Author: Aaron Gordon
Date: 15:34:17 07/02/03
Go up one level in this thread
On July 02, 2003 at 13:02:21, Sune Fischer wrote: >On July 02, 2003 at 12:38:42, Aaron Gordon wrote: >> >>I test the non-modified chips for 24 hours with Prime95 & BurnK7. The dual chips >>get tested for 48 hours. These are GOOD tests, as some people have >>non-overclocked chips that fail in BurnK7 and Prime95. You obviously haven't >>bothered reading the "About our pre-tested chips" on my webpage. >> >>>Overclocking at your own risk dudes. > >I agree with Vincent. >Your tests do not seem thorough enough. >24 hours is a good start, but like I said it takes weeks. >To give you a concrete example that happened to me: >I was once OC'ing my 1 gig to 1133 MHz, it posted and booted just fine. >I crunched 3 seti WUs in a row with no signs of instability. >The next time I booted it didn't even post, I had to clock it back down to 1060. > >My dual celly ran 433@559 for years. Now they won't run stable at anything above >500. I can't explain it, but things are not as simple as they might appear. > >>>If i overclock i take the risk that i lose bits simply. In fact you can easily >>>measure that using programs giving given outcomes like some scientific matrix >>>calculations do. Those run at a PC a few months / years before producing >>>results. Bit less if you run them dual. >> >>You keep thinking the chips are technically overclocked, which infact they are >>not. These chips have 2.4GHz cores and run those speeds with ease, at low >>voltages. If you push up to 2.5-2.6GHz, you have to raise the voltage and then, >>and only then are you actually 'overclocking'. I'm sure you would still consider >>an Athlon XP 1700+ (1.46GHz) with a 2.4GHz core set to 1.53GHz "overclocking". > >You keep thinking that all chips are identical. >They can be very different I think, my (plausible?) theory is that some may >contain a few bad wires that makes them a bit more sensitive to high MHz. >Considering that they have 30+ million transistors, it doesn't take a lot to be >slightly broken. Like maybe the silicium wasn't 100% perfect but only 99.5%. >They test this at the plant and sell them at a bit lower freq instead of tossing >them out entirely. > >The problem is that it may only be part of the chip that is half broken, and the >code you run may not be testing that particular part. > >-S. Any chip that has any overclocking problems has failed extremely fast with BurnK7 and Prime95 from my experiences. I've had friends that played Quake, did other games & encoding.. I told them that is definitely NOT good stability testing. They swore up and down their PC was stable, soon one of my friends fired up BurnK7 the machine rebooted instantly. That is just how harsh it is. These chips run BurnK7 flawless, and aren't clocked anywhere near the "edge" of stability. It's a pitty people don't understand AMD/Intel make all chips 1 speed and then mark them to whatever the market demands (be it 1700+ or 3200+). See my post here for more details on my testing methods: http://www.talkchess.com/forums/1/message.html?304354 If the chip is that "on edge" stability wise, to where there are potential problems running your every day applications BurnK7 will push it over the edge just from the heat increase alone (you can verify this via my formulas in the URL above), or at the very minimum make the system extremely unstable instead of rebooting/locking up. Once you find out what your chip can do then you can adjust the clock/voltage to run completely stable speeds. Most people are too fixated with, "overclocking == running on the complete edge of stability". Thats not whats going on here at all.
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.