one more try.. please R300 *only* and 2 Tmu's

It'd be interesting to see an R300 and NV30 clocked the same as a Parhelia and benched on a "next-gen" game such as UT2K3 or the like, just to see what speed differences the different architectures yield. Or perhaps on a very configurable engine like SS/2, if it's been patched to optimize for each VPU.

I mean, it's obvious more than clockspeed is hurting Parhelia. I'd like to see a game in which its 4x4 config shines.
 
alexsok said:
Which has to make you wonder... wtf was Matrox thinking???

Remember that this is Matrox's first exposure to the high end market, their first experience in this field.

So I guess the G400 wasnt aimed at the highe end....

or the g200....

at least get thier history right man.

They had bump mapping out with the g400, made me really want that card when I saw what the water could look like.
 
I'm amazed at those mentioning parhelia in the same breath as r300 or NV30. Don't get me wrong, Parhelia would be a good budget card if it were ever priced that way. ( No offense to Parhelia owners)
 
3dcgi said:
Parhelia should be able to execute longer shaders than GF4 and R8500 in fewer cycles. Also, it's possible that the extra texture units don't take up as much space as more pipelines would.

Exactly this makes me wonder, too.

How many transistors more would a 16x1 VPU need as opposed to a 8x2 VPU/GPU?

or, phrased differently: wouldn't it be better to just use "a few" more transistors and go directly with another pipeline instead of just adding a second "TMU" ?
 
Back
Top