Jump to content
HWBOT Community Forums

Recommended Posts

  • Crew
Posted
5 hours ago, frag_ said:

Is 478-775 adapter allowed (hope it's not because it would give huge advantage and it's hard to get)?

I suppose yes but I think it'll be useless.

  • Crew
Posted (edited)

You say you want a revolution?

Do you see anything strange over here?

Spoiler

TL;DR:
So all you need to do is solder a jumper from AL2 pin to Ground.

So here's the explanation. CPUs starting from P3 Mobile have different performance states. P4 have several technologies for power-saving (C1E, EIST) but I'll talk about something different. Thermal monitor. The one that prevents CPU from damage in case of overheat. The older TM1 is known as throttling when a CPU skips a percent of clocks. The newer one, TM2 drops the multiplier to minimum possible for given core and voltage to a lower state.
image.png.f2a78b0cde813a35d940310d6f59622e.png

image.png.5d8ca3c21830bb56067dfa94cfb07622.png

So we can see that the PROCHOT# signal triggers a multiplier drop but the signal is present only in a high temperature situation... or not?

image.png.101f7040f23f86ec4db08da340fdf431.png

So after all the PROCHOT# signal not only works as an output letting the system aware of a CPU overheat condition but also as an input to be aware of a system (VRM, etc) overheat condition. Once PROCHOT# is triggered (pulled low), the CPU uses it's safety mechanisms - TM2 or TM1 (whichever is enabled by BIOS or software). If it's TM2 it'll drop multiplier to its minimum, hence, 14x for Prescott.
This trick works on all LGA775 CPUs even without C1E like Celeron 3xx, 4xx and so on.

Now as you see, 478 adapters are pretty much useless for this competition. ?

snaphsot0003.png

Edited by Antinomy
  • Like 3
  • Thanks 3
Posted (edited)
9 hours ago, Antinomy said:

You say you want a revolution?

Do you see anything strange over here?

  Reveal hidden contents

TL;DR:
So all you need to do is solder a jumper from AL2 pin to Ground.

So here's the explanation. CPUs starting from P3 Mobile have different performance states. P4 have several technologies for power-saving (C1E, EIST) but I'll talk about something different. Thermal monitor. The one that prevents CPU from damage in case of overheat. The older TM1 is known as throttling when a CPU skips a percent of clocks. The newer one, TM2 drops the multiplier to minimum possible for given core and voltage to a lower state.
image.png.f2a78b0cde813a35d940310d6f59622e.png

image.png.5d8ca3c21830bb56067dfa94cfb07622.png

So we can see that the PROCHOT# signal triggers a multiplier drop but the signal is present only in a high temperature situation... or not?

image.png.101f7040f23f86ec4db08da340fdf431.png

So after all the PROCHOT# signal not only works as an output letting the system aware of a CPU overheat condition but also as an input to be aware of a system (VRM, etc) overheat condition. Once PROCHOT# is triggered (pulled low), the CPU uses it's safety mechanisms - TM2 or TM1 (whichever is enabled by BIOS or software). If it's TM2 it'll drop multiplier to its minimum, hence, 14x for Prescott.
This trick works on all LGA775 CPUs even without C1E like Celeron 3xx, 4xx and so on.

Now as you see, 478 adapters are pretty much useless for this competition. ?

snaphsot0003.png

It seems this could also be possible without any hard mods. Using the Intel Integrator Toolkit & D975XBX2KR, the multiplier can be forced to 14x through BIOS editing. I have just tested it with success:

EDIT: Nevermind, in-OS it is still 20x multi...

IMG_20221101_180758007.jpg

IMG_20221101_180806216_HDR.jpg

Edited by MachineLearning
  • Like 1
  • Crew
Posted
39 minutes ago, MachineLearning said:

Nevermind, in-OS it is still 20x multi...

Because this BIOS options works with EIST and it's blocked on Celeron CPUs. BTW, any interesting timings available through Intel Integrator toolkit on your Bad Axe?

  • Like 1
Posted (edited)
4 minutes ago, Antinomy said:

Because this BIOS options works with EIST and it's blocked on Celeron CPUs. BTW, any interesting timings available through Intel Integrator toolkit on your Bad Axe?

That makes sense. :)

Unfortunately no, nothing unusual. Same RAM timings available as usual. I can PM screenshots if you're interested. Also have details on the board/other changes here: https://warp9-systems.proboards.com/thread/1838/intel-bad-axe-d975xbx2-kr At the bottom is a changelog, including all timings I could alter. And a link to a .Zip including ITK 4.0.x, .ITW file, .BIO, everything necessary.

Edited by MachineLearning
  • Like 1
Posted (edited)

Awesome i searched datasheet for something like this cause i knew 14x multi exists, just wasn't sure how to do it... Guess i just suck at finding the things i want first time around ?

Anyways this will be so much better now as any celly will do, no need for unicorn hunting...

Edited by TAGG
  • Crew
Posted
3 hours ago, TAGG said:

any celly will do, no need for unicorn hunting...

That's exactly why I disclosed it. I didn't invent this trick, learned it years ago from one Russian BIOS guru.

  • Thanks 1
Posted

@Antinomy Did you find any way to fix hwinfo freakout? my hwinfo dispalys correct multi but bus speed readout is completely trashed with this mod, CPU-Z works though...

But effi is there that's for sure :) sub 1m easy

  • Like 2
  • Crew
Posted
2 hours ago, TAGG said:

Did you find any way to fix hwinfo freakout?

I don't use HWinfo usually. CPU-Z works fine and it's good enough for me. But yeah, I suppose some timers break that are used by some utilities.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...