Slashdot Mirror
Teach An Old Athlon New Tricks
budn3kkid writes "Seems like Upgradeware have a new gadget out for those overclockers looking to upgrade their age old Athlon mobo (KT133, KT266 etc.) with a spanking new AMD Barton CPU. Also, saw an article at ol' Tom's about it right here as well."
"Now you don't need to paint the L3 bridges with a silver pen or cut the L3 bridges with a penknife on your Athlon XP processor jut in order to adjust multipliers. XP-TMC total multiplier controller can help you to do the job without any mod on your processor. The warranty of your processor can be kept alive"
So now stores that sell these CPUs now have even less to work with when determining whether John Smith walking into the store asking for a CPU exchange actually had a faulty CPU to begin with or was doing mad l33t overclocking in his basement last night.
Why is this a good thing? Someone that's willing to buy this device seems like they'd also be of the mindset to lie at the return counter if their CPU kills itself early after an Overclocking Session Gone Bad (TM).
The results of that? 1) The people working the front counters of the computer store hating their jobs due to arguing with customers about whether or not they tried overclocking their CPUs. 2) The prices for these CPUs going up due to the return/exchange rate on them. 3) Like in the case of some stores with Apex and other high-return brands, they just take a harder stance on people that want a straight exchange of the product.
Who wins here? The company selling the adaptor. No one else.
Because the chip rests on top of an adaptor, it would make the CPU sit higher. Wouldn't this present a problem with heatsinks? i.e. clamps that may use more force, or not able to reach the attachments.
So what kind of fans and heatsinks are compatible with this? Tom's made the point that the processor sits up higher, and that Upgradeware's heatsinks and fans must be used.
How reliable/good are these heatsinks and fans? And what other companies manufacture compatible heatsinks and fans?
The Political Programmer
I've had an Athalon box running 24/7 for over a year now. No problems at all.
Of course, it's running FreeBSD, so maybe your problems lie, ahem, elsewhere.
--
the strongest word is still the word "free"
I don't think they show a substantial speed increase, but I think they show a larger increase than by simply ramping up the CPU clock.
I know that if I clock my Athlon higher first using the CPU multiplier, and then using the FSB, my results show that the FSB boost is far more effective. Recently AMD has had more significant FSB increases than CPU speed increases (move from DDR200, DDR333, DDR400).
And the jump from 333 to 400 is incredibly small and had Tom's Hardware jumping all over the claimed speed of the Athlon XP.
Granted, an increase in the FSB helps a lot, but not as much in, say, a Pentium 4 or an IBM 970, where a large FSB is vital. The Athlon is comparatively simple, and the 400MHz FSB is overkill.
Again, an Athlon XP 3000+ will be faster on a DDR400 system than in my old 133MHz FSB system.
The Political Programmer
Actually when I upgraded from my socket-7 mobo to my Abit Kt7 RAID my uptime in windows 98 went from about 12-24 hours to a solid week. I was amazed. The old k6 system crashed all the time. With win2k I routinely get uptimes of several weeks, and almost always the reason for the reboot is due something other then a whole crash (like my sound drivers or explorer or whatever will get 'weird').
And I've got cheap-ass ram in here to boot. I've been pretty impressed with the reliability so far.
I do wonder though if this adaptor will cause any problems. For one, the thermal sensor will no longer be in contact with the actual device, and for another the pin lengths will be longer. Could cause some problems...
autopr0n is like, down and stuff.
My point is that, instead of sending a HTML page (possibly in another language) when an HTTP error is generated, it would be better if the server simply returned the HTTP error, and let the browser decide how to present it to the user. This removes the language barrier.
It was perhaps unfair of me to present this as strictly an IIS problem, because it is also possible to present custom error messages in Apache (and probably other web servers as well). The reason IIS deserves bashing here, is that it's default behavior is to send custom error messages instead of standard HTTP error codes. Apache will send standard HTTP errors by default. In my opinion this is the correct behavior.
It's kinda kewl that this adapter might squeak out some more life for a kt133 system (since millions of computer bits are winding up in landfills or 'recycled' often dangerously in the 3rd world or whatnot. but otoh a new nf2 mb can be had for $100, how much does this thing cost? cause if it is $40 or so plus, I'd be temped to just get a new board and add the old board to my 'bone pile' or ebay it. granted I might have to buy new ram to ... but a new nf2 board would crush one of these adapters running on a kt133 board.
actually I am happy to see you, however that is in fact a banana in my pocket.
I have come to the conclusion over the years that the most important aspect in computer stability is the motherboard, and often the problems are driver-related. For instance, I used to have a VIA-chipset athlon thunderbird. That machine was horribly unstable when I got it. By the time I got rid of it VIA had updated the assorted chipset drivers and everything worked great.
My XP has the SiS 745 chipset, and yet it is quite stable (and speedy.) However it does have a gig of more or less brand name ram. One stick is kbyte, and I forget what the other is. Makes a big difference.
The Power Macintosh G3 I'm sitting at now (Can't play games on it or do any real content creation but Mac OS X is winning me over quickly) has 512MB of cheap-ass bargain-basement PC100 in it - actually, 128MB of it is PC133. I've only had it up and running for three days, but it doesn't crash or exhibit any strange behavior at all. While comparing Apples to PCs is generally a silly comparison to make (because of the zealotry it inspires) my point is that it's a good, stable system with a well-manufactured and designed motherboard, using high quality components. (The IDE chipset, for example, is found in many other Unix workstations, such as those from Sun.) The Power Supply is a valid issue if you are not getting enough juice to run everything, and the ram is important, but the motherboard is crucial.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
"It is? If you're having trouble with the stability of your system, it's most probably either your Power Supply Unit or RAM, don't blame the CPU."
Though I can support your stability claim, I learned a lesson about cooling those things. When I lived in my centrally air conditioned apartment, my athlon worked great. When I moved into an apartment that was considerably warmer, I started seeing instabilities plus the bios would scream at me to shut it down.
The truth of the matter is that the CPU can be blamed, particularly if the environment exceeds its configuration. I suppose one could blame the user, but then again, if there's already a fan on it what do they do?
My dual athlon at work has impressive uptimes, even with Win2k, but darned if it doesn't sound like I work in an airport runway. *Whirrrrrrrrrrrrrrrrrrr*
"Derp de derp."
"This upgrade raises the concern of the choked bus. I mean, who really wants to run a 3000+ on a 133MHz system bus?"
Depends on what you're doing. I did some informal (and unscientific I should mention) tests using Lightwave on an an Athlon vs. a P4 with Hyperthreading etc. The specs on the P4 sounded like should have blown the doors off of the Athlon. Nope. The Athlon was surprisingly quick. It's not even a modern Athlon, either. It's got a 266 bus.
I don't think LW cares too much about the bus. Then again, my results aren't exactly conclusive so take what I've said with a grain of salt. I had to admit, though, I was pretty stunned by my 1.5 year old Athlon's performance. (happy too!)
"Derp de derp."
I'll second that... I've got a dodgy-ass Gigabyte KT266 board with an XP2100+ on it, and it's rock solid in Slackware 9 and 2000 Pro.
The weird part is it's running off a really old AOpen 230W ATX PSU, which I would think was massively under-spec considering the box has got a Geforce 5600 Ultra, 1.5 Gig of DDR2100, 2 SCSI cards, 2 burners, and a Seagate SX44..... 'Elite' (47 gig double height USCSI disk, eats about 50W on its own....).
This box has NEVER crashed on me, except when I booted a 2.5.something kernel and it didn't like the GART driver (I think.)
Athlon instability is either a myth or due to bad luck.
You're doing it wrong.