CPU-Z Validation Changes, bogus dumps removed.

[Charles Wirth]

Hi guys,

Just a note to tell you I just changed some steps of the internal validation check. Now, to get your dump validated, you need a clean frequency over time ("time" is, of course, not an hour but 2-3 seconds max.). This was added in order to reject all dumps where the CPU was throttling or unable to keep its rated frequency more than some milliseconds. Some guys spammed the validation key as long as needed to get a result "validated". We don't think a quarter second frequency might be called a "validated overclocking", so I added that check.

This issue is mainly affecting AMD K10 CPU (Intel doesn't seems to be concerned). No problem found so far with the Bulldozer OC Contest currently in progress.

PS : many bogus dumps were removed from world records.

Source - http://forum.canardpc.com/threads/60...h-overclocking

[bassplayer]

I can see this fixing QPI error runs on X58, but 90% of WR's aren't going to be valid anymore.

Will they just get the Red X now? I assume the bot will continue to accept those?

[massman]

As for the bot question: problem with the Red X-mark is that it's also given to perfectly valid scores.

[ChanceCoats123]

I'm confused as to what the red-X even means...

It's never made sense to me, to have three different responses to your checksum. It should either be 100% valid, or if there is even a hint of in-validity in the cpu-z file, then it's not validated. There should be no "in-between" valid and invalid.

[Sunfire]

"Bulldozer OC Contest" ???

[EniGmA1987]

"Bulldozer OC Contest" ???

^^^

[PatRaceTin]

hmmmmmmmmm interesting

[BossBorot]

indeed

[dinos22]

As for the bot question: problem with the Red X-mark is that it's also given to perfectly valid scores.

no good

seems like a bit of a silly rule this. Who makes this rule and why is a fraction of a second stability worse than 3 second stability and to whom? Why dont we then make it 1 minute stability or 1 hour, where do you draw the line? Pretty silly IMO

if its max frequency benching who cares if it's a nanosecond or 3 seconds seriously

[BeepBeep2]

no good

seems like a bit of a silly rule this. Who makes this rule and why is a fraction of a second stability worse than 3 second stability and to whom? Why dont we then make it 1 minute stability or 1 hour, where do you draw the line? Pretty silly IMO

if its max frequency benching who cares if it's a nanosecond or 3 seconds seriously

I care about this a lot, due to HTT float (only affects CPU-Z, I can check with any other program for example THG Clock Tool and it stays still) on my motherboard about half of my validations are invalid.
I sat in windows for a while at that clock? Completed Sandra bandwidth/latency for a competition at 1990 6-6-6-17, (took about 3 minutes) CPU-Z dump was still invalid...

[dinos22]

how much does your HTT "float around"?

[Mafio]

completely agree with dino.

if CPUZ is the only program suffering from FSB fluctuation they should solve the problem, not just set this silly 3 seconds rule.

[[XC] Oj101]

That's ridiculous, CPU-Z is now a stability tester for three seconds? Wow, how long before CPU-Z starts putting 100% load on all cores because -

We don't think a 0% load might be called a "validated overclocking", so I added that stress.

Absolutely silly to be honest, it's kind of like moving the target halfway through the race. FAIL

[BeepBeep2]

how much does your HTT "float around"?

2.2-2.3 Mhz HTT, it only shows in CPU-Z. It happens in both Vista/XP 32-bit + Win7

[Kayin]

Very familiar with this on AMD platforms. My wife's PC fails validations because of it. And it's only CPU-Z that it bounces around in. It seems to me the problem is the program, not the machines. So what gives?

[BeepBeep2]

Very familiar with this on AMD platforms. My wife's PC fails validations because of it. And it's only CPU-Z that it bounces around in. It seems to me the problem is the program, not the machines. So what gives?

Obviously its to make it harder for you to beat old frequency records or even match them on better hardware, and to fix the problem by adding more problems.

[[XC] gomeler]

lolwut. Sounds to me like it is a band-aid to fix an underlying problem.

[G H Z]

Obviously its to make it harder for you to beat old frequency records or even match them on better hardware, and to fix the problem by adding more problems.

Get serious, Franck was very clear in the motivation for making the change. If it helps clean up the rankings it's a good thing, we'll still see 8+GHz shots with the new version, trust me.

[BeepBeep2]

Get serious, Franck was very clear in the motivation for making the change. If it helps clean up the rankings it's a good thing, we'll still see 8+GHz shots with the new version, trust me.

Why can't he just fix his software first? I'm pretty sure almost all of my validations will now be invalid due to CPU-Z's inability to read the correct clockspeed.

[V2-V3]

Interesting

[der8auer]

what the?

I totally agree with dino. What a stupid rule to set. Since when is max clock related to stability?

[bassplayer]

Didn't you hear... To hit a clockspeed it has to be stable for 3 seconds. That's so obvious how didn't you know that?

[/sarcasm]

[flesheatinvirus]

Absolutely, get your straight first and perhaps this would not happen in the first place.

[sin0822]

this is dumb rule.

[l0ud_sil3nc3]

this is dumb rule.

agreed

max frequency should not require a time limit