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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: revision1 who wrote (81225)	5/31/2002 8:10:54 PM
From: wanna_bmw	Respond to of 275872

Revision, Re: "Sounds like the Patent examiner in the early 1900's who said "..every thing that can be invented has already been invented.." Several uses for 64 bit (and 128 bit in the future) will be extended-extended precision arithmetic, FP and Integer, handling ultra large databases, encryption and security, and there is always packed data handling." I really meant for that comment to be taken in the context of the current computing paradigm, which I expect to stick around for the next 10-20 years. After that, with quantum computing or what have you, anything can change, including who will be big in terms of compute manufacturers (since semiconductors may not even be the in thing with computers). In the mean time, I don't know what you mean by extended-extended precision arithmetic, 64-bit registers can hold integers up to 18,446,744,073,709,551,615. For numbers larger than that, people use floating point. Double precision IEEE-754 floating point has 52-bits for the significand field, which gives plenty of accuracy for even the most demanding scientific applications with 17 orders of magnitude in resolution, while an 11-bit exponent field allows very large or small numbers to be represented. And if that's not enough for you, 80-bit extended floating point should do the trick. Exceeding these specifications won't be happening within the current computing paradigm, that's for sure. As for ultra-large databases, the size is dependent on the memory in your system, not the size of your registers. 64-bit addressing (which is not going to be available on Hammer, or on Itanium 2 - rather these two use 40-bit and 50-bit addressing, respectively) can address 16,777,216TB worth of memory. Given that memory requirements double roughly every 1.5 years, I won't expect current large databases to break this until ~2044, and again, that's way outside of the current computing paradigm. I guess I could be wrong, but I think I'm right at least in the scope of today's 64-bit and greater needs. I'm not here to predict the future. I was just arguing against the prospect of 128-bit and 256-bit processors being the norm in 10 years. But I could be wrong about that, too. wbmw