To: fooledalot who wrote (16688 ) 5/28/1999 5:27:00 PM From: Eric.sun Read Replies (1) | Respond to of 64865
fooledalot, I didn't read those articles before, but it shows compaq is trying to pull this one off. Like the guy in TI said, it would have been better if alpha could scale up 32 or 64 ways two years ago. And compaq intends to use window NT on it, I wonder how powerful the 32 or 64 way alpha server is in window NT. Even though the single alpha processor could be the best processor ever designed in industry, alpha server/workstation is less than 5 percent in the market place. Remember on one hand is how many ways CPU can scale up and on the other hand how much performance it scales up. And what kind of support people gonna get and what kind of front end application available. One more thing is Java, since Java gets lion mindshare among programmers, people would have strong bias toward SUN server/solaris unless alpha server can beat SUN server with much better performance and much cheaper price. IMHO, time is on SUN server side, alpha server will be left out in the dust. K7 is an interesting story and I know a company named Poseidon is developing multiprocessor chipsets for AMD. If AMD really can pull this one off, it will have dramatic impact on intel. But for SUN, probably it will be minor. Because K7 most likely targets the low end window NT workstation/server which SUN doesn't care much but that is where intel profit margin locates. If K7 does succeed, the semiconductor industry will see a big change with true low priced microprocessors across the board and it may not be a good sign for the whole industry. But as a underdog to intel for so many years, one would question what credential AMD has to beat intel in the high end. All just my personal feel at this stage, with SUN execute well so far, SUN might eventually challenge IBM mainframe with ultra_III series and that will make SUN the top dog in computer industry.