Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : 3DFX -- Ignore unavailable to you. Want to Upgrade?

To: Michael G. Potter who wrote (11273)	3/20/1999 7:05:00 PM
From: Patrick Grinsell	Read Replies (3) \| Respond to of 16960

Wicked PGP sucks and here's why... After reading the whitepaper on Wickeds web site I'm not very impressed. On my pentium II 375 my Voodoo3 2000 is getting about 50 fps at 1024x768 and my SLI is at about 66 fps. Obviously I don't have the drivers that are able to kick SLI off the map...if they even exist. I'm going to have to see if I can pick the new drivers up. Both were using timedemo 1 in Quake II. According to this web site, Banshee PGP is getting 36 fps at 1024x768 on a 400 mhz Pentium II: 3dconcept.ch More than all this though, the implementation is kludgy and expensive. The boards have to be modified so that they can make use of PGP. This means that the base board itself will be more expensive, even if you aren't using PGP. On top of that, the setup requires that you have some sort of external device to combine the output of both cards before forwarding the information to the monitor. I can see plenty of nightmares arrising from this technology. First, they are going to have to do some sort of fix to make sure that filtering and blending goes seamlesly across the card boundaries. Each card only goes up to the edge of it's half of the screen, but methods like trilinear filtering use a sort of averaging method to blend with the pixels around it. If course the card can only do this if it has the data for the surrounding pixels and this is near to impossible if the data isn't combined until after it's been kicked out the RAMDAC. Second, I foresee major compatibility issues. How are you supposed to read information from the framebuffer? How are the polygons split in between the two cards? Will these extra driver manipulations cause serious performance issues in next-generation games (Q3) that are driver limited? Finally, S3 has admitted that they are doing another rendition of their S4 chip later this year that will be essentially "two savage4 chips in one". Nividia seems to be going the parallel pipeline route too with NV10. Conclusion: Why would anyone invest in a kludgy technology that's expensive and might have compatibility issues when only a couple of months later better solutions will be available? It isn't worth the investment when the product will only have a lifespan of a couple of months. And I'm not the only one with this opinion: We then asked Creative about Metabyte and their PGP technology. Jim stated that while there are certain problems with PGP that Metabyte engineers have done a great job solving, the jury-rigged nature of it makes support issues much more complicated for Creative. Quite simply, Creative's installed base is huge, and they can't afford to offer something that they can't guarantee 100% firingsquad.com Pat