HP3000-L Archives

September 2002, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Arthur Frank <[log in to unmask]>
Reply To:
Arthur Frank <[log in to unmask]>
Date:
Fri, 6 Sep 2002 15:57:49 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (25 lines)
What Wirt writes here is absolutely true, about overclockers pushing the CPU beyond it's rated performance.  However, the original point -- that Intel and AMD intentionally cripple their chips -- still stands.  In the Coppermine-core Pentium III days, Intel's Celeron line was a "crippled" Pentium III.  Half of the on-die L2 cache was deactivated (it was still there, just unavailable for use) and the front-side bus was set slower (at first 66MHz vs. 100MHz for the PIII, then after a period of selling 133MHz FSB PIIIs, the Celeron was bumped up to 100MHz.  The current crop of Celerons are "crippled" P4s, with slower clock speeds and smaller L2 caches, though the FSB is the same.  AMD played similar tricks with their Duron line.

It's also quite possible that Intel and AMD give some chips a slower speed rating than they ought to have, simply to meet demand.  Take this plausible scenario:  Because the 2.0GHz chips sell for less than the top-of-the-line 2.6GHz chips, there may be greater demand for them.  And because of advances in manufacturing technology, perhaps most of the chips coming off the line might be judged suitable for a 2.6GHz rating.  However, they don't need so many 2.6GHz chips because of the market, so they just take some of that batch and rate them for less.  This is the rationale that folks use for overclocking.

Years ago, Intel starting incorporating FSB-multiplier locks to prevent overclocking (and AMD has since followed suit) but that hasn't stopped people from trying.

Art Frank
Manager of Information Systems
OHSU Foundation
[log in to unmask] 
(503) 220-8320

>>> Wirt Atmar <[log in to unmask]> 09/06/02 02:14PM >>>
<snip>

I don't presume any sort of malfeasance to Intel's or AMD's motives in this
instance. In fact, what they're doing is what every engineering organization does: providing a guaranteed working system by building and delivering a better system than the one that they're promising you, one with a safety factor built in. Every bridge, every building, and ideally every tire from Firestone will operate and survive at much higher conditions than that at which it is rated.

The World Trade Center buildings almost survived the impact of much larger aircraft than they were designed to survive. If a slightly larger safety factor had been put into their design, they would be with us now. There's fundamental no reason they had to fall down. Intel and AMD are "giving" you much better machines than they promised only to insure that they will meet or exceed their marked specifications under all conditions. It's standard engineering practice. Those people who push the chips and busses beyond their rated values are working in that "safety factor" zone, but they're also doing it at their own risk.

Wirt Atmar

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2