Uebernimmt Sun Micro AMD??? - 500 Beiträge pro Seite

[woernie]

es gibt geruechte in usa ueber eine moegliche uebernahme! was haltet ihr davon?


A Sunny Future for AMD?

This rumor is admittedly extremely tenuous, but it is worth mentioning because it is so intriguing. Allegedly, AMD`s wild success this year has caught the attention of computer giant Sun Microsystems. Additionally, the server heavy computer company is purportedly very impressed with the potential performance of the Sledgehammer server processor which, the rumor goes, is already in alpha silicon. Sun is supposedly considering server lines based on this 64-bit processor.

That Sun might be considering an acquisition makes sense given the sophisticated technology that AMD is throwing at their 64-bit platforms and the unbelievably undervalued state of AMD. The Texas centric chip company might currently be one of the best bargains in the computer industry, making AMD and its wealth of top notch engineering expertise a mouth-watering takeover target for wealthy companies like Sun.

One rumor that we can dispel is that, due to patent issues, the Sledgehammer will not be a multi-core chip. We have been assured by an AMD representative that the Sledgehammer will indeed have two cores.

Extreme Rumors

While we are venturing this far out, we might as well revisit the unsubstantiated rumor above that Sledgehammer has already reached alpha silicon. The 64-bit chip will supposedly be ready to launch in early Q3 of 2001 if necessary. Also, for those worried about AMD`s production constraints, the Sledgehammer will allegedly be produced on an SOI process in an IBM copper fab. These are hefty rumors!

This contradicts what is being reported elsewhere and attributed to AMD representative, Bob Mitton. Mr. Mitton disclosed in an interview with the Japanese PC Watch that the first Hammer has just now completed layout and should reach "tapeout" before the end of the year. Hardware sampling should begin early next year.

In the interview, translated on JC`s PC News and Links, Mr, Mitton also states that SSE2 will only be implemented for its IEEE floating point interface and that integer SIMD instructions will not be supported. If true, we find this disappointing, since integer operations will be processed normally. A full implementation of SSE2 would give the Hammer line key strategic footing with Intel`s coming processors on apps expressly optimized for this standard.

It must be stated that Mr. Mitton has also told members of the press at the recent Microprocessor Forum that the Mustang will carry no architectural improvements over the Thunderbird other than a larger L2 cache. This claim is widely dismissed by industry observers as smokescreen, since many believe the chip has a better branch prediction unit and has other enhancements to help it reach higher clock speeds on a given process than the Thunderbird (I was told directly by another rep that the Mustang`s BPU is improved).

[schmackson]

Danke schön.
Aber wie wär`s mal mit `ner Quellenangabe? Dafür nochmals und schon im Voraus DANKE.

[woernie]

http://www.tomshardware.com/column/00q4/001101/bb-07.html

und

http://www.tomshardware.com/column/00q4/001101/index.html