Originally posted by mozez
am i the only person who would rather apple go to amd than ibm
yes, you are.
I, for one, know that the best thing Apple could do is change the preferred architecture that OS X runs on. I mean developers will LOVE to learn how to optimize their code for X86, 3dNow, 3dNowPro, MMX, MMX2, and SSE. They will throw LOTS of extra resouces into developing software for the 4% market of the Mac so they can have their code run fast on two different architectures. Apple won't short change PPC development for the new Architecture either. They will dump tons of R&D to support the Billions of dollars of legacy hardware and software that Mac users have invested in PPC machines.
And the users... they will flock to the new x86 OS with the pretty buttons and no applications. Grandmothers will jump on the bandwagon because they like recompiling open source code so that the unix-ported-to-PPC-OSX projects will run on their new x86 boxes. And the developers will see this huge influx of new app-less users and they will roll the applications out lickety split.
<--- I'm being sarcastic of course
at the end of this year or next year, it will be still waaaaaaaaaay behind x86, and so really, what was the point?
Waaaaay behind? what are you smoking? IBM has announced pre-production SPECfp at over 1050 for a 1.8GHz part and they said that this was conservative and that it would go up.
That score is slightly behind the reported SPEC of the Athlon64 at 1.8 GHz... it's is slightly under the 3GHz P4 and it is way ahead of the AthlonXP 2800+. Even the mighty 1 GHz Itanium 2 which sells for what? $4000 per processor module?... only rates about 20%-40% greater performance than the 970's preliminary SPEC.
The Athlon appears to be stalled, Tom's Hardware is saying that AMD needs to move on.
The P4 STILL shares registers between FP and SSE2. Altivec is far superior to SSE2 (and AMD doesn't even do SSE2 yet).
Now consider that IBM will have a robust .09 micron production line up at Fishkill long before the 970 is released, and that IBM is already hinting that they may debut over 1.8 GHz. I think the 970 will do just fine against x86 offerings, especially since it will run nicely as a dual rig with those power consumption rates.
Also, consider that the Athlon64 appears to have been pushed back until the 3rd quarter, and Intel has a history of only releasing chips that are slightly faster than AMDs. Intel won't release a 3.6GHz P4 if AMD is only shipping 3200+ chips this summer... they have no reason to because their yeilds are higher when the clock range is lower. They are happy just being faster, and it appears they will remain faster until the Athlon64 begins to ramp up speed.
no, i would really like to build a custom mac, alot of techies would and alot of people are building their own machines today, so why not apple mac a profit off it. why would osx sell?
gee, everyone I know steals MS Windows. M$ makes their money from OEM licensing on their 95% marketshare, not from home builders. BYO Powermacs would
reduce OS X sales, not increase them.. as fewer Mac/OS bundles would be sold.
the 970 to me seems the wrong way to go, considering that spec scores show it got it's ass handed to it by several other processors, many of which were cheaper.
Now I know you are smokin... The 970 hasn't gotten it's ass handed to it by any processor. The estimated specs are impressive and they dont take Altivec into account. Not only that, but they will go up before the Chip hits market.
What I'm really curious about is how you think the chips that apparently beat the 970 were cheaper? How much does a 970 cost? You don't know because NO ONE KNOWS. Duh. From what I've seen, it doesn't look all that big. If IBM gets good yields (and their process is generally very good) the chip won't be expensive.
Do a little research before you start posting crap.