I have a couple of old systems which both have AMD K6-2 CPUs. I recently swapped the CPUs around and at the same time put some fresh thermal paste in (that was really needed, it was like chewing gum). Please excuse the not-perfect pics. Here are the complete specs of the machines: System 1 (from around 1998): - AMD K6-2, stock 500MHz - Some random Compaq OEM mobo, says Foxconn UZI on it - SiS 530 chipset - Integrated graphics with 8MB VRAM - SDR SDRAM: 2x Toshiba 128MB 100MHz + 1x Dell 256MB 133MHz = 512MB (thanks to my friend for grabbing these from work for me to upgrade the stock 64MB) - Realtek RTL8139 10/100Mbit/s PCI network adapter, really low profile - Currently using a WD 40GB HDD - Mitac 110W PSU - Pic: System 2 (from around 1995, propably upgraded in ~1998): - AMD K6-2, stock 400MHz - Asus P/I-P55T2P4 (am I sick when I remember that?), Baby-AT form factor - Intel Triton 430HX chipset - Nvidia Riva TNT2 M64, PCI model with 32MB of VRAM - 2x EDO 32MB + 2x FPM 32MB = 128MB RAM - Creative SoundBlaster AWE64 ISA sound card - 3Com EtherLink III 10Mbit/s ISA network adapter - Currently using a Seagate 3GB HDD - Real Power 200W PSU (talk about overkill!) - Pics: WARNING! Do not blow chunks when you see the PCB color... You can also see the fan I'm using in these experiments in this last pic in the lower left corner. Studied a bit before starting. I found out that the Asus mobo on system 2 is relatively limited. Only supports bus speeds of 83, 75 and 66MHZ (and lower but I'm not interested in those); voltages of >=2.5V and with hacks down to 2.0V; and multipliers of 2.5-3.5 and 6.0 (the K6-2 interprets the 2.0 multi as 6.0). So basically I could only control the CPU speed with the FSB, and only in large steps. A very helpful and recommended site: http://web.archive.org/web/20040805122822/http://www.jump.net/~lcs/kalle/rev3.htm The Compaq thing, then, I didn't find any web resources, but examination of the board physically tells me that various bus speeds ranging all the way from 66 to 133MHz, all useful multipliers, and voltages <=2.3 (and then jumps to >=2.8V) are supported. Okay, let's get started with overclocking. But wait, I've already done it! Well I'll just explain what I've done then. Results so far: I started with system 2. Tried it on stock (2.2V; 6x66MHz, a beast of a configuration) after the CPU swap to test if it works. It sure did. 6x75 was a bit unstable so I bumped up the voltage to 2.3V. Wanted to go higher. Put it to 6x83, but the CPU didn't even start. Tried at 2.4, same result. Then tried at the maximum AMD allows (CPU can get damaged after this), 2.5V. Now it started, but it was just way too unstable, usually didn't even finish POST. I thought "Well I guess I can't go higher than 450MHz then " but later I decided to try at 2.7V (the mobo doesn't allow 2.6V for some reason). I worked abolutely fine! I'm a bit worried about the voltage though. I think I'm going to manually force it to 2.6V by using a resistor (read more on Kalle's page). Okay, onto system number one. Problems from the beginning. POST took ages! After a year it showed me a message: "The following configuration options were automatically updated: No physical disks present." even though the 40GB WD was connected and spinning. Then restart. After another year it showed me "The following configuration options were automatically updated: Disk1 WD 40GB". Restart once again, and yet another year waiting for it to POST. I thought the system was broken or didn't handle the higher FSB required by the 500MHz CPU. But I was wrong: after the third reset it booted succesfully to Windows and it was rock stable! Well, tried with a lower FSB, same result. After a while of tinkering I tried with another HDD, a Seagate 4GB, and POST took about 10 secs and it booted straight out. I guess the problem is somehow related to the 40GB disk (dying?). Okay, I thought, onto overclocking, and put the 40GB back and I'll just have to live with the 3-year POST time. Overclocking results: 112x4.5. I'm not impressed. The CPU is running at 504MHz... The problem is the limited voltage selection I spoke about earlier. It can't go any higher than 2.3V which is just not enough for 560MHz (112x5), the system is very unstable then. I'd just want to get it one notch higher (or maybe two-three notches if I aim for ~600MHz) FFS! (Or more specifically it can go higher, but it goes TOO high (2.8V)!) So, I guess that's it for today... Any comments, suggestions, questions are welcome! Also you can share any memories aswell. And I can take more pics if wanted. I'm open to ideas on what to do with system 1, I don't want to have it at just 4MHz over stock. Any electronic gurus here who would have suggestions how to measure the resistances of the voltage selectors, or is it even possible?