The K6-2+ 500 will be much faster especially when clocked at 550 or 600 MHz, the K6-2+ is ~2% slower than the K6-3(+) clock for clock.

Haha, Wikipedia is funny! https://en.wikipedia.org/wiki/AMD_K6-III

Nevertheless, the K6-III 400 sold well, and the AMD K6-III 450 was clearly the fastest x86 chip on the market on introduction, comfortably outperforming AMD K6-2s and Intel Pentium IIs.

Completely ignoring the fact that Intel had a PIII-500 out at the same time:
http://www.anandtech.com/show/254/7
Not to mention the fact that even a PII-400 is faster than the K6-III at 450 MHz in games: http://www.anandtech.com/show/211/4
They also included the K6-2 in some of those those tests, and the K6-III is clearly a lot faster than even the 500 MHz K6-2.

Skyscraper wrote:

The K6-2+ 500 will be much faster especially when clocked at 550 or 600 MHz, the K6-2+ is ~2% slower than the K6-3(+) clock for clock.

Thank!. I consider myself lucky to find this one cause the III+ are really difficult to find. Here and there many K6-2 mostly 400/450 are not that difficult to find but + versions really was rare even in their times.

Scali wrote:

Haha, Wikipedia is funny! https://en.wikipedia.org/wiki/AMD_K6-III […]
Show full quote

Haha, Wikipedia is funny! https://en.wikipedia.org/wiki/AMD_K6-III

Nevertheless, the K6-III 400 sold well, and the AMD K6-III 450 was clearly the fastest x86 chip on the market on introduction, comfortably outperforming AMD K6-2s and Intel Pentium IIs.

Completely ignoring the fact that Intel had a PIII-500 out at the same time:
http://www.anandtech.com/show/254/7
Not to mention the fact that even a PII-400 is faster than the K6-III at 450 MHz in games: http://www.anandtech.com/show/211/4
They also included the K6-2 in some of those those tests, and the K6-III is clearly a lot faster than even the 500 MHz K6-2.

Well, the + versions were probably oriented for mobile low cost solutions but I still think that the original K6-3 should have been the real and only K6/233 replacement just like the VSA-100 should have been the real Voodoo2 replacement.

The 100 MHz higher clock speed beats the 128 KB higher L3 cache 😀

What about say a K6-3 400 vs a plain K6-2 550? I'd mainly be interested in games.

mwdmeyer wrote:

What about say a K6-3 400 vs a plain K6-2 550? I'd mainly be interested in games.

Well I can't say about the 550Mhz version I didn't use it recently, but the K6-3 400 feels imho faster (considering we're talking about lower clock) than the K6-2 500Mhz both in win usage (es. browser) and games (Thief2, 3DMark).

mwdmeyer wrote:

What about say a K6-3 400 vs a plain K6-2 550? I'd mainly be interested in games.

You might find this one interesting: AMD K6-2 vs K6-2+ vs K6-III+ Comparison

Ok, I updated the bios and now posting with the K62+ cpu. One strange thing at 2.0v the 5AGM2 mobo could not boot it, but it can boot only at 2.1v. On the plate there's voltage at 2.0v spec. Will I destroy it with the 0.1v more or it's safe and probably the mobo give less power with jumpers settings?

2.1V is totally safe but do not use more than 2.2V

Skyscraper wrote:

2.1V is totally safe but do not use more than 2.2V

Thank. Actually it feels a bit faster than the K6-3 400, with browser seems to render faster the pages. Not a big jump but I can try at 550Mhz if it will be stable.

One thing I don't like... it feels less vintage..😁 My K6-3 400 is dated early 1998... this one early 2000...

I think you should be more worried about the motherboard than the CPU.
If the caps are old/poor quality, they may die when the CPU draws too much current.

Scali wrote:

I think you should be more worried about the motherboard than the CPU.
If the caps are old/poor quality, they may die when the CPU draws too much current.

Yeah I was thinking to that too. From my experience with this mobo the K6-233/3.2 seems to be the hottest running I felt similar (but i think a bit higher) to the K6-3 400/2.3.
I don't find the specifics wattage of this K62+ but I would bet about 20 watt max at 500Mhz similiar to the 2.2v 0.25u standard K6-2 500? I should be under the 30w socket 7 "limit".

According to this list: http://www.cpu-data.info/index.php?gr=15
They should be well under 30w.
K6-III goes up to 29w, so that can be dangerous: http://www.cpu-data.info/index.php?gr=16&lng=1

Yeah. Probably the K6-3 400 feels a bit cooler that the K6-233 cause being much faster, it work much less in the usual desktop enviroment. I can't imagine the 2.4v K63 at 450 (or 500).

philscomputerlab wrote:

mwdmeyer wrote:

What about say a K6-3 400 vs a plain K6-2 550? I'd mainly be interested in games.

You might find this one interesting: AMD K6-2 vs K6-2+ vs K6-III+ Comparison

Hey Phil, I have to agree with you on the mobo cache. Disabling it seems to have more adverse effects (if anything) than just leaving it on.

Did you try the 'performance boost' from the Win98SE unofficial service pack? It states it requires at least 256 MB of RAM to operate. I wonder if that would increase performance when cacheable memory is changed from 128 MB to 512 MB.

I doubt the K6-2+ will use more power even at 2.1V and 550Mhz than the non plus K6-3 do at stock 400MHz. I have really pushed my K6-3+ CPUs and looking at the heat produced I do not think they have used much more than 25-30W when runnig 600+ MHz with 2.2V. Its still a good idea to have some airflow over the VRM MOSFET and dodgy caps can always be an issue.

raymangold wrote:

Did you try the 'performance boost' from the Win98SE unofficial service pack? It states it requires at least 256 MB of RAM to operate. I wonder if that would increase performance when cacheable memory is changed from 128 MB to 512 MB.

Nope. Never tried the service pack. I don't see a need for it to be honest, but I might do a review one day, comparing before / after. But what would be a good machine to test this?