News Blog

Update at July 15, 3:00 a.m. PDT with additional information and corrections concerning the Intel-Nvidia dispute.

Nvidia said Monday that its multichip technology will be architected to work on Intel's upcoming Nehalem chip platform.

Nvidia SLI technology supports multiple graphics boards.

(Credit: Nvidia)

This announcement may help Nvidia to work around a standoff with Intel over whether Nvidia can make chipsets that work with Intel's next-generation Nehalem platform, due later this year. And also demonstrates that despite Nvidia CEO Jen-Hsun Huang' s rhetoric, Nvidia must cooperate with Intel in order to thrive.

Monday's announcement has no relation to separate licensing negotiations, according to sources familiar with the discussions. In other words, Nvidia is not announcing a chipset for Nehalem--which would require a license. It is simply a statement that Nvidia will support Nehalem with its nForce 200 Scalable Link Interconnect (SLI) chip.

However, some reports say Nvidia has reached an agreement with Intel to license Intel's Quick Path Interconnect (QPI) technology, paving the way for Nvidia to design chipsets for Nehalem.

The nForce 200 chip will work with Intel's "Bloomfield" line of Nehalem processors and the accompanying Intel chipset. SLI allows Nvidia to use multiple graphics boards in one system.

Upcoming SLI motherboards will use Nvidia nForce 200 SLI silicon, Intel Bloomfield processors, and Intel Tylersburg (X58) chipsets, Nvidia said in a statement.

"The nForce 200 SLI processor features patented SLI technology for graphics bandwidth management and multi-GPU peer-to-peer communications, both required to optimize graphics performance," Nvidia said. GPU stands for graphics processing unit.

Future systems "can be powered by one, two, or even three Nvidia GeForce GPUs, including the new...GeForce GTX 280 and GTX 260 GPUs," according to Nvidia.

Nvidia included statements from system suppliers in the Monday release. "It's great to see that Nvidia opted to enable SLI on the future Intel Bloomfield platform," said Rahul Sood, CTO Voodoo Business Unit, HP. "Make love not war I say...and Nvidia's (enabling) of Intel chipsets to support SLI will make our jobs much easier."

Nvidia claims that nForce 200 SLI silicon with Intel's new Bloomfield processor and Tylersburg chipset core logic chipset will deliver up to a 2.8X performance boost over traditional single graphics card platforms.

Motherboards and PC systems that will use the Nvidia nForce 200 SLI chip, Nvidia GeForce GPUs, and Nvidia SLI technology will be available from companies such as Acer, ASUS, Dell, Falcon Northwest, Legend, and Velocity Micro.

Nvidia has slashed the price of products with its newest GTX 260 and 280 graphics processors only a few weeks after it launched the chips, in response to stiffer competition from Advanced Micro Devices' ATI unit.

Nvidia said Sunday night that the GeForce GTX 280 is now available for $499 and the GTX 260 for $299. The high-end GTX 280 was originally $649, while the 260 was priced previously at $399. Both products were rolled out less than a month ago.

Nvidia's graphics boards are now more in line with ATI's newest offerings. At $299, the GTX 260 price now matches that of ATI's comparable HD 4870.

There's more to come from ATI too. Later this quarter, ATI is expected to launch the 4870 X2, which combines two chips on one board. This will be ATI's high-end offering for the enthusiast gaming market. The lower $499 price for Nvidia's high-end GTX 280 should bring it close to 4870 X2 pricing.

ATI appears to be faring well in this round of graphics chip competition, putting more pricing pressure than usual on Nvidia. Not only are its individual chips more competitive than previous generations, but its strategy of building smaller, lower-cost chips is paying off. Instead of building one large, expensive graphics processor as Nvidia does, ATI is building less power-hungry chips for the mid-range market, then ganging them together to boost performance for the high-end enthusiast market.

Advanced Micro Devices' bet on a new approach to graphics chip design appears to be paying off, according to analyst Jon Peddie. This could put AMD's ATI graphics chip unit on top again--or at least on equal footing with Nvidia, the graphics leader over the last few years.

Peddie heads Tiburon, Calif.-based Jon Peddie Research, which specializes in graphics chip market research.

Test reports on AMD's and Nvidia's newest graphics chips are pouring in. Both companies are racking up good scores. (See Diamond Viper Radeon HD 4850 review here.)

But beyond the day-to-day test scores, AMD's ATI graphics chip unit may be winning the longer strategic battle. ATI has gambled on a radically different strategy for its latest series of chips--the HD 4850, HD 4870, and upcoming dual-chip R700.

"(AMD) is starting in the middle of the market and scaling up. That's a break with tradition," said Peddie. "We always started at the very, very tippy-top and build the most powerful thing you could and then let it scale down over time."

But Peddie said this traditional approach just isn't practical anymore. "The chipsets keep getting larger and larger despite the fact that we were going to smaller and smaller (manufacturing) process nodes. The chips grew faster than the process nodes shrunk and the consequences of that is that the power consumption went up, the costs went up, and it got to the point where it's kind of impractical to continue along that way," he said.

In essence, AMD's ATI unit strategy is to build smaller, less power-hungry chips and then gang them up to get better performance. Nvidia's strategy has been to build one large, extremely fast--and extremely power hungry--chip.

For ATI, the execution of this chip-ganging strategy is the key. And this is where ATI appears to have been successful. "The inter-processor communications. Getting that to work has been the trick. This is what ATI has done. They've come up with this stellar way of doing inter-processor communications so they can in fact get the scaling," according to Peddie.

And there's more than meets the eye. ATI has also cut in half the number of bits in the memory interface, Peddie said: down to 256 bits while Nvidia has remained at 512. "That has the benefit to ATI of reducing a big hunk of the power consumption."

Peddie said in the past this kind of approach would have been suicidal because it would have decimated ATI's test scores. "The argument against this is that graphics performance is a function of memory," he said. "Typically you want wider and wider (bit width)."

But ATI has countered this by using the fastest memory standard available. "So to compensate for shrinking down the bit width, ATI has jumped to the next-generation in memory design called GDDR5. GDDR5 is approximately three times faster than GDDR3--which is what Nvidia is still using and what ATI uses on their smaller cards," according to Peddie.

"So with three times the speed but half the width, they end up with 1.5 times the processing capability with the memory."

"A very clever thing that they did but mind you it was a gamble that looks like it's going to pay off," he said.

ATI has more processing units than Nvidia inside its chip too. "The other thing is that ATI has 800 processors in their chip and Nvidia has 240. That has a processor count advantage," Peddie said.

Though it remains to be seen if this advantage is borne out in testing over time, he added. "Nvidia and ATI keep improving their drivers so they'll seesaw back and forth with their scores, almost from week to week."

But in the long run, Nvidia may be forced to adopt ATI's strategy to keep pace in these week-to-week battles. "If ATI is successful, as we expect that to be, then Nvidia will have no choice but to adopt (ATI's) approach, just out of practicality," Peddie said. "It just makes a whole lot of sense."

AMD-ATI's upcoming R700 (rumored to be called the 4870 X2) two-chip graphics board will be the ultimate test of this strategy.

"It's a new proprietary inter-processor communication technology. If they put these two chips on one board and it does scale properly, then they have pulled off a coup," he said.

"When you gang up graphics chips (using the traditional Scalable Link Interface or CrossFire technologies) they roll off pretty fast. ("Roll off" implies that performance doesn't scale up well.) "So when you put two boards in, you don't get twice the performance but you (only) get one and a half. You put four boards in and you (only) get about 1.7, 1.8. What ATI is saying is that with two chips using (their) proprietary inter-bus, they will get 1.8 (the performance) with two chips. If that's true, you can expect to see four of them giving you something around 2.5."

Getting 2.5 times the performance from four boards would be a masterstroke for ATI.

The previous ATI dual-chip solution was very different, Peddie said. "The HD 3870 X2 was not a proprietary bus but a CrossFire connection. The CrossFire connection and the SLI connection are at the very, very end of the pipeline. Not the most efficient place to do an inter-processor communication. That's one of the reasons ATI has abandoned it."

AMD's ATI unit is also better positioned than it was before in manufacturing. "Part of the reason that Nvidia has been ahead is that ATI has been suffering over the last three or four years with manufacturing problems. It's not that ATI didn't have a good chip, the problem was that ATI couldn't build enough of them."

This should change with the newest series of chips. "This (design)--so they say--will really go into high-volume production." Though he cautioned this still remains to be seen.

"The (new ATI chip) is a really efficient, tight design. They used to do this all the time but they kind of got off that trail. And now they're back on it."

The Linux Foundation is trying to push Nvidia to make its graphics drivers more accessible. The Foundation's beef: closed drivers make Linux look unstable to end users.

Though a statement issued Monday does not cite Nvidia by name, Linux Foundation Technical Advisory Board Chair James Bottomley cited Nvidia in a phone interview. "My intent is to point out the problems Nvidia has been causing themselves with their binary-only (drivers)," he said. "They are certainly one of the few companies sticking to a binary-only strategy." Binary-only means the drivers are essentially closed.

"We...consider any closed-source Linux kernel module or driver to be harmful and undesirable," the official statement begins. "Vendors that provide closed-source kernel modules force their customers to give up key Linux advantages or choose new vendors."

But Bottomley gets much more specific than this. "Their (Nvidia's) binary module is one of the top causes of kernel crashes, which makes Linux look bad," he said.

"Nvidia does a reasonable job of Q-and-A-ing (quality assurance) of a certain number of configurations but the problem is that their configurations (are) a lot less than what's actually out there on the market," Bottomley said.

In the past, Intel had been the target of open-source advocates, but the chipmaker is now a leading open-source code provider. And graphics-chip supplier ATI Technologies, acquired in 2006 by Advanced Micro Devices, is open source too, Bottomley said. He did, however, cite some outstanding problems with an ATI "FireGL" driver.

"It's basically a reflection of the fact that graphics is one of the most complex and most difficult areas of technology that sits in a computer nowadays," he added.

Nvidia says it provides a high-quality Linux driver. "Nvidia supports Linux, as well as the Linux community and has long been praised for the quality of the Nvidia Linux driver," Nvidia said in a response to an e-mail query.

But the graphics chip maker defends its binary-only policy. "Nvidia's fully featured Linux graphics driver is provided as binary-only because it contains intellectual property Nvidia wishes to protect, both in hardware and in software," according to Nvidia.

"We try to make things open source whenever it makes sense," Nvidia said. The company cited examples here and here.

"To assume that customers won't have access to open-source updates from Linux kernel.org if they use closed source modules is not correct," Nvidia said. "Nvidia's Linux graphics driver kernel module is structured so that all the code that is Linux-specific is provided in source code as a 'kernel interface layer.' When customers upgrade their kernel to get the latest from kernel.org, they have everything they need to rebuild (and even patch, if necessary) the Nvidia's driver's kernel interface layer."

See: Linux developers petition for open Linux kernel drivers and ZDNet report here.

Nvidia has posted a video that involves the new GTX 280 chip, overclocking, lots of liquid nitrogen, and the Nvidia labs. Oh, and no quad-core processor. Get the point?

Nvidia NForce 790i Ultra SLI motherboard

(Credit: Nvidia)

"A lot of people believe you need an Intel quad-core or Intel quad-core Extreme to build an extreme PC," says the post by "Steffee" on the Nvidia Web site. "Today I'm going to build a gaming PC using the Intel Core 2 Duo. That's duo. Got that? Duo, two cores."

I think the point the blogger is trying to make is that the test system has only has two cores, though I could be mistaken.

Here's an excerpt from the blog: "Think you need a quad-core CPU for an extreme gaming PC with impressive 3DMark Vantage numbers and gaming performance?...I overclocked and hyper-cooled an SLI gaming rig using two of our latest and greatest GPUs--GeForce GTX 280...and a sub-$200 CPU."

Intel, of course, has a different take on this. "Most of what people do today with their computers requires powerful processors. Examples of processor-intensive applications include: creating content, viewing/editing high definition video, using social media sites, office tools, downloading music, and editing photos," Intel said in a statement.

In the test, the blogger uses an NForce 790i Ultra SLI motherboard, Intel core 2 duo E8400, in addition to the GTX 280 graphics card.

With the single GTX 280 board, the score is a 3DMark Vantage "X4796."

Then the blogger (an Nvidia employee) takes a hike to the Nvidia lab. "Now we're going to take a field trip to Nvidia labs to do some serious overclocking," she says. After adding what seems to be prodigious amounts of liquid nitrogen and adding another GTX 280 SLI board, the 3DMark score jumps to "X10,282."

Intel could respond (which it hasn't) by saying that the E8400 is not a slow processor: it runs at 3.0GHZ and has 6MB of cache. And overclocked with liquid nitrogen, it would probably get some pretty good scores too. And then, of course, it might be simpler to just get a quad-core Extreme CPU.

In the test, the Nvida GTX 280 core clock was overclocked to 727MHz and the shader to 1458MHz. The core clock is normally 602MHz and the shader clock 1296MHz.

System specifications:
--2× NVIDIA GeForce GTX 280 graphics cards running in SLI nForce 790i Ultra SLI motherboard
--Intel Core 2 Duo E8400 CPU
--4 GB SLI-ready Corsair DDR3 memory
--PC Power and Cooling TurboCool 1200W power supply
--Windows Vista 32-bit operating system

Nvidia has announced it will release a graphics chip that matches Advanced Micro Device's manufacturing process--an advantage that the latter has often touted.

Board using Nvidia GeForce 9800 GTX+ chip

(Credit: Nvidia)

The Nvidia GeForce 9800 GTX+, announced Thursday, is made on a 55-nanometer process. Current Nvidia processors--including the just-released GTX 200 series--are made on a slightly "fatter" 65nm process. AMD has moved most of its graphics chips to a 55nm process.

Typically, the smaller the process, the more energy-efficient a processor is. Smaller processes also typically offer better performance. Tech Web site PC Perspective has photos showing that the chip package is smaller than a 65nm Nvidia part.

The GeForce 9800 GTX+ will be competitively priced. "It's an enthusiast category graphics product, but it's going to be offered at $229 when it hits retail next month," said Ken Brown, an Nvidia spokesperson. It is slated to hit retail shelves on July 14.

AMD introduced new Radeon HD 4850 and 4870 graphics chips on Monday. These will start at just less than $200 and range up to about $300, making the 9800 GTX+ a direct competitor.

(Update: PC Perspective had this to say about the HD 4850 versus the 9800 GTX+: "The AMD Radeon HD 4850 and the NVIDIA GeForce 9800 GTX+ appear to be nearly equally matched in price and performance." The HD 4850 is now being offered by resellers. Prices range roughly from $189 to $220.)

"We've come to expect these cunningly timed product announcements from Nvidia every time AMD announces a new Radeon graphics card, said CNET's Rich Brown in a post.

Specifications for the 9800 GTX+ include a core clock speed of 738MHz, a shader clock running at 1836MHz, and 512MB GDDR3 memory rated at 1100MHz. The prior-generation 9800 model had a core clock speed of 675MHz and a shader clock speed of 1688MHz.

The platform also supports adding a second or third 9800 GTX+ board using the Scalable Link Interface.

Support for game physics and transcoding is also key to the product. "One of the key differentiators is something we introduced with the GTX 260 and 280 (just released Monday) and is moving down through the rest of our product offerings--support for Physics and, with CUDA, it supports transcoding," Brown said.

CUDA is Nvidia's C language programming environment. Transcoding is the conversion of a movie, for example, from one format to another.

Brighter, crisper images are the goal for top Intel researchers in their work on future graphics technology.

I talked Wednesday with Intel's Jerry Bautista, the co-director of the Tera-scale computing research program, and Daniel Pohl, an Intel researcher. I focused mostly on a concept called ray tracing but also questioned them about Intel's upcoming Larrabee processor.

Reflections: ray tracing versus rasterized graphics

(Credit: Intel)

First, some background. Ray tracing--whether you agree or disagree about its viability--has been a fairly hot topic. It has been mentioned frequently by Intel over the last six months. An Intel blog titled "Real Time Ray-Tracing: The End of Rasterization?" and later comments by Intel executives that the company is looking at doing ray tracing on its processors set the stage for debate on the viability of ray tracing in mainstream gaming.

Ray tracing is a technique for rendering three-dimensional graphics using complex light interactions, allowing the creation of extremely detailed reflective surfaces, for example, with stunning photorealistic results.

In the future, ray tracing may compete with today's traditional raster-based graphics used in games running on Nvidia and AMD-ATI graphics processors. Intel claims ray tracing runs better on general-purpose processors, such as its Core 2 Quad processors, than on traditional graphics processors. Ray tracing may also run on future processors such as Larrabee.

Intel CEO Paul Otellini alluded to this at a Sanford C. Bernstein & Co. Strategic Decisions Conference last month. Asked who Intel's major future competitors are, Otellini responded, "In graphics, as we move up the food chain, we're bouncing into ATI via AMD and Nvidia more than we used to. And I don't expect that to abate anytime soon."

Flexing its muscle at online retailers.

(Credit: Nvidia)

If you've gone video card shopping online lately, you may have noticed that you certain retailers are now asking you to click through to your shopping cart to see a price for Nvidia-based 3D graphics cards. As H Enthusiast reported earlier this week, that newly inconvenient shopping experience is part of a calculated effort by Nvidia to regulate the advertised pricing of its 3D cards.

The gist is that Nvidia is conducting a test run of this pricing scheme, and if online retailers like NewEgg, Buy.com, and others don't comply, they will face a series of penalties, and ultimately they will be cut-off from Nvidia-based 3D cards after a certain number of infractions. This strategy, called Manufacturer Advertised Pricing, is not exactly price-fixing, but it's illegal in other countries, and, as H enthusiast reports, it's only recently legal in the U.S.

Compliance thus far seems spotty. Some cards listed at NewEgg obfuscate the price, others with the same chip do not. The same with TigerDirect. Amazon, Buy.com, and ZipZoomFly all list prices as normal.

Nvidia's reasons for implementing this strategy are unclear. H Enthusiast was not able to get anyone on the record, although the going theory seems to indicate that it's a way to regulate non-US certified board partners from undercutting the Nvidia-approved US competition.

With only some vendors complying with Nvidia's request, and only then partially, it seems that Nvidia's plan is only an occasional annoyance. Whether Nvidia makes the plan permanent, cracks down on stray listings, and forces other retailers to go along all remain to be seen.

UPDATE: On Monday, Advanced Micro Devices and Nvidia are launching graphics chips based on distinctly different design philosophies.

Nvidia chips are big and powerful

(Credit: Nvidia)

Nvidia's GTX 280 and GTX 260 are designed to deliver the biggest performance bang per chip. A so-called "monolithic" approach packs 1.4 billion transistors and 240 processing cores onto one piece of silicon.

(See Peter Glaskowsky's review of the GTX 280.)

AMD's modular approach tends toward less is more: smaller, less power-hungry chips that can be strung together to achieve higher performance. The company plans to implement this strategy with the HD 4850 and HD 4870 graphics processing units (GPUs) that are being introduced on Monday. (See "Notes" below.)

This design philosophy is based on the same multicore strategy now employed by AMD (and Intel) for the CPU, the main PC processor. Instead of making one chip bigger and faster, AMD boosts performance by stringing together multiple chips. This approach also allows a chipmaker to avoid the time and expense of designing separate processors for the midrange and very-high-end segments.

(Clarification: In the case of a CPU, a core refers to a "scalar" x86 processor core found, for example, in a quad-core Intel or AMD chip. The GPU core is a "parallel processing core." GPUs these days typically contain hundreds of these cores.)

"The beauty of this design is that it's scalable. You can put one or two (chips) on a board," said Matt Skynner, vice president of marketing at AMD's Graphics Products Group.

In the midrange segment ($200 to $300) AMD uses a single chip--for example, an HD 4870. At the high-end ($500 and above), it adds another chip to scale up to better performance. This dual-chip design--code-named the R700--will be marketed as the 4870 X2.

On the other hand, Nvidia says its emphasis on a single, very-high-performance chip is necessary to keep it out front. "At the high end, there is no prize for second place," Ujesh Desai, general manager for GeForce products at Nvidia, said in an interview with Nanotech: The Circuits Blog last month.

(Nvidia also offers multichip designs, but it puts each chip on a separate board, while AMD puts two chips on one board.)

And Nvidia is trying to raise the bar with GTX 280. "We're rendering about 3 million triangles per frame," Curtis Beason, an engineer at Nvidia, said last month at an event where Nvidia previewed the GTX 280 chip.

AMD targets smaller chips that can be strung together to get better performance. AMD chip on left, Nvidia chip on right.

(Credit: AMD)

"With (the previous-generation) GeForce 8800, what we achieved is a very photorealistic character. Very detailed skin. But it was a single character," Jason Paul, the GeForce product manager, said at the Nvidia event last month. "With GTX 200 what we're moving to is multiple highly realistic characters."

Nvidia is also boasting that a dual-core GTX 280 can convert a high-definition movie into iPod video format in 35 minutes, compared to about five hours for a quad-core CPU system with low-end integrated graphics.

Hewlett-Packard's Voodoo unit will be one of the first to adopt the new Nvidia GPUs. "We are excited to be one of the first companies in the world to offer the technology in the new Exhilaration Edition of the...HP Blackbird 002," said Rahul Sood, chief technology officer, HP Voodoo Business Unit.

Notes: updated 6/16, 12:10 PM:

--AMD 4800 series processors will be available starting next week at Besy Buy, according to AMD VP Rick Bergman, speaking Monday at an AMD event. "In just a little over a week from today. You'll be able to walk into a Best Buy and buy this chip (4800 series) on a graphics board for about $200. A teraflop for $200," Bergman said. He added that systems will also be available from Falcon Northwest, Velocity Micro, and ibuypower. "We're also introducing a system that can take four of these boards," he said. "That's almost five teraflops of performance in a personal computer."

--Both AMD and Nvidia say their GPUs can achieve about one teraflops (trillion floating point operations per second) of performance.

--Nvidia's GeForce GTX 280 will retail for $649 and be available on graphics boards starting Tuesday. The GeForce GTX 260 will be priced at $399, with availability slated for June 26.