new scalable graphics chip for PCs on the horizon: Mali200 from Falanx

[Guest]

http://www.beyond3d.com/forum/showthread.php?t=22310

http://www.hardocp.com/article.html?art=ODAy

http://www.falanx.no/


the Mali200 graphics chip is still a long way from reaching the market.
the design is highly scalable.

a late 2006 graphics card using a bunch of these small cores could be a
viable solution

quote from Hardocp:
" Mali200 employs a combination of tile-based and immediate mode rendering
for efficient bandwidth usage, which results in several competitive
advantages over using a single approach."

the best of PowerVR-style rendering plus the best of SGI-style rendering ?


we certainly need some competition in the graphics industry, currently
dominated by Nvidia and ATI.

 

[Kevin C.]

the Mali200 graphics chip is still a long way from reaching the market.
the design is highly scalable.

a late 2006 graphics card using a bunch of these small cores could be a
viable solution

The scalability of the core is irrelevant since the end user product has a
fixed profile. Without a cost-performance analysis the whole thing is
meaningless. The Voodoo cores were also scalable so this is nothing new.

Both the article and Falanx's company line imply that the chip is really
intended as an embedded solution for closed boxes. It will be competing
mainly with Intel and nvidia's embedded solutions on the PC platform.

we certainly need some competition in the graphics industry, currently
dominated by Nvidia and ATI.

Neither of those companies are exactly rolling in the dough. Bringing in a
3rd party will merely result in the elimination of one. The GPU market seems
to have room only for 2 major players. There aren't enough sales and profits
to support a 3rd.

 

[J. Clarke]

The scalability of the core is irrelevant since the end user product has a
fixed profile. Without a cost-performance analysis the whole thing is
meaningless. The Voodoo cores were also scalable so this is nothing new.

Both the article and Falanx's company line imply that the chip is really
intended as an embedded solution for closed boxes. It will be competing
mainly with Intel and nvidia's embedded solutions on the PC platform.


Neither of those companies are exactly rolling in the dough. Bringing in a
3rd party will merely result in the elimination of one. The GPU market
seems to have room only for 2 major players. There aren't enough sales and
profits to support a 3rd.

The trouble is that Intel decided to try to take over the graphics market by
making their own graphics part of their chipsets. This has been bad in a
variety of ways--the Intel embedded graphics is crap, but it's essentially
free, so we have bargain basement system with really poor graphics and
nvidia and ATI forced to compete with free for the mass market. Both are
now trying to do so with their own chipsets--nvidia's nforce chips are
quite good for the AMD market but neither has so far been any more
successful than SiS or Via in competing with Intel for the Intel market.

Maybe if AMD wins their lawsuit then we'll see a more even playing field.