Yes, it is possible to modify the CPU microcode.
Yes, it is possible to modify the CPU microcode.
You explored ways to boost performance by altering CPU microcode, aiming to mimic another processor's speed. The idea involves downloading an X5690 microcode, adjusting clock values, and applying it to the system. Whether this is feasible depends on technical constraints, compatibility, and the availability of accurate microcode files.
It's feasible, though very risky... you'll likely have to address cooling as well. If I were you, I wouldn't spend my time on something with such high stakes and minimal benefits.
I haven't tried it myself, but I saw something similar before. It's important to remember there are more issues than simply changing the microcode. Microcode can be updated via BIOS or operating system. If you're using an older BIOS version, the OS might replace it. I think this is why non-K overclocking stopped working later because subsequent microcode disabled that feature. You'd also need to address it at the OS level to avoid this problem. In short, it's a lot of effort with little benefit. If you're determined, go for it, but I don't think it's worth the hassle.
The available info on "microcode hacking" is limited to a few sources. A webpage on Ars Technica discusses the topic, while another Stack Exchange question addresses processor microcode changes and their feasibility. Hackaday provides insights into modifying BIOS settings, which can influence CPU features like multiplier locking. Research from Ruhr University involved reverse engineering AMD microcode with some success. Overclock.net shares forum discussions and technical notes on motherboard configurations and overclocking techniques. A user reported overclocking dual X5650 CPUs using specific BIOS tweaks, though results varied.
I managed to adjust the BIOS settings on a G31M-VS2 motherboard that didn’t support unlocking the core multiplier. After tweaking microcode and dealing with blue screens, I was able to bypass it. However, when testing with an Xeon E5430, I discovered the board couldn’t handle the 10m power draw—so I’m limited to using the Xeon model with 8m. It was a frustrating process, but I enjoyed the challenge. P.S. The Xeon E5430 is a 771 processor. P.S²: The G31M-VS2 is a 775 board. P.S³: This machine is my guinea pig. If you need images, I can send them later.
the key point is not the challenge of breaking microcode, but its protection through signed microcode. if you update a BIOS with microcode that doesn't match the signatures, it could damage the motherboard.