Sphere478 wrote on 2024-05-30, 04:07:

I've never been a fan of putting a MMX at 3.3 or 3.5v. but yes, I have heard of people doing it. I think it's stupid.

Intel MMX chips were manufactured at 0.35 μm. 3.3V is normal operating supply voltage for that node https://en.wikichip.org/wiki/350_nm_lithography_process There is nothing stupid about doing this.

rasz_pl wrote on 2024-06-28, 22:57:

Sphere478 wrote on 2024-05-30, 04:07:

I've never been a fan of putting a MMX at 3.3 or 3.5v. but yes, I have heard of people doing it. I think it's stupid.

Intel MMX chips were manufactured at 0.35 μm. 3.3V is normal operating supply voltage for that node https://en.wikichip.org/wiki/350_nm_lithography_process There is nothing stupid about doing this.

I wonder what the dissipation is at that voltage? 🤔

Here were some specs I found previously for the mmx series, as well as a comp. with a pentium 133 since that's the official max for my socket 5 board, plus some K6 cpus.

Power dissipation looks substantially higher with the mmx chips at similar voltage compared to the p1 non-mmx., though lower than the K6. I agree, its best to play it safe with an interposer or overdrive, especially for anything rare or valuable. There's clearly a reason Intel didn't market this stuff as Socket 5 upgrades. Thankfully both the cpu and board I'm playing with are quite common. As for temp, it felt cooler to the touch after hours of use than my pentium 133 did. I've enclosed it in the case at this point, which is a circa 1995 desktop case without any significant venting, and ran it for another 4 hours without an issue. Neither the case nor the air coming out of the vents or psu fan felt warmer than usual by the end of that session.

Pics for fun. The benchmarks are more than twice as fast as this PC performs with it's stock configuration.

Just coming back now that it's been 4 months to report that I've encountered no particular issues with using the split voltage 233 mmx CPU on the single voltage socket 5 board during this time.