A powerful low-end machine sounds like an contradictions in terms, but I stumbled across one during the research for my previous blog entry. First let me state that I have no preference for Gateway Desktops, never having owned one.

The Gateway desktop in question is the Gateway DX4200-UB001A Desktop Computer, which has a Quad Core AMD processor machine with a huge amount of memory and a 64-bit Windows operating system. To be more specific, the Gateway DX4200-UB001A has a 2.2GHz AMD Phenom X4 9550 processor, 6GB (yes that’s not a typo) of RAM, and a 640GB 7200 rpm Western Digital hard drive. The 6Gb of RAM would really be wasted if it was for the fact that the computer ships with 64-bit Windows Vista Home Premium (64-bit Operating Systems are able to make full use of 4GB or more of RAM). It also has an integrated 256MB ATI Radeon HD 3450 graphics card, which is actually one of the weaker components of the system. It is because of the graphics card, which may prove difficult to upgrade, that the system is fine for casual gamers but not for serious gamers.

Additionally the system has a 15-in-1 media card reader, HDMI and DVI output connectors, 6 USB ports, and 1 Firewire port.

One thing to note is that the system doesn’t include a monitor.

NVIDIA is a company name you may not recognize (unless you’re a gamer), but it’s managing to do some of the most interesting graphics chipset work out there for motherboards and video cards. Some of the brands include GeForce, GoForce, Quadro and nForce. Now, you should add supercomputing to the work that NVIDIA is putting its imprint on. Popular Science just recognized its CUDA with a “best of what’s new” award.

The Cuda is a C language software development kit that can help a programmer transform a “graphics processing unit” (GPU) into a supercomputer. These are the devices that scientists and researchers rely on for computing jobs that require massive amounts of calculations. Frequently, they involve parallel processing — where a number of CPUs are applied in tandem to the work.

With CUDA you can turn your notebook to that kind of computing work because it can get threads on NVIDIA GPUs to work together in solving a problem. The programming kit includes examples for performing matrix multiplication, image convolution, Monte-Carlo option pricing and plenty of other samples on topics about which I have no clue.

I don’t know how this computing power will manifest, but no doubt somebody will create an application that will enable ordinary mortals to restart hearts with a cell phone and remove calories from Thanksgiving dinner via the family room Wii.