To: Tenchusatsu who wrote (36428 ) 8/28/1998 7:18:00 AM From: Maxwell Read Replies (3) | Respond to of 1572361
Tenchusatsu: <<No doubt about it, DirectX 6.0 is a big win for 3D-Now, but I think your 50% figure is a little exaggerated.>> Exaggerated? You must be wishing. Go check out the 3D benchmark on AMD homepage. The 3D benchmark on the K6-2-300 matches the PII-400. In that benchmark, the program calls for drawing 500+ 3D figures. It is a simple program and it shows how powerful K6-2 3D engine is. If games are written in 3D and are well coded, the K6-2-350 leaves the PII-400 in the dust. The next cool game out is Trespasser. It will show you why K6-2 is better than PII. <<Why? You can assume that all it takes to port 3D-Now drivers over to KNI will be a recompile.>> Sorry to disappoint you. It ain't that easy like you wish. Perhaps you could start your company porting 3DNow drivers to KNI. KNI will probably end up as a dead bead just like MMX. <<Well obviously Intel sees AMD as a real threat. But I'll bet you it's not because of die size. (I think Jim McMannis is overblowing the advantages of a smaller die size.)>> Why AMD is a threat to Intel? #1) AMD has 3DNow! and DirectX 6.0 #2) AMD has a SMALLER DIE SIZE In semiconductor, life and death is depended on the yield and die size. These 2 factors will determine who has the lowest cost. You'll be a fool if you think Intel 154mm^2 die is cheaper to produce than 80mm^2 K6-2. <<As for 3D-Now's chances vs. KNI, I have two words for you: Critical Mass. Can AMD achieve it, considering that everywhere you go, the K6-2 is being pitted against the Celeron? >> How many installed units must there be for you to consider critical mass? In this market, it is the retail consumers that define the critical mass, not the business market. The retail market people play games and demand faster games. AMD already has 38%+ of the retail and is growing fast. Mendocino is a dead end product. Intel will move to socket 370 and it is the end for people who want to upgrade. Interestingly enough, CPQ, IBM, and HWP think the CeleronA is ideal for business market, not retail. It is too bad for Intel. Intel wanted the CeleronA side by side with K6-2 but OEMs say NO! The reason is that the CeleronA would kill their PII sales. <<Until then, critical mass just won't be achieved, and 3D-Now will be in "danger of becoming a footnote in history," as Microprocessor Report puts it.>> You have been reading and believing the Microprocessor Report too much. It is a nice journal but the writers think little of AMD. Let me put it this way. AMD may not have the marketing muscle like Intel but AMD DOESN'T CHEAT consumers like Intel. Intel promised the MMX and delivered NOTHING! They provided NO SUPPORT MMX. MMX is a lost cause. MMX is a blow in Intel face. MMX showed what Intel is made of, just skin deep. MMX sucked! 3DNow! on the otherhand is an INNOVATION. After revealing it at the Microprocessor Forum last year Intel decided to dispatch a team of engineers to COPY AMD 3DNow concept. Intel is best for copying other people ideas just like they copied DEC in branch prediction, etc. AMD supports 3DNow! They even rewrite QuakeII in 3DNow to show what 3DNow can do and to motivate ISV to write. Intel wouldn't never have done this. They think that if they make anything people would come and support them. Well, WRONG! MMX sucks. In less than 3 months after the release of K6-2 there is a dozen of softwares already available. ISV couldn't believe the performance of 3DNow. They couldn't believe AMD could have come up with such an innovation. They now believe in AMD, buy AMD parts, and support AMD. They figure it is the next best thing to Intel. Intel is all hot air and too difficult to do business with. That is why I said all along Intel will one day pay for harassing their OEMs. Maxwell