Date: Sun, 28 Sep 1997 22:49:49 -0500 From: Ben Bryan <bbryan@umr.edu> To: hardware@freebsd.org Subject: Re: supermicro p6sns/p6sas Message-ID: <3.0.1.32.19970928224949.006a229c@pop3.umr.edu> In-Reply-To: <Pine.BSF.3.95q.970928194724.21363A-100000@misery.sdf.com> References: <199709282221.RAA23164@compound.east.sun.com>
next in thread | previous in thread | raw e-mail | index | archive | help
At 19:58 28-09-97 -0700, you wrote: > >On Sun, 28 Sep 1997, Tony Kimball wrote: > >> Quoth Michael L. VanLoon -- HeadCandy.com on Sat, 27 September: >> : Socket 8 indeed appears to have been a limited detour. Slot 1 may in >> : time be superceeded by something better. >> >> IIRC, Intel plans to phase out slot 1 during the coming year. >> >> : Socket 7 has a very limited future. It simply doesn't have any way of >> : coping with some of the new technology that will be necessary to make >> : any sort of performance boosts possible beyond 300MHz. >> >> The supporting argument is insufficient for the conclusion. AMD and >> Cyrix are actively engaged in providing alternate paths to cache. One >> proposed solution is to make a low-profile Socket 7 module with an >> on-board cache controller, making the Socket 7 interface effectively >> a point-to-point bus -- analogous to Slot 1. > > None of these methods exist yet. I'm little dubious about how well AMD >can make it work. K6, due to bugs, was unable to run FreeBSD reliably up >until a few weeks ago (see archives about which stepping are known to >work, and which are not). Actually, the on chip cache controller does exist, sorta. The Cyrix MediaGX line of processors have an on chip memory controller and run without L2 cache yet perform almost identically to the 6x86 line. > >> : AMD and Cyrix may indeed be able to perpetuate several years more of >> : Socket 7 chipsets, in the secondary market. But it will be a typical >> : AMD/Cyrix market, where performance is secondary, and low-cost is >> : king. >> >> K6 and P-II numbers are within a few months of each other, so I can't >> agree with the implication that it will therefore not be possible to >> maintain a Socket 7 system with competetive performance. > > Is CPU performance measured in months now? :) > >> I do believe that it is quite clear that Socket 7 has superior life >> expectancy to both Socket 8 (already hard to find many P6 parts) and >> Slot 1 (Intel-planned short lifespan, to be replaced by Slot 2 in 98). > > I don't find it hard to find socket 8 parts (motherboards and CPUs). >Everyone has them, and lots of stock. > > Why is it clear that socket 7 has superior life expentency to socket 8? >Socket 8 is better (better bandwidth) than socket 7, so why doesn't AMD >just make socket 8 processors rather than messing around trying to sqeeze >more out of socket 7? I agree with Michael. AMD/Cyrix is targetting the >secondary market, otherwise they'd be cranking out processors to use with >the faster socket 8. > > Also, you can use socket 8 processors in a slot 1 with an adapter. >I have no idea why AMD isn't exploiting this to put CPUs onto slot 1 >motherboards now. > >Tom Why don't they use socket 8? Simple. Intel patented it just as they did slot 1. Slot 1 will be dead by next July. Socket seven will be in use by AMD and Cyrix for the next year at least. After then they will announce some new socket/slot design for the K7/7x86 or whatever they're going to call the next generation. And given the performance of a Cyrix chip at 225Mhz (winstone w/in a point of a PII 266) I'm pretty sure most users will be happy with a Cyrix chip at 300Mhz. And those who want more FP power can get one of the new K6+ chips that are due to be announced next month. Or they could just go buy a 21264 Alpha... Ben Ben Bryan ME/CS University of Missouri-Rolla Listen to KMNR, you fools. 89.7fm http://www.umr.edu/~kmnr
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3.0.1.32.19970928224949.006a229c>