TXE trouble

Special Topics Intel Management Engine
1

Hi, I updated my ZBOX CI323 nano (Intel Celeron N3150) BIOS to the latest version 2K160913 and it went fine, but a day or so later I noticed the CPU is slower than usual. It’s not using the turbo clock anymore.
Looking for errors I noticed in device manager the Intel Trusted Execution Engine has a yellow exclamation mark

and in event log there is an error for TXEIx64


Also the TXE Information section inside the BIOS settings shows "N/A" for both Sec RC Version and TXE FW Version, this was definitely not the case right after updating the BIOS.
I already tried re-flashing the BIOS again, it shows no errors but doesn’t help.
The change-log specifically mentions changes to the update procedures for ME / TXT updating, I’m wondering if this BIOS is just bad or the TXE unit dying on its own is a thing (would be quite the coincidence).
Is TXE even related to the turbo clock not working anymore?
I’m on Windows 10 Pro.
https://www.zotac.com/files/download/by_…type=238&os=All

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
 
-------[ ME Analyzer v1.6.7 ]-------
            Database r70
 
File:     B301P018.bin
 
Firmware: Intel TXE
Version:  2.0.2.2092
Release:  Production
Type:     Region, Extracted
FD:       Unlocked
SKU:      1.375MB
VCN:      4
PV:       No
Date:     06/10/2015
FITC Ver: 2.0.2.2092
Platform: BSW/CHT
Latest:   No
 

I already tried TXEInfo/TXEManuf but they error out, even the efi-versions.
[Guide] Clean the Engine Initialized DATA section of a SPI dump would this help me? I'm not sure which section to follow for Braswell.


/e:
I noticed they use fpt in their flash script
1
2
3
4
5
6
7
8
9
10
11
 
set AfuName  AfuEfix64.efi
set BIOSName B301P018.bin
set FptName  fpt64.efi
 
set BlkNum 0
if exist fs%BlkNum%:\%AfuName% then
fs%BlkNum%:\%AfuName%  fs%BlkNum%:\%BIOSName% /p /b /n /x /r
fs%BlkNum%:\%FptName%  -f fs%BlkNum%:\%BIOSName% -txe
fs%BlkNum%:\%FptName% -greset
goto success
endif
 

When I run the fpt64 -f B301P018.bin -txe part on its own it says Platform: Cherry Trail.. that can't be right, can it?
2

First try "fptw -greset" command and check if the problem is solved after the reset. Otherwise, follow the cleanup guide (Pre-Skylake) and reflash the SPI image with a cleaned+configured TXE. Your problem seems to be TXE related. I suspect Zotac used a previously initialized dump at that BIOS. Both BSW and CHT use TXE 2 firmware.

3


Thanks for the advice, -greset is already part of the official flash script, so it didn’t help separately either.
Tech support came back in the meantime, claiming probably a bios ic issue or a corrupted bios, suggesting I replace it under warranty.

So I figured I might as well go ahead with your flashing guide, took it off power completely to move it near my other PC where it just sat for a while and what do you know… it’s working again, TXE Version Info is back in BIOS, no complaints in device manager and turbo boost is functional again. What the actual fuck, I literally did not even do anything to it?

/e: HAHA, as I was sending, this happened:

4

You did do something. Removing the power from the system for more that one minute is equal to a -greset so that’s how the TXE got fixed.

The visual corruption problem is definitely not related. It should be gpu driver related. Otherwise, if it happens outside of the OS, it’s gpu hardware’s problem and you can use the warranty in such case.

5


But then -greset should have fixed it much sooner.
Bloody technology, when did turning it off and on again actually become the proper fix for things? I’m just going to pretend this never happened.

/e: Turns out to be a problem with S5 Deep Sleep Support, if its enabled in BIOS and the PC is shut down, TXE will disappear until power is removed.