To: stock talk who wrote (3721 ) 7/9/1999 9:11:00 AM From: SBHX Read Replies (1) | Respond to of 5927
I was aware of both SIII and ATY during the musical chair 'switchover'. AND, you probably have to be canadian to know this, but a long time ago, ATI was the market leader with their EGA wonders and Graphics Solution. Way before SIII and cirrus logic showed up. ATY has had its turn at relinquishing their lead for the musical chair thing, think they're willing to do the same thing twice? I think the conditions are vastly different. ATY did not exactly miss 2 design cycles. SIII did. The savages were not just late, it was not even past design stage when their competitors are ramping up production. When rageIIs and voodo1s were shipping in volume, S3 did not have a 3D chip. All they did was repackage the Virge and added 3D behind the label --- with tragic results for their sales force. When the ragePros and voodoo2s and riva128 were shipping in volume, S3 did not have a 3D chip. Morale there was terrible for a long time. ATY did not see any shrinkage in unit volume in the most profitable high volume segments. SIII did --- they basically had no margins for a long time. In fact, ATY seems to have deliberately abandoned unprofitable segments to SIII at the right time. First, chips with no video scalers, then chips with no 3D. What's next? No margins = no profits. I just don't see the same parallels. From what I heard, the savage4 is not a bad chip, but it is not as fast as people thought at first (that was just a benchmark trick [as exposed by zdlabs]), in terms of actual performance, it places well behind G400, TNT2, voodoo3, and rage128. This means that there will be some market acceptance of this chip.... but this chip is not a lot faster than a Whitney (i810), and hence could face some price pressures. This is why, based on day to day operations, it should be harder for SIII to turn a profit. They really need a faster chip to differentiate from whitney. It also seems that everyone else (matrox, nvidia[of course] and TDFX and ATY) has learnt a lesson and went for the ability to do a bleeding edge performance (price is no object) chip and then a build a cheaper slower version. SIII's savage5 might be an interesting chip, but it would again, be one-gen late. The N10 is going to have some useful features : T&L. Apparently, ATY also are about to reveal their new chip that competes with the N10. TDFX has made no secrets about their T&L plans. Technically, NVDA is a tough company to force off in terms of their willingness to blow away moore's law, so if they can actually deal with this diamond setback, they are a force to reckon with. [And after all those unfrozen shares are absorbed --- buy :-)]. As for ATY. It has always shown the right product mix with the right features and have always attacked the right segment of the market(s). Other than voodoo3's 3500, I don't see any competition for their All-In-Wonder128 product. Where else can you get a real-time MPEG2 video capture, a TV tuner with your 3D accelerator that still plays Quake3 at bleeding edge frame rates? The markets have in turn, rewarded them with good financial results. My guess is they will still be the winner in terms of profits and market share. I'm even willing to bet money on that. There is a level below which, I am accumulating ATY. I actually do sell above a certain price, but only because of short term market sentiments. In recent history, there is some worry about the strength of their competition, and precisely the factors you pointed out, and this is reflected in their relatively flat stock price. ATY did deny the market share loss as claimed by Ken P. So someone is lying or has wrong info. Who has more credibility? :-) Even if I were to lighten up on ATY in my portfolio in the short term, I still want to be holding lots of ATY when their settop box HDTV decoders go on sale. But I can't imagine switching from ATY to SIII until I know that KenP is not just another Banatao. I'm investing in high-tech companies, not investment VC firms. I think if we had a chart of ATI and S3 for 1999 & 2000 and 2001. It would look the exact same with one exception, I think S3 and ATI are once again changing places. Like they did in 1997 and 1998. Next cycle of musical chair leaders.