In general, all CXT K6-2s seem to manage 50mhz. Best overclockers are lowest core volts, highest heat range. Beware of very early non-CXT 400s which were basically rare freaks to manage 400 at all before the CXT changes. They are going to be mid 98 date codes though, the plain jane chompers.

I can tell you my worst - it was a K6-2 533 and it took a voltage boost to get it to run at 550 MHz! Can't tell you the stepping.

Repo Man11 wrote on 2026-05-24, 22:46:

I can tell you my worst - it was a K6-2 533 and it took a voltage boost to get it to run at 550 MHz! Can't tell you the stepping.

2.4 volt chips, 533 and 550 chips were the worst overclockers

Back when the K6-2 550 was current, I had mine running at 660 (120x5.5) on an ASUS P5A with air cooling. It was the 2.2v version.

With extreme cooling, there have been instances of people about to get them to 720Mhz, don't remember if anybody successfully got one to go higher than that.

My AMD-K6-2/333AFR 2.2V CORE/3.3 V I/O A 9830EPAW doesn't even run stable at 350 MHz.
BitWrangler, you're right. It is a pre-CXT.

Sounds like it's not a CXT

Disruptor wrote on 2026-05-25, 22:26:

My AMD-K6-2/333AFR 2.2 V CORE/3.3 V I/O 9830EPAW doesn't even run stable at 350 MHz.

Reminds me of the older 300/66 k6 chips that became unstable on 100mhz FSB and required the strange 3.52 VIO despite the 2.2 vcore

I don't know if my 266 was newer than that then, had it up to 4.5x75 before some K6-2-400 went on sale and I grabbed one of those, which ended up running 6x75.

rmay635703 wrote on 2026-05-26, 01:22:

Disruptor wrote on 2026-05-25, 22:26:

My AMD-K6-2/333AFR 2.2V CORE/3.3 V I/O A 9830EPAW doesn't even run stable at 350 MHz.

Reminds me of the older 300/66 k6 chips that became unstable on 100mhz FSB and required the strange 3.52 VIO despite the 2.2 vcore

My AMD-K6/300AFR 2.2V CORE/3.45V I/O A 9827BPHW runs stable at 3x100 MHz.

Disruptor wrote on 2026-05-26, 04:48:

rmay635703 wrote on 2026-05-26, 01:22:

Disruptor wrote on 2026-05-25, 22:26:

My AMD-K6-2/333AFR 2.2V CORE/3.3 V I/O A 9830EPAW doesn't even run stable at 350 MHz.

Reminds me of the older 300/66 k6 chips that became unstable on 100mhz FSB and required the strange 3.52 VIO despite the 2.2 vcore

My AMD-K6/300AFR 2.2V CORE/3.45V I/O A 9827BPHW runs stable at 3x100 MHz.

Is it marked 66mhz only?

There was a period of time where certain chips were marked 100mhz FSB and others were marked with a different model number that indicated a 66mhz FSB limit

The earlier 66mhz chips that existed before pc100 was common don’t count.

BitWrangler wrote on 2026-05-24, 21:46:

In general, all CXT K6-2s seem to manage 50mhz. Best overclockers are lowest core volts, highest heat range. Beware of very early non-CXT 400s which were basically rare freaks to manage 400 at all before the CXT changes. They are going to be mid 98 date codes though, the plain jane chompers.

Oddly there were somewhat rare earlyish 2.4 volt 475mhz rated parts.
Those must have been some real dogs releasing before faster 2.2 volt chips were common.

Bringing me down memory lane with all the god aweful strange k6-2 clockings, there was a 466mhz rated chip, strange FSB.
Oddballs like 366, 380, 433, 475, 533, 570 never made much sense

rmay635703 wrote on 2026-05-26, 17:06:

Disruptor wrote on 2026-05-26, 04:48:

My AMD-K6/300AFR 2.2V CORE/3.45V I/O A 9827BPHW runs stable at 3x100 MHz.

Is it marked 66mhz only?

No.

The 333 is a 95FSB part. I've just checked and it's either 66 or 95 FSB.