Author: Robert Hyatt
Date: 18:05:31 02/23/03
Go up one level in this thread
On February 23, 2003 at 01:39:52, Jeremiah Penery wrote: >On February 23, 2003 at 00:37:56, Robert Hyatt wrote: > >>On February 22, 2003 at 02:01:30, Jeremiah Penery wrote: >> >>>You're making some seriously unfounded statements about these issues, that have >>>absolutely no basis in reality. >> >> >>There is a _lot_ of reality in his statements. To wit: > >The unfounded assumptions I was referring to were that using 'cheaper' parts >means your machine will be less reliable, and that using those parts >necessitates exotic cooling methods in order to get good performance. > >I have said absolutely nothing referring to any of the below. > >>1. If he has a failure in the hardware, Dell will have it fixed _tomorrow_ >>with no questions asked. Nearly everything they ship comes with a three year >>warranty that includes next day on site maint. >> >>2. If your bolt it together yourself machine dies, tomorrow you will be >>sitting on the phone discussing the problems with a variety of vendors, >>having to justify why _they_ should replace your CPU when you obviously >>bought a MB capable of overclocking. > >If you build your machine yourself for 1/5th the price, you have a lot of extra >money to replace the broken parts yourself and not worry about the cost. :) Yes, but you will experience a week of downtime to get that replacement part, which might kill a stock trader... > >But seriously, most parts you order come with some kind of 90-day warranty from >the store or perhaps 1-3 year manufacturer's warranty. Some parts even have >lifetime warranty from the manufacturer. Last year, I had a video card fail. I >called the manufacturer, sent the card in, and had a replacement within a week, >for free. Most of the time, it's not much more of an ordeal than it would be >with Dell. Dell guaranteed next day repairs. You just mentioned a week, which has been my experience in getting drives replaced when I ordered them myself. I do this still, but I know the risk. For critical machines, everything comes from Dell, and then I know that if it dies today, it will be operational tomorrow morning if not sooner. > >>3. He also bought a _bunch_ of disk performance. Which will cost the same >>whether the MB is cheap (as you have to add a decent U320 SCSI controller) >>or not (MB comes with onboard U320 SCSI). > >I never said one word about the disk performances or cost of anything. > >>For someone with a machine that _has_ to be up, there is absolutely no >>comparison between a do-it-yourself overclocker and a commercial off the >>shelf with onsite warranty system. > >I work at HP. I build computers every day. I would place _far more_ trust in a >computer I built myself to be fully reliable than the 'commercial off the >shelf...' system - onsite warranty or not. If I had to have absolute >reliability or whatever, I wouldn't overclock. I don't overclock now. I just >don't hold the paranoid view that overclocking will absolutely cause failures. > I think the math is simple. If you believe that overclocking _can_ cause a failure, then doing it long enough is going to guarantee a failure. The only issue is whether it is a detected failure or not. Many are not. >>Not everyone just uses their machines for games. Some do real work. Some >>depend on them for their very livelihood. > >No matter how careful you are, failures will happen. How many disk failures >have you had in the past few years? I can remember at least twice. I've >_never_ had a disk failure of any kind. Does that tell much? Not really. I >could say you're treating your drives badly somehow, but I'm sure that's not the >ca I have had at least six failures. Three on my quad xeon (9 gig 10K scsi drives). Two from IBM, one from Seagate. average time to repair was one week, them shipping me the replacement. I had a 40 gig drive fail which drove me to upgrade the old ftp machine to two 80 gig maxtors. One of those failed which led to the recent upgrade to 3 X 146 gig scsi drives. The other failed a few weeks later. Both have now been replaced. I'm not counting failures on my cluster, or our suns, or our SGI boxes, or our PCs and so forth. I am only talking about crafty.cis.uab.edu and ftp.cis.uab.edu... se. > >Every single person I know who has a Dell/Gateway/HP system has had nothing but >problems from day one with them, with one exception for a friend of mine who >upgraded the machine himself. I am another exception. I have about 200 dells in labs here. They run day in and day out, 24 hours a day, 7 days a week, with hardly any failures of any kind. Most failures are floppy drives where students stick things in them that don't belong. My cluster hasn't had a hardware problem in 2.5 years, with 9 quad xeons. My office machine is now a Dell that has only been running a few months, but zero problems so far. Our faculty PCs are dual-xeon 650's from dell, no problems. I can't imagine that someone would have problems with dells, assuming they know what they are doing and don't put coffee cups on the CD tray or whatever. :)
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.