Antinomy

@Leeghoofd, I do have one with a screen if it changes anything. hwbot.org/submission/3834852_antinomy_cpu_frequency_celeron_533mhz_(mendocino_s370)_760_mhz

Well, rules were unclear about this one. And not sure whether there was a statement about such buffed up cards. There's much fun in finding exotic hardware and even knowing of its existence, completely different case than which is a total cheat. I do agree on all the rest you've said.

How much did they pay you?

It's not for HWBot validation, it's for OSBIS competition. Carry on.

Yesterday I've got caught a glitch from the submit page. It's quite an old one but annoying nevertheless. I think we should work hard pushing the hardware, not fighting with submission engine. Bug report: 1) Images selected for attachment are removed if you get a error message, both "verification screenshot" field and "pictures of your system" (except ones being prepopulated from previous submissions) . Adding the fact you have to scroll way down to re-check this, it's a nasty one. 2) internal exception errors are seen, "Property score threw exception; nested exception is java.lang.NullPointerException". And there are two of them, above and below the input field. Feature requests: 1) On CPU-Z result page we have two "CPU frequency" input fields. Can we grey-out the second one and dupe from the other as long as one of them gets filled? It'll be the best solution. Or at least make one input field. It's kinda counter-productive having to input the same thing twice. 2) We have a motherboard database, why don't we fill up motherboard vendor and chipset if you select a certain model? If it's not in the database, O.K. let them fill all three fields. But if it's in, you can prepopulate the other two from model name. 3) Last but not least - could we nail the "I've read the rules and agree to them till the next time they change". I understand why this thing got light when rules have changed. But this doesn't happen often. We could nail it once per benchmark and reset when benchmark rules or general rules have been change. It is annoying. TIA

Sucker, I've just noticed my screenshot didn't get attached despite I did. WTH

I didn't have a screenshot but when I saw a message from Mr. Scott, I had to rebench the setup to get one. The rules tell screenshot is required. I think a photo might work too but it's up to you. I'm afraid, can't help you with this one. Start checking voltages.

@DR4G00N what do you mean by pin 20? Try shorting PS_ON pin in ATX power cord: http://pinouts.ru/Power/atxpower.shtml you need to short pins 14 and 15. If your system boots fine with this, it's the multiIO chip dead (most likely). Two hours to the end and I've just found my WinXP CD and it's got read errors. A new level of extreme.

Were did you buy SL7MG? That's the right question, adapters are much easier to get. Making FSB Northwood work on P5K-E is impressive too!

What problems did you have? Unhuman Gumanoid =) excellent work, congratulations. Looks like highest Mendocino frequency.

On Asus CUV4X chipset ID is from PM133/PZ133 so CPU-Z can't detect it's timings despite them being the same as Apollo 133. I have the same situation.

I don't see a problem here The real problem is that about a half of my hardware isn't in HWBot database at all.

Thank you! Now it makes sense. And you're just in time before I've taken my Aladdin Pro 5 off the shelf

@Strunkenbold So, which non-Intel chipsets are allowed? Everything except Apollo 266 or only the ones listed on the stage?

I love this argument. No, got nothing. Doesn't mean I'll stick a webcam in my bedroom. And yes, Scotty did figure this was a joke.

As long as you don't send our likes to NSA, it's fine

Strangely enough, UCbench results give points though the whole benchmark was removed from HWBot. http://hwbot.org/benchmark/ucbench_2011/rankings?cores=2#start=0#interval=20

@TASOS More interesting over here: http://www.plasma-online.com/english/upgrade/tweak/fixes/fix_converter.html

A bit late but nevertheless. P55 core (classic Pentium MMX) has two multiplier pins BF 1:0. Two pins give us 4 combinations and they are 2x, 2.5x, 3x, 3.5x - 133MHz (not presented in desktop CPUs), 166MHz, 200MHz and 233MHz. So anything over 3.5x on P55 is software rubbish (Pentium Pro has the same problem, BTW and maybe Klamath). But the CPU frequency should be O.K. As for unlocked - Intel started fighting with overclocking and cut down higher multis on latter P55 and some P54C. I suppose it has been done by cutting (or not wiring) the package (plastic or ceramic) to the core. Early samples didn't have this making all multi available. There is a number of things CPU-Z can be better (showing onboard cache like it did on old versions but stopped) we might want a thread for that. There's been only but maybe a separate for old hardware would be nice, sometimes it's hard to find a work-around working single as it requires different hardware...

What's the point removing VIA Apollo 266 and leaving CLE266 (that's an Apollo 266 + Unichrome iGPU)? Considering SDRAM only. And Aladdin Pro V with SDRAM being allowed.

Nice score! I thought LN2 should give more than 250MHz over water.

Nice BIOS modding, congratulations!

Post some photos of the board. Did you put the BIOS chip the right way? There is a small crack both on chip and socket marking the way to insert them. Check out here: Badflash Bad Flash

Fastest Aquamark on Socket 3. That's gonna help you kill [quite] some time

But why? SDRAM wasn't too popular on s423. I think I have 2 or 3 DDR boards but none based on i845 SDRAM. My thoughts: 3DMark 2001 SE socket 370. Or maybe even socket 370/socket A (Athlon non XP, i.e. before Palomino). Nice overall system performance.