Nvidia’s next generation CUDA architecture

Posted on 13/11/2009 by

1



Hey everyone,

I know that this statement would seem a bit “deja vu”, however, it has been a quite a while since i last posted a post (whats up with the word play 😮 ?). But i’ve been busy with university work (third year is not as easy as people perceive it to be!!). Anyway, since i had time this morning (finally), thought that it would be a good idea to write a review. More specifically, write a review about the NEXT generation CUDA architecture aka Code named FERMI made by everyone’s favorite graphic card manufacturer, NVIDIA!.

Note: I do not claim that I like Nvidia, I am, after all, an ATI fan.

Anyway, time to get started with the review. Firstly, to eliminate the possibilty of confusion and damage to my credibility, I’ll have to  explain what CUDA actually stands for. CUDA is an acronym for Compute Unified Device Architecture, in other words, it is a parallel computing  architecture that has been developed by Nvidia. One would say that it is the same engine that can located in most of Nvidia’s GPUs (Graphics Processing Units).

Now Nvidia has made the next generation CUDA, aka FERMI. One look at this beauty and you would actually think that it is the most powerful GPU computing architecture ever built (It is !!). The power of this machine is phenomenonal !!! Though, it would most useful to those who prefer multimedia creation, application development and game development. It is essentially designed for C++ and is also available with a Visual Studio development environment. This system makes has over 3 billion transistors built in and powers 512 CUDA cores.

I would like to explain a bit more about this, however, if I do, i would leave you all in a sense of utter confusion.
So here’s the link, do have a look at it.

http://www.nvidia.co.uk/object/fermi_architecture_uk.html

Peace out.

JaGENAW

Note: Here’s a picture

Code named FERMI

 

Advertisements
Posted in: Hardware Reviews