked versions in a bid to boost demand.
The 8800GTS 512MB cards are expected to retail for around US$299-349
Crysis 2
Popular Crysis news site inCrysis has loosely translated an interview with Crytek's Steven Bender, where he reportedly suggested that there were plans for a sequel. Furthermore; the main focus for Crysis 2 will be to beef up the game's graphics:
According to him, the developer's attention will be even more focused on graphics. There will be numerous improvements regarding textures (especially facial textures) and 3D-animations. Their ambitious aim is to push the game's visuals to an advanced level, that could be considered near photorealistic movie quality.
Will the Crysis sequel satiate the thirst for visual eye candy? That we know: Far Cry 2 not.
According to him, the developer's attention will be even more focused on graphics. There will be numerous improvements regarding textures (especially facial textures) and 3D-animations. Their ambitious aim is to push the game's visuals to an advanced level, that could be considered near photorealistic movie quality.
Will the Crysis sequel satiate the thirst for visual eye candy? That we know: Far Cry 2 not.
Friday, November 9, 2007
AMD FireStream 9170: un procesador de 320 núcleos
AMD presentó ayer su chip FireStream 9170, un procesador de 660 millones de transistores y 320 núcleos. Es un paso más en el proyecto “AMD's Fusion Project”, que pretende combinar en un mismo chip el microprocesador principal y la unidad de gráficos.
El concepto de “stream computing” se basa en el uso masivo del paralelismo en los procesadores. La idea es simple: en lugar de un solo núcleo potente, caro y complejo, se utilizan muchos núcleos pequeños, baratos y sencillos. Esto hace que el chip funcione a menos velocidad y que por consiguiente consuma menos energía y no se caliente tanto. AMD ha empleado esta filosofía de diseño para dar forma al FireStream 9170.
Los microprocesadores empleados en las tarjetas de video (GPU, por Graphics Processing Unit, o Unidad de Procesamiento de Gráficos) actuales son, a menudo, más complejos y potentes que el procesador principal del ordenador. De hecho, son capaces de hacer mucho más que mover y generar pixeles en un videojuego. NVidia lo ha demostrado con su proyecto TESLA, que emplea uno de sus GPUs como procesadores principal.
De hecho, los primeros ordenadores concentraban las tareas de procesamiento y de generación de gráficos en un mismo chip. Con el paso del tiempo, las tarjetas especializadas en gráficos se volvieron populares, y en medio de una guerra por lograr el mayor rendimiento, empresas como AMD y NVidia desarrollaron una serie de chips cada vez más poderosos. Tanto, que ya es posible utilizar uno de estos microprocesadores como cerebro de un ordenador.
AMD ha reducido la distancia que separa a los microprocesadores y los chips gráficos de alto rendimiento. El FireStream 9170 no es ni más ni menos que uno de los chips gráficos más potentes de ATI al que se le ha añadido memoria y unidades de coma flotante de doble precisión. El resultado es una bomba que estará en la calle el primer trimestre del próximo año, a un precio de 1.999 dólares.
Entre las características más importantes del FireStream 9170 se cuentan sus 320 núcleos, unidad de coma flotante de doble precisión, 660 millones de transistores y un rendimiento de 500GFlops. AMD ha preparado una tarjeta que se inserta en un slot PCIe 2.0 x16, equipada con este procesador y 2GB de RAM GDDR3 on-board, que puede funcionar en ordenadores Windows XP, XP64, Linux 32 y Linux 64 para experimentar con el nuevo chip.
Pero AMD quiere integrar este tipo de tecnología directamente en un ordenador, reemplazando al procesador principal. Y pronto: su proyecto Fusion contempla esta posibilidad en un par de años.
El concepto de “stream computing” se basa en el uso masivo del paralelismo en los procesadores. La idea es simple: en lugar de un solo núcleo potente, caro y complejo, se utilizan muchos núcleos pequeños, baratos y sencillos. Esto hace que el chip funcione a menos velocidad y que por consiguiente consuma menos energía y no se caliente tanto. AMD ha empleado esta filosofía de diseño para dar forma al FireStream 9170.
Los microprocesadores empleados en las tarjetas de video (GPU, por Graphics Processing Unit, o Unidad de Procesamiento de Gráficos) actuales son, a menudo, más complejos y potentes que el procesador principal del ordenador. De hecho, son capaces de hacer mucho más que mover y generar pixeles en un videojuego. NVidia lo ha demostrado con su proyecto TESLA, que emplea uno de sus GPUs como procesadores principal.
De hecho, los primeros ordenadores concentraban las tareas de procesamiento y de generación de gráficos en un mismo chip. Con el paso del tiempo, las tarjetas especializadas en gráficos se volvieron populares, y en medio de una guerra por lograr el mayor rendimiento, empresas como AMD y NVidia desarrollaron una serie de chips cada vez más poderosos. Tanto, que ya es posible utilizar uno de estos microprocesadores como cerebro de un ordenador.
AMD ha reducido la distancia que separa a los microprocesadores y los chips gráficos de alto rendimiento. El FireStream 9170 no es ni más ni menos que uno de los chips gráficos más potentes de ATI al que se le ha añadido memoria y unidades de coma flotante de doble precisión. El resultado es una bomba que estará en la calle el primer trimestre del próximo año, a un precio de 1.999 dólares.
Entre las características más importantes del FireStream 9170 se cuentan sus 320 núcleos, unidad de coma flotante de doble precisión, 660 millones de transistores y un rendimiento de 500GFlops. AMD ha preparado una tarjeta que se inserta en un slot PCIe 2.0 x16, equipada con este procesador y 2GB de RAM GDDR3 on-board, que puede funcionar en ordenadores Windows XP, XP64, Linux 32 y Linux 64 para experimentar con el nuevo chip.
Pero AMD quiere integrar este tipo de tecnología directamente en un ordenador, reemplazando al procesador principal. Y pronto: su proyecto Fusion contempla esta posibilidad en un par de años.
Subscribe to:
Posts (Atom)
