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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Tenchusatsu who wrote (135137)	5/15/2001 4:30:41 PM
From: pgerassi	Respond to of 186894

Dear Tench: Given Intel's past specs on P3, 10C (18F) equals about 66MHz. The mere fact that overclocks on P4 are at most, 15%, shows that it is very likely to not like you doing it. I do not run my Tbird at max attained speed either to give myself some margin. I'm at 150MHz (300 DDR) FSB due to one PCI card's limitations. It goes to over 160 without that card. Pete

To: Tenchusatsu who wrote (135137)	5/15/2001 4:38:51 PM
From: tcmay	Read Replies (3) \| Respond to of 186894

Overclocking has more risks than just junction temperature! Tenchusatsu writes: "Pete, so you're saying that if the Pentium III is heat-limited (according to your definition), I can easily overclock my processor with the stock heatsink and fan, since I would have a lot of headroom in regard to max die temperature, right? "If so, I'll pass. I have a overclocking "batting average" of about 0.300 in my lifetime, and I don't feel like wasting too much time trying to raise that average." I used to be a reliability engineer at Intel. Overclocking is fraught with dangers, not just thermal ones. Parts of a circuit may not "run at speed." Various reasons for this, including varying RCs in some circuits, drive currents being inadequate to get a node up to where it needs to be in time. (My lab also used to use scanning electron microscopes and image analyzers to locate "critical speed paths." A more persuasive argument is perhaps this one: If AMD, Intel, etc. could have sold their chips at 1.5 GHz instead of at 1.3 GHz, don't you think they would have? A lot of bucks in that extra margin of speed! And bear in mind that a chip vendor generally knows more about testing the various functions and interdependencies than an end user does. (I say "generally"...sometimes the customers find things the vendors missed.) Running "Quake" at 1.5 GHz for several hours may not be an adequate test. In fact, most CPUs probably _can_ run at overclocked speeds...just not reliably, consistently, and safely. I've overclocked two of my machines: I overclocked a 6 MHz 80286 to 8 MHz, largely because we all knew the CPUs were rated for 8 MHz but that IBM had designed their motherboard conservatively for 6 MHz. And I overclocked an Apple Macintosh PowerPC 7100av from 66 MHz to 80 MHz for largely the same kind of reasoning. But I'd be mighty leery of overclocking a part AMD or Intel says has been tested to perform reliably at 1.3 GHz to "as fast as it can go before Quake stops running." Maybe if some test sites do a _lot_ of studies (but then the question is why the vendor didn't pocket the extra $$$ by selling it as a higher-speed part), maybe if some special cooling (a la cryocoolers) is used. Most of the "uberclockmeister" types of sites I've been to don't have _any_ results of reliability tests, comprehensive test vector reports, etc. A 20% boost in speed hardly seems worth this kind of risk. --Tim May