AMD Optimal BIOS settings + Overclocking Guide
JMke writes "Here's a step by step guide on how to get the most out of your AMD setup. Overclocking tips and BIOS tweak settings discussed, as well as an overview of the more popular overclocking tools. Start your overclocking here!" Lots of good info here for getting the last bit of performance out of your system while also watching out for dangers that could fry your processor.
That's bizarre. I have an 1800+ on my desktop, with the stock heatsink/fan.
I've never seen it go above 45C, even after hours of playing something like BF1942. It idles somewhere around 33C. My video card is a heatsink-only Radeon 9000, and I've got 3 hard drives in the case - only a regular mid-tower too. No case fans at all, no hard drive fans.
What do you keep your room temperature at? Or could different motherboards have different heating effects? I always figured a setup like mine should run really hot, but it doesn't from what you're describing.
Any experts out there wanna field this one?
Endless arguments over trivial contradictions in books written by ignorant savages to explain thunder in the dark.
"case modding and quiet PCs"
That's the real key.
How about a guide on underclocking something far enough that it can be passively cooled? Most chips are so much too fast that it hardly matters. (As always, results may vary)
Or, what about the power-on-demand-or-whatever-they-call-it that mobile chips have? How about a guide on setting up a heat-sensitive fan so your system is silent when idle?
Motherboard temp sensors can be off a lot. On my Abit NF7-S they had to fix it in a bios update because it was off by up to 10C. Your 1800+ might also have the Palomino core which ran extremely hot. A good HSF and arctic silver are very good at lowering temps. I have a 2100+ running at 3200+ speeds with full load temp being 52C.
It's pretty much what I did when I bought my new system a few months ago (Barton 2500+, 8RDA+ mobo). Their recommendations match the experience I had. I used (and deleted, stupid stupid me) a spreadsheet to track my changes as I stepped through FSB, multiplier, voltage and memory settings until I found my maximum performance.
Then I backed off by about 10%. I'm more interested in overall reliability and longevity than maximum performance, since I want this system to last for at least three years as my primary.
-Thomas
I for one don't need the horsepower of a top-end CPU. I have a KT266a-based board (max FSB=133/266) and I put an Athlon-xp 2500/333 into it.
My CPU is underclocked from 1.83G to 1.46G, it dissipates about 45W, which is about the same as a G4, and HALF what a modern P4 drops. It's stable as all hell and I'm very happy with the speed.
I do the same thing to my G3/450, I use it as a fileserver so the 450MHz is totally wasted. I turned it down to 300MHz with less than 2% 'real world' performance difference from the client machines. It also generates less heat and uses less power now.
Any of you living on your own and paying electric bills would be well-served by underclocking, as the VAST majority of our CPU cycles go to waste anyway.
"Sometimes, I think Trent just needs a cup of hot chocolate and a blankie." -Tori Amos on Nine Inch Nails
While the 2600+ is a good overclocker, it's a thoroughbred model. You'd be better off getting a 1700+ if you want a thoroughbred. They'll both hit about the same speed, which is usally 2.5-2.7GHz on some good air cooling. The Barton 2500+, though, is the lowest end of the Barton line of processors, and generally hit about 2.4GHz with air cooling, which actually equates to about 3400+. With some exotic cooling, like phase change, you will probably get a few hundred MHz higher out of either chip.
hey!
What are the applications that I can use to check cpu temperature/fan speed under linux?
preferably something that is common on most distros, as I'm using knoppix right now and running off the cd, so apt-getting it won't work with my current setup.
If I had windows I could check temps during daily bootups in the bios, or with the windows software that came with the motherboard, but since I'm running knoppix, it's been almost three months since I last rebooted, and I don't know which app to use for hardware monitoring.
Suggestions?
There are different methods the motherboard can use to measure the die temperature. One is to place a thermal diode on the motherboard in the center of the socket beneath the CPU. Other motherboards report the temperature directly from the CPU's internal on-die thermal sensor.
r 7a-r/sensor_small.jpg
Usually the report from the on-die sensor is 10-20C higher than the external sensor reports.
Here is what the in-socket thermal diodes look like: http://www.overclockers.com.au/techstuff/r_abit_k
Whenever the offence inspires less horror than the punishment, the rigour of penal law is obliged to give way...
Believe it or not, one of the most common problems with frying a processor is applying way too much thermal grease to the core. Some people get crazy with this when attempting to overclock. It only takes a couple of drops, not a layer of grease to do the job. In fact, too much grease does the exact opposite and acts as an insultor which causes the processor to fry even faster.
That's bizarre. I have an 1800+ on my desktop, with the stock heatsink/fan.
I've never seen it go above 45C, even after hours of playing something like BF1942. It idles somewhere around 33C. My video card is a heatsink-only Radeon 9000, and I've got 3 hard drives in the case - only a regular mid-tower too. No case fans at all, no hard drive fans.
I built a couple systems recently as my new servers, and bought exactly the same parts at the same time. I got a Asus A7N8X Deluxe motherboard and Athlon XP 2000 for each system. Same memory, heatsink, etc, etc.. They are even in the exact same cases, except one is higher on the rack than the other (by only 1 foot). Whenever I check the temperatures on them, the CPU temp is always about 35C on one, and 40C on the other. ALL the time, no matter what I am doing. I have tested it with SETI@Home or other number-crunching type applications, and they STILL keep the same difference in temperature. I think it all comes down to what batch your chip comes from. And buying 2 of the same chips from a company doesn't always ensure that they are from the same batch. Not sure why else they would show a difference in temperature all the time.