During the past few months one product has been on every gamers
mind: Nvidia's GeForce 256 Graphics Accelerator or as they would
like you to call it "GPU". This chip and the
boards based on it have received more hype than any other graphics
product in recent memory. To find out if the GeForce 256 really
lives up to the hype we've gotten our hands on a fresh new Creative Annihilator
Pro card featuring 32 MB DDR SGRAM memory.
Why makes the
GeForce Different?
I know by now you've
probably read tons of stuff about the GeForce and all the
technological advancements it brings to the PC Graphics Area but I
just want to run through all that info really quick in a hopefully easier
to understand manner. The GeForce is different from any other
graphics chip out there with the exception of course of the Savage
2000 in that it accelerates the entire graphics pipeline or to put
it in simpler terms its the first of a new generation of chips that
can do pretty much all the calculations that go into displaying
those pretty pictures on the screen. Previous accelerators only did
calculations for triangles and rasterization but the GeForce accelerates
the geometry portion of the scene. Geometry is composed of
transformation and lighting (T&L). T&L operations are the
most complex of the entire 3D pipeline and until now they have been
the exclusive domain of you CPU. The GeForce changes all that by accelerating
T&L via its geometry engine thus the CPU is finally free to do
other cool stuff like physics calculations for realistic 3D models.
So what's the catch?
Well, in order for the GeForce to work its magic the application
must be programmed to support the T&L extensions. So far there
isn't really any titles out there that have this capability, but its
definitely the future of 3D gaming. With that said, if you buy a
GeForce your really buying it for what it will do down the road ,not
today but don't get me wrong the GeForce is one hell of a
contender for today's apps as you will see in out benchmarks
presentation.
DDR vs. SDR
DDR memory is a new type of memory
that is faster than conventional SDR memory. DDR memory transfers on
both the rising edge and the falling edge of the input clock, while
SDR memory transfers data only on the rising edge of the clock. This
effectively doubles the peak bandwidth of the memory without
requiring a wider interface. The picture below shows this
"doubling" of peak bandwidth.
I know all this sounds a little complex but essentially if you
buy a DDR equipped card your getting double the bandwidth thus, your
likely to find vendors such as Creative and Guillemot to
advertising their DDR SGRAM as running at 300 MHz even though
its actually running at 150 MHz. 150MHzx2 = 300MHz got it? Good. As
you will see later in the review DDR lives up to the hype and in my
opinion proves itself to be the graphics memory technology for the
21st Century. Now that we've removed
the confusion surrounding the technology behind DDR and SDR lets
take a look at the board itself.
PAGE
2: The Board
