| View previous topic :: View next topic |
| Author |
Message |
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
 Posted: Sun Apr 21, 2019 12:42 pm Post subject: Methods for dealing with unstable ES chips? |
 |
|
I recently bought a whole bunch of unmarked ES i7 1st gen mobile chips, and I've got a few interesting ones.
Some of them have sharpie marking "A2", which would make them very early stepping, as cpu-world only shows B1 revision ES chips.
This was the CPUID dump from one that I got working
The chips are horrendously unstable in normal Windows. I cannot get Win10 to boot normally (locks up with the spinning wheel thing), the installers for windows 10 and 7 both lock up (mouse stops moving, everything freezes).
I have been able to successfully boot linux and win10 safe mode + networking (how I managed to get the CPUID uploaded). I am unable to get cpu-z to run happily in safe mode, same goes with HWinfo64 (was hoping it would get the Q-code)
Any suggestions for getting a cpu-z validation or getting some program to extract the q-code? I've got a few of these chips working, and they all behave the same.
My test system is an Alienware M15X with the A05 bios. I have tested multiple other chips in this machine with no problems.
EDIT: LESSON LEARNED! ALWAYS UPDATE THE BIOS FIRST!
_________________
My collection
Last edited by crazybubba64 on Sun Apr 21, 2019 7:24 pm; edited 1 time in total |
|
| Back to top |
|
 |
frag_
Joined: 17 Nov 2008
Posts: 4015
Location: Estonia
|
| Posted: Sun Apr 21, 2019 3:26 pm Post subject: |
|
|
Maybe installing very lightweight win like XP can help?
And how it's possible HWinfo64 can extract q-spec?
To the best of my knowledge none of i7s contain any sspec info. |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Sun Apr 21, 2019 5:04 pm Post subject: |
|
|
| frag_ wrote: | | And how it's possible HWinfo64 can extract q-spec? |
Beats me, it might just look up the specs and make a good guess? I can check a couple of my other newer and older ES chips to see if it's reliable.
I know xsecret mentioned something about the internal serial number having a way to identify the q-spec with his LGA chip tester thing.
If it's not a reliable way, then I've got a whole bunch of chips with no q-spec now lol.
| frag_ wrote: | | Maybe installing very lightweight win like XP can help? |
I hadn't thought of this, might run into some driver issues, but I'll give it a shot!
In other news, I had one of the chips almost work, as it made it into windows, and I was able to get cpu-z to start, but the machine locked up soon after. 
_________________
My collection |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Sun Apr 21, 2019 5:24 pm Post subject: |
|
|
Just tried XP, got a BSOD during the install
EDIT: STOP: 0x0000007E (0xC0000005, 0x808C4E23, 0xF78A2234, 0xF78A1F30)
_________________
My collection |
|
| Back to top |
|
|
cuttingedgecs
Joined: 08 Oct 2017
Posts: 1764
Location: Australia
|
| Posted: Sun Apr 21, 2019 5:34 pm Post subject: |
|
|
| You could install Open Dos 7.02 and use the built in debug program to send machine code CPUID instructions and read the registers. The less complex code might be more stable on an ES chip than Windows. You can get the model and stepping information that way, but I agree with fraq_ that sspecs aren't stored in the chip. |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Sun Apr 21, 2019 5:57 pm Post subject: |
|
|
Noted. I did make a CPUID dump using the tool on this website, doesn't that extract all of the raw info? The raw CPUID data is here
_________________
My collection |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Sun Apr 21, 2019 7:01 pm Post subject: |
|
|
Did a BIOS update and got it to boot successfully.
cpu-z validation
HWinfo64 reports it as a "Q1TK", not sure what to think of it though. It also reports the microcode as "FFFF000D" and the TDP at 62W. The base clock in HWinfo64 is jumping all over the place though (between 80 and 130MHz), however cpu-z simply reports a 132MHz base, which is expected.
Base clock is 1.33GHz, boost is 1.60GHz. 4 cores, 8 threads.
A2 stepping.
EDIT: HWinfo64 reports the chip as unlocked! Gonna see if the instability was due to a BIOS issue.
EDIT2: Yep, it was a bios problem, chips are running fine now.
_________________
My collection |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Sun Apr 21, 2019 7:34 pm Post subject: |
|
|
Tried some overclocking, chip didn't like it all.
Makes sense though, XTU reports the chip as a 62W TDP (which is kinda insane for a laptop part) and a turbo boost current at 62A (also INSANE for a laptop part)
_________________
My collection |
|
| Back to top |
|
|
Mr.Scott
Joined: 15 Jul 2014
Posts: 267
Location: Upstate NY USA
|
| Posted: Sun Apr 21, 2019 7:43 pm Post subject: |
|
|
| Quote: | | Intel Mobile Core i7 i7-820QM 1733/8M Q1TK ES (Clarksfield) mFCPGA8-988 ENG SAMPLE |
Vegeta sold these a while ago.
http://www.cpu-world.com/forum/viewtopic.php?t=26355
Last edited by Mr.Scott on Sun Apr 21, 2019 7:47 pm; edited 1 time in total |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Sun Apr 21, 2019 7:47 pm Post subject: |
|
|
They must be VERY very early 820QM's then, as the TDP is crazy and the chips are unlocked. The only thing they share is the 8M cache. Base clocks are much much lower too.
It could be that HWinfo64 compares some value to a known database of part codes?
_________________
My collection |
|
| Back to top |
|
|
wren4777
Joined: 13 Dec 2016
Posts: 571
Location: Litija, Slovenia
|
| Posted: Sun Apr 21, 2019 8:22 pm Post subject: |
|
|
| crazybubba64 wrote: |
It could be that HWinfo64 compares some value to a known database of part codes? |
Yes, that's what HWinfo does if I remember correctly.
For what it's worth, always be careful when updating the BIOS on a board you want to run ES on. Make sure that the correct microcodes are present and that the ME version isn't one new enough to disallow booting pre-retail steppings.
I think the issue you had with your chips was that the microcodes weren't present in the original bios, but were in the upgraded one. This is a common issue with ES (I got my QL32 for 1/3 the price they go for because the seller thought it was the chip's fault it was freezing when the board just didn't have the microcodes  )
Also A2 stepping would make sense, there are A2 desktop chips with the same architecture that also don't report their stepping in CPU-Z (Q2AP and Q2AL I believe). The CPUID matches between those two chips and yours, as well.
_________________
My WTB List
My collection
Have a nice day!  |
|
| Back to top |
|
|
|
![[ Hidden ]](https://cdn.cpu-world.com/forum/templates/CPUWorld/images/lang_english/icon_email.gif "[ Hidden ]"){kind=icon}
