To: steve harris who wrote (172099 ) 12/5/2002 12:56:59 AM From: Windsock Respond to of 186894 "AMD has planted each of its feet on the deck rails of two different ships heading in two different directions." techupdate.zdnet.com Battle of the bits By Bill O'Brien, Tech Update December 2, 2002 10:43 AM PT There's no denying that the clock's ticking for AMD. Like Intel, AMD has probably seen the writing on the clean room wall: "Profit margins in the consumer space are very low." They're so low, in fact, that they won't (and haven't been able to) support the level of R&D it needs to remain competitive. And like Intel again, AMD needs the higher ticket IT market so AMD can carry on at least into the rest of this decade as a viable company. But while AMD has competitive products in the works, the Opteron and the Athlon 64, Intel has at least a one-year head start and a steamroller attitude. So just what are AMD's chances? A little timeline snooping might provide some clues. Don't worry, it's a short line. October 18, 2002: AMD announces that it suffered a net loss of $254M in the third quarter of 2002. Whether you read that as a 34 percent drop in third-party sales compared to Q301 or as a 32 percent drop during the first nine months of 2002, it's not good news. That's after signing a deal to supply Compaq with Mobile Athlon XP processors earlier in the month and just prior to announcing that supercomputing giant Cray adopted the upcoming AMD Opteron processor to fulfill a Sandia National Labs computer contract. At that point in time, the Opteron existed only on paper and in AMD's lab, and it wasn't until a month later that the 2200+ Mobile Athlon XP processor was announced. October 29, 2002: AMD embarks on coast-to-coast tour to unite with computer enthusiasts and channel partners. "PC enthusiasts are core AMD customers," said Ed Ellett, vice president of client business for AMD's Computation Products Group. AMD's PC enthusiasts may indeed be enthusiastic, but they're apparently not paying the bills. Fifteen days later, AMD announced it will lay off about 15 percent of its global work force by the second quarter of 2003. November 18, 2002: AMD announces that Covalent and Red Hat are developing 64-bit Apache Web Server software for the upcoming AMD Opteron processors. The two telling words in this announcement are "developing" and "upcoming." At this point, according to Intel, more than 80 percent of the software needed to make the Itanium a viable choice has been developed, debugged, and is waiting in the wings. November 18, 2002: AMD demonstrates a prototype 64-bit Microsoft Windows operating system running servers based on its upcoming Opteron processors. Same problem words as above: "prototype" and "upcoming." At the very least, NEC is shipping Itanium 2 servers running a customized version of Linux based on a Red Hat release. IT likes Linux--Windows is still an unknown quantity in the high-end server space. November 19, 2002: AMD decides to name the CPU it codenamed "Clawhammer." The new name is AMD Athlon 64, to denote the 64-bit nature of the processor (it will also handle 32-bit code). Quoting AMD's Ed Ellet from the press release: "There's unstoppable momentum building towards 64-bit computing because it will overcome the inherent limitations of 32-bit systems. The upcoming AMD Athlon 64 processor will bring the performance benefits of 64-bit computing within reach of all desktop and mobile PC users." Excuse me? So the 32-/64-bit Opteron and this new 32-bit capable Athlon 64 have "inherent limitations?" Worse still, five days earlier, while waiting for 64-bit code to reach the workstation level, Intel released a Hyper-Threading version of the Pentium 4 that will pump up (by a double digit percentage in some cases) productivity. AMD will release an Athlon XP 2800+ that might have given this new 3.06GHz Pentium 4 a challenge--except it will be a restricted release to a small number of vendors who build gaming PCs. Think whatever ill thoughts you will of Intel, but it took a stand on the 64-bit issue that dominates the RISC server market. It brought on Xeon processors to power the mid- and low-end market, and then added a Hyper-Threading Pentium 4 to cover its bases in those marginal cases where even multi-way Xeons might not be cost-effective. In contrast, AMD has planted each of its feet on the deck rails of two different ships heading in two different directions. If it's going to be competitive it must develop a similarly consistent approach. Otherwise, the Opteron will be great if there's ever a 64-bit version of Quake. Do you think AMD has lost more ground to Intel? Can it ever catch up in the enterprise? TalkBack below or e-mail us and tell us what you think about AMD's chances.