SI
SI
discoversearch

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: Chip Anderson who wrote (12881)5/21/1999 12:12:00 PM
From: Sun Tzu  Read Replies (1) | Respond to of 16960
 
>> but the data will come from a connector cable to the main card

Sounds like a big waste of bandwidth, especially since it is already a big bottle neck. The whole concept of SLI in its V2 incarnation should be changed IMO. What you need is a way for multiple cards to get and process their data independently while still producing syncronous output. This could be done via drivers and connector cable. I don't see (besides convenience) why the processing load has to be evenly distributed between two cards. Why can't the cards roughly distribute the load based on a mix of their processing power and bandwidth availability? Idealy, I'd like to have 3 monitors connected to 3 graphics cards like I have on my unix work stations, but that's another story.

ST



To: Chip Anderson who wrote (12881)5/21/1999 12:16:00 PM
From: Piranha  Read Replies (1) | Respond to of 16960
 
<< SLI will be done via a daughter card architecture >>

Do existing V3's have a connector for this? (NOT a rhetorical question, I honestly don't know) If not, you can't market a daughtercard as upgrade to existing V3's, which (based on the PC Data report) is a very large number of units.

Unless you're not talking specifically about V3.

Piranha