Despite constant rumours that Nvidia is focusing on the potentially lucrative HPC industry with Fermi - rather than gaming - company spokesperson Bryan Del Rizzo dismisses such rumours as "unfounded" and "ludicrous".

"Gaming remains our bread and butter focus area. However, there are other opportunities for us to explore as the company grows, such as the HPC sector," explained Del Rizzo.

The reason why all details on Fermi so far points towards a focus on compute processing rather than gaming graphics is probably because the two events that have covered Fermi - GTC and SC '09 are both compute oriented. Del Rizzo insists Nvidia have still not revealed "a lot" of details about the gaming capabilities of Fermi.

More next page.

ASUS has adopted award-winning NVIDIA® ION™ graphics for many of its top-selling “Eee” netbooks and nettop PCs. All of the NVIDIA ION-based PCs and motherboards from ASUS fully support Windows 7, which takes advantage of graphics processors to accelerate a variety of applications, freeing the CPU to handle other tasks. The new ASUS products include:

• ASUS Eee PC 1201N: An incredibly fast netbook that combines NVIDIA ION graphics with a dual-core CPU to deliver a superior multimedia experience.
• ASUS EeeTop PC ET2002T: A stylish all-in-one PC featuring a 20-inch Touchscreen display and built-in DVD drive.
• ASUS EeeBox PC EB1012: The latest version of ASUS’s pioneering nettop delivers an incredible media experience with support for HD video, HDMI, and casual gaming.
• ASUS EeeBox PC EB1501: A full PC experience with built-in DVD burner, up to 4GB of memory, wireless connectivity and Windows 7 Home Premium.
• ASUS AT3N7A-I: A mini-ITX motherboard that is ideal for building a media centre, small form factor desktop, or home theater PC.

Following the release of ATI's flagship Hemlock, we first witnessed the first picture of a working Fermi based Geforce system, and now at SC '09, Nvidia have demonstrated Fermi going up against GT200. All in one day!

The demonstration is a N-body simulation running on CUDA with 20,480 body interactions in double precision. This demonstration was designed to show off Fermi's massive DP performance increase over GT200. Of course, being a supercomputing conference, these results are hardly relevent for gaming. The GT200 GPU was a Tesla C1060 and the Fermi was an Tesla 20 sample. The results? 3.5 fps for the C1060, 21.72 fps for Fermi.

More details next page.

Finally, we have a picture of what looks like real Fermi based Geforce product rendering graphics! Probably not a co-incidence it was revealed on the same day as AMD's Radeon HD 5970 release.

The picture, posted on Nvidia's Facebook page, is of a GF100 product running Unigine's Heaven DX11 benchmark. Visually, the card look similar to current GTX 260/275/285 products.

More next page.

Galaxy releases the new GeForce GT240 D5 and Galaxy GeForce GT240 D3 with 1GB/512MB. It is based on 40nm process with 96 stream processors, 128-bit memory interface, 512MB GDDR3/5 memory. The cards come with its own custom cooler to keep the GPU operating 10-15 degrees lower than the standard cooling system.

We all know it is constantly being delayed - but at the end of all those delays, what we were expecting was a stellar product from Nvidia. However, with today's press release, certain inconvenient details are revealed. Let's forget about the delays for now, and just consider the product itself.

The first Fermi GPU - GF100 - as we know for a while now, is a 3 billion transistor giant, taking a die size of around 500 mm². Compare this with the 2.15 billion transistor, 330 mm² Cypress on the same 40nm TSMC process, and you would be expecting a different class of product. Unfortunately, the details revealed today about cast an uncertain shadow over this basic assumption.

More details next page.

Nvidia have "paper launched" the HPC version of its next-generation Fermi architecture. Branded as Tesla 20 series, the HPC GPGPUs will release in Q2 2010. Like the Tesla 10 series based on GT200, Tesla 20 will feature single GPU and four GPU options.

The single GPU will come in to variants - C2050 and C2070, with upto 3GB and 6GB GDDR5 memory respectively (with ECC enabled, 2.6GB and 5.2GB). The prices? An enormous $2,499 and $3,999, respectively.

More next page.

A day before release, Fudzilla have finalized specifications and price for what is likely to be Nvidia's final release before Fermi. The product is the GT 240 - based on the 40nm GT215 shrink. It packs in 96 shader units and an unusually low clock speed for a product of it's kind of 550 MHz. Another evidence of Nvidia struggling with the 40nm process? Of course, the GT 240 releases many months late, and really, it would only be competitive had it released in 2008.

The GT 240 will ship with either a GDDR3 version or a GDDR5 version. Considering it is tied to a 128 bit memory interface, we can expect the GDDR5 version to be much faster. Performance wise, the faster GDDR5 version is expected to just edge out the $69 HD 4670, but fall much short of the 9800 GT or the HD 4770. Not particularly remarkable? But the worst part of the GT 240 is its price.

More next page.

Nvidia diversifying to the CPU business is an ancient rumour, which recurs with every generation or two. No surprises then, that a version of the same rumour - that Nvidia is developing an x86 CPU - has been doing rounds on the internet over the last week.

In an interview with CNET, Nvidia CEO Jen-Hsun Huang has outright denied the rumour, with a "No.". Mr. Huang goes on to say "Nvidia's strategy is very, very clear. I'm very straightforward about it. Right now, more than ever, we have to focus on visual and parallel computing".

More next page.

Fudo of Fudzilla, one of the few reporters calling for a November/December 2009 release (a rumour which has now been squashed), suggests Nvidia Fermi is only going to ship with an A3 revision.

Currently, we are at A2, which was rumoured to have been taped out in WW42, which was end of October 2009. What is baffling is how Fudzilla is speculating a January 2010 release date when we are still one silicon behind, and they themselves claimed in a previous article GPU retail availability takes 3-6 months from tape out.

More next page.