Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : All About Sun Microsystems -- Ignore unavailable to you. Want to Upgrade?

To: techtonicbull who wrote (44359)	8/4/2001 1:59:00 AM
From: Prognosticator	Read Replies (1) \| Respond to of 64865

When I talk about 10THz I'm talking about the effective speed that an instruction-level parallelism CPU such as Itanic would have to run at to keep up with 1024 10GHz CPU's that exploit thread-level parallelism. I believe this is achievable, and more. Moores law only applies to single CPU cores, and they will hit the quantum/relativistic barrier by 2015 for sure topping out at 10-20 GHz. Not shabby, but two orders of magnitude too slow to compete. Provided you can make each CPU small enough and low-power dissipation, there is no such Moores law limit to what can be achieved with multi-core CPU's. Intel may be working quietly on such a beast, but the key is that Java software is inherently able to take full advantage of the thread-level parallelism, while C++/C software is not. However, Itanic is taking an opposite direction, instruction level parallelism with horrendously complex and therefore buggy compilers and unreliable systems, as noted many times by Qwik. It's a dead end. IMO by 2010 SUNW will completely dominate the server space unless they fail to execute. IBM and INTC will have to play catchup as usual. But they'll all be running Java byte-codes and licensing from SUNW. M$FT will still be under anti-trust scrutiny for desktop monopoly abuse, and .NyET will be in the same file-cabinet as .BOB. P.