CeBIT 2008

Guru3D events and tradeshows 19 Page 3 of 15 Published by

teaser

3 - NVIDIA

 

NVIDIA then. Originally I had scheduled a thirty minute appointment with team green, yet there was so much going on that I think after two hours I already missed 2-3 other appointments. NVIDIA has loads and loads of interesting stuff going on. Unfortunately most of it is under NDA (disclosed) disallowing me to talk about it.

I'll just highlight the items that already where published on the web and thus is nothing "that" new, I just can't (and will not) discuss in-depth details. 

A pretty new market for NVIDIA is the mobile segment, smartphones. Their new NVIDIA's APX 2500 Cellular Phone Technology is also understated, very interesting. I had a brief peek at this piece of equipment and from what I can tell, it's the new shiznit in smart phones for sure, though I'm not the expert on that. From the pictures and video its looking like the NVIDIA/Microsoft partnership will take the Windows Mobile platform farther than any other platform currently that you can think of. Think of a low scaled GPU as the core for the smart phone. All functionally it integrated onward to the end-user GUI, yet being power efficient. All kinds of GUI 3D rendering and effects such as lighting and reflections as we spot while using the device. Another feature is just mind-blowing as it can decode and encode high-definition video up to a resolution of 720p. You can watch it on the device, or as I experienced it as well, connected to a plasma screen, and watched a pretty okay looking HD movie. Early stages stuff, but massive impressive.

You already heard about it here at Guru3D, early staged but obviously NVIDIA is working hard on Hybrid solutions as well. Very simply put, after nForce 790 all their mainboards will get an integrated graphics core. In desktop mode you use that one, in 3D mode you switch to your secondary graphics card. Why would you do this ? Energy consumption as in desktop mode you can power down that big ass graphics card of yours. You can also manage lower specced SLI functionally depending on which graphics core is integrated, yet that's not really a big thing as far as I'm concerned. 0.1+0.1=0.2  But you want 1.0 performance in gaming right ? Though that assumption might be tad unfair as Guru3D is really driven to enthusiast gaming & hardware.

Right now NVIDIA is working really hard on the concept. Yet there still seem to be some significant drawbacks. We think that the first implementation of Hybrid SLI technology not only does not support multi-monitor output capability, but also will not support dual-link DVI output, which means that such a system cannot support 30 high-end displays with resolution of 2560x1600. That still remains to be seen though.

Physics; as you guys know NVIDIA recently acquired Ageia Technologies, the company behind the PhysX software and hardware components. The acquisition will give Nvidia a physics element for its CUDA parallel processing systems. Gaming wise NVIDIA is working really hard on drivers. Though I can think of a thousand implementations of CUDA versus Physics, the idea is to implement a PhysX layer in the drivers, probably run it through CUDA coded in C and then give you the opportunity to use the GPU for physics processing. From what NVIDIA could tell me, the original idea still stands. And follow me closely now as this would be great: compatibility starts at GeForce 8, yet you can very well imaging using a second graphics card, and assign the physics functions to that second card. And that would be wonderful, imagine upgrading to a new graphics card, and you end up for example with a 8800 GT in the future .. wouldn't that make an excellent physics unit ? I expect to see the first driver implementation in roughly 4 months, and let's just pray that games will indeed start to support the feature.

Mainboards, obviously nForce 790i is around the corner. nForce 790i will be NVIDIA's first true PCIe 2.0 chipset. The 780i which we reviewed earlier is basically a 680i chipset with a PCIe 2.0 bridge chip. nForce 790i will also be the first NVIDIA platform sporting DDR3, and we think that NVIDIA has gone all the way and added support for up to DDR3-2000 (and maybe higher)  with the flagship model. There should be a couple of new SLI tweaks on the board as well .. but more on that piece of technology real soon though when we'll review it in-depth for you guys.

Graphics cards .. surely the most interesting topic to cover here at Guru3D, obviously you guys every friggin single bit on both the new 9800 GX2 and "that other one" already. First up (pretty soon) will be that GeForce 9800 GX2. That GeForce 9800 GX2 is basically  two 8800 GTS cards that have 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 as well. This really is all I can tell you though.

That second card is currently being discussed everywhere on the web, a lot of sites are already showing everything about it. Photo-shoots, benchmarks ... we spotted it all already. Since there already is so much info out there on the web I decided to do the following, specs wise you will not hear a single but from me until the release. Alright guys, from here on we'll slide into the photo shoot, dozens of pages loaded with all kinds of goodies combined with some loose comments here and there. I hope you enjoy this article.

Before we start I'd like to add we are showing the pictorial in random manufacturer order. There's no preference as to which company is shown first.

** Update - we had some photo's of that "new" product. Yet decided at the request and out of respect of the manufacturer & NVIDIA to remove them.

Anyway: the article is so big that we'll split it up and divide coverage over a couple days. Here's the CeBIT pictorial Part #1

Share this content
Twitter Facebook Reddit WhatsApp Email Print