EVGA is a company that's pretty important in the world of computing-
they're responsible for producing both Nvidia graphics cards and Intel
motherboards. To challenge AMD Radeon's new dual core card, they
announced one of their own on Friday- the 2win Geforce GTX 46o. The short version of this story: it's shaping up to be a fine piece of hardware.
For one, it utilizes dual Fermi parallel processing cores. And...I should probably explain that term. "Fermi" is what Nvidia's decided to name one of its CUDA-or Complete Unified Device Architecture. This one launched in March 2010. To state it simply, it's a microarchitectural design for Nvidia graphics cards, one which is named for the physicist Emrico Fermi, one of the men responsible for nuclear power. I won't get into the specifics of CUDA's advantages over typical chip
architecture here. It's enough to say that the design allows the chip to work fairly quickly and efficiently, though like anything, it has its limitations.
Regardless, given the name, Nvidia was obviously pretty confident about what this design could do when they developed it. Looking at their new card, I can sort of see why.
Well, first of all, the fact that Nvidia's saying this new card will outperform what's currently their most powerful on the market; I think we can be rather optimistic about the hardware specifications, no? Not only that, they're going to be selling it at a lower price than the GTX 580. More powerful and cheaper too? Not too shabby.
That said, for those of you who want something more concrete, the EVGA's Specs are as follows:
- 700 mhz core clock speed, with a memory clock speed of 3600 mhz
- 672 parallel processing CUDA cores, Fermi design
- 2 GB of GDDR5 Memory
- DVI and Mini-HDMI Interfaces
- 230.4 gigabytes per second memory bandwidth
- Support for DirectX11
- PCI-E 2.0 bus
- Runs three to four simultaneous displays
- Air-cooled with three 80 mm fans
Pretty sweet, isn't it?
What does all of that mean?
Those of you who can actually understand the above without an explanation can skip this next part. As with my Intel article, I'll offer those of you who aren't techheads an explanation of the above terms- albeit brief. Still, it should be enough that you get a general idea.
The core clock speed and memory clock speed relate to how the processor interacts with the rest of the computer. The first relates to how many instructions the processor can execute in a typical cycle, the second-700 megahertz would typically mean 700 million. The second refers to how quickly data can be allocated into the processor's memory each cycle. Typically, a computer cycle is one nanosecond. GDDR5 memory is basically 'active' Random Access memory that's allocated
specifically for graphical computer operations. Random Access Memory is
memory which only 'holds' data when the computer's power is on. It's
used for running computer processes rather than storing data. DVI and Mini-HDMI stand for Digital Video Interface and High Definition
Multimedia Interface. They're basically two different types of video; I
won't get into the details here. Memory bandwidth is how much data, in bytes, can be processed in per second. Finally, the PCI bus is basically the physical interface by which components of a computer are attached to one another.
Catch all that?
Well, there you have it. It's powerful, it's not going to be terribly expensive (as graphics cards go, anyway) and it'll hit the market...well, we don't actually know yet, unfortunately. Nvidia and EVGA have both been rather vague regarding both the exact price point and the exact day it'll hit shelves.
Guess we'll just have to wait and see.