Intel (Converged Security) Management Engine: Drivers, Firmware and Tools (2-15)

[gihan]

many thanks bro I’ll try and let you know thanks!!!

[Fernando]

@Horstfuchs

Can you verify the statement made in the following post on Guru3D?
A user claims that timer resolution optimization was removed in the more recent releases of the MEI driver. Is this true?

Today I have done a benchmark comparison test with my Z97 system running in AHCI mode.

Update (dated 09/18/2015):
Since I got interesting results while testing different Intel MEI drivers, I have tested later on additionally the Intel MEI driver v10.0.30.1058, which is the last one from the v10 series, and the Intel MEI driver v11.0.0.1157 WHQL, which is the latest from the v11.0.0.115x branch.

Here are the results sorted from the left to the right according the release dates of the Intel MEI drivers (v10.0.30.1058 WHQL > v11.0.0.1144 BETA > v11.0.0.1157 WHQL > v11.0.0.1166 WHQL):

Performance ranking:
1. Intel MEI driver v11.0.0.1157 WHQL (dated 07/07/2015)
2. Intel MEI driver v11.0.0.1144 BETA (dated 06/09/2015)
3. Intel MEI drivers v11.0.0.1166 WHQL (dated 08/31/2015) and v10.0.30.1058 WHQL (dated 09/23/2015)

Conclusions:

• At least it was the Intel MEI driver v11.0.0.1157 WHQL (and not the v11.0.0.1144 Beta), which gave my Z97 AHCI system the absolutely best READ and OVERALL performance. Even the WRITE scores were a little bit better than with all other 3 test candidates.
• The Intel MEI driver v11.0.0.1144 Beta was less performant than the v11.0.0.1157 WHQL, but gave me much better READ and OVERALL scores than the other 2 test candidates.
• The benchmark results, which I got with the latest Intel MEI drivers of the v10 and v11 series, were nearly identical. This verifies, that only the Intel MEI drivers of the v11.0.0.114x and v11.0.0.115x development branches have a noticeable positive impact on the performance of my Z97 system.
• It seems, that Intel had added a special Intel MEI performance boost feature, when the company began with the v11.0.0.xxxx development tree, but for unknown reasons they obviously have removed this feature, when they started with the 11.0.0.116x development branch.

[Horstfuchs]

Thanks for performing these benchmarks!
I really find it hard to believe that different MEI drivers could cause such cause substantial performance differences. Is it possible these results are within the margin of error? What would happen if you were to actually never install/update the MEI driver?

[plutomaniac]

Intel ME 10.0 Apple Mac 1.5MB Firmware v10.0.35.1012
Intel ME 9.5 Apple Mac 1.5MB Firmware v9.5.4.1856
Intel ME 7 Apple Mac 1.5MB Firmware v7.0.1.1205
Intel ME 1 Tekoa GbE AMT Firmware v1.4.3

[Fernando]

@Horstfuchs

Is it possible these results are within the margin of error?

No, I have done 2 tests with the same result.
Furthermore the measured differences are outside the usual tolerance level.
EDIT:
Just to avoid any misunderstanding: I have not installed the Intel MEI Software, just the pure Intel MEI driver from within the Device Manager.

[gihan]

@plutomaniac

@ gihan :

There are three ways to combat this: a) use FWUpdate to downgrade the ME before flashing and then update it again, b) replace the ME Region with the one from the A13 update and then flash bios + update me manually again, c) modify the BIOS update files to ignore the newer ME and flash only BIOS

a) You cannot downgrade using FWUpdate because the 9.5.14 ME that is inside the A13 update has a VCN of 5 whereas your 9.5.45 ME has a newer VCN of 9.
b) Run Flash Programming Tool with command fptw -d SPI.bin and check if it reports Error 26. If it does, you cannot replace the ME Region with the one from the A13 update as your Flash Descriptor is locked.
c) I have attached the files that the A13 executable extracts (use WinRar v5.x). You can use SctWinFlash32 to manually flash your BIOS. The file you need in that case is DOH50A13.ROM.

Note: I take no responsibility for option c. It should work but I cannot even check the arguments of WinFlash without a PhoenixBIOS system.

@ those who send pms about the archives abeing "corrupt" :

They are not! WinRar v5.x or later is required. Read the first post carefully!

thanks bro it worked really appreciate !!! you’re a genius!!!

[plutomaniac]

@gihan

thanks bro it worked really appreciate !!! you’re a genius!!!

It would be nice if you could tell me what you did to solve the issue. Method C by running the BIOS Flasher directly with the .ROM image?

Also, run MEInfo (shouls display info and no red errors) and MEManuf (should report MEManuf Test Passed) to check that everything is ok with the ME firmware. Other than that, you are good to go. But do let me know of how you solved the issue for future reference.

[gihan]

@plutomaniac

Zitat von gihan im Beitrag #1627

---

thanks bro it worked really appreciate !!!! you're a genius!!!!!!!!

It would be nice if you could tell me what you did to solve the issue. Method C by running the BIOS Flasher directly with the .ROM image?

Also, run MEInfo (shouls display info and no red errors) and MEManuf (should report MEManuf Test Passed) to check that everything is ok with the ME firmware. Other than that, you are good to go. But do let me know of how you solved the issue for future reference.

---

your C method worked for me

[plutomaniac]

Intel MEI Drivers & Software v11.0.0.1167 for Corporate systems (driver 1166)

@gihan

your C method worked for me

Good to know, thanks for the report. Enjoy your updated system.

[gihan]

i should thank you bro and all tests were passed again thanks!!!

[Fernando]

@Margus

Thanks to one guy on guru3d forum he linked special performance MEI driver for z97 so far i test it its still best then all others what intel released intel removed these performance settings on latest drivers its up on biostar download page its not even in staition-drivers. grab here http://download.biostar.com.tw/upload/Dr…_1d5M_Win10.exe but i can tell its work perfect.

I can confirm this statement.
Today I have done some benchmark tests using the following Intel MEI driver versions:
1. v10.0.30.1058 WHQL (latest Intel MEI driver from the v10 series)
2. v11.0.0.1144 Beta
3. v11.0.0.1166 WHQL (currently latest Intel MEI driver)

You can find the results >here<.

[N6O7]

Thanks!
I couldn’t find the MEI Driver v11.0.0.1144 (1,5Mb) for Operating System others than Windows 10 (aka Biostar motherboard: http://www.biostar.com.tw/app/en/mb/intr…php?S_ID=802#dl).

Is there one for Win 7 x64 around?

[Fernando]

@N6O7

I couldn’t find the MEI Driver v11.0.0.1144 (1,5Mb) for Operating System others than Windows 10 (aka Biostar motherboard: http://www.biostar.com.tw/app/en/mb/intr…php?S_ID=802#dl).
Is there one for Win 7 x64 around?

Yes, I have attached the "pure" Intel MEI driver v11.0.0.1144 Beta.
It is digitally signed by Intel and usable with all 32/64bit Windows Operating Systems from Win7 up.

Intel MEI Drivers v11.0.0.1144 Beta.rar (3.18 MB)

[Fernando]

Today I have tested the impact of the Intel MEI driver v11.0.0.1157 WHQL on the performance of my Z97 AHCI system.
These are the surprising benchmark results I got with this driver:

So not the Intel MEI driver v11.0.0.1144 Beta, but the newer and WHQL certified Intel MEI driver v11.0.0.1157 seems to have the biggest positive impact on the performance of an Intel 9-Series Chipset AHCI system.

Due to these new test results I have updated again >this< post and its conclusions.

[kurtextrem]

Either I’m blind or I can’t find the INF only version of 11.0.0.1157. Can you point me to it, or attach it?

Thanks in advance, very interesting results

[plutomaniac]

@kurtextrem

Either I’m blind or I can’t find the INF only version of 11.0.0.1157. Can you point me to it, or attach it?

The latest INF is 11.0.0.1166 and can be found at Section A1.

[Fernando]

@kurtextrem

Either I’m blind or I can’t find the INF only version of 11.0.0.1157. Can you point me to it, or attach it?

The "pure" 32/64bit Intel MEI drivers v11.0.0.1157 WHQL are attached.

Intel MEI Drivers v11.0.0.1157 WHQL.rar (3.18 MB)

[plutomaniac]

@ bicalov:

I remember that last time I checked you had the correct BIOS variant applied. I had given some advise and tests for you to run but they were never done. This could be an HP BIOS issue, a corrupted ME Region, a corrupted BIOS Region, wrong NVRAM settings, the fact that the motherboard is in a bad state etc. Too many variables. Without an unlocked flash descriptor we cannot exclude the ME Region as the issue, although as I said earlier I believe it’s not ME-related. If your chip is removable I suggest you buy a cheap programmer to flash it freely or a least unlock the flash descriptor with the “pinmod”. Go for the pinmod first and test if it works by trying command fpt -d spi.bin (no error 26 should be shown this time, a full 8MB dump is the resulting file).

Maybe @CodeRush can think of something else. To not waste his time searching, this is the beginning of the conversation.

[gpvecchi]

Guys, I tried updating the ME firmware with 11.0.etc.1173 on my new Asus Z170 Deluxe, but I get a generic Error 1 (unknown error). I have the full MEI drivers package and I’m using the correct (x64) version of fwupd

[plutomaniac]

@ gpvecchi:

Yes, we have been having this issue since Skylake launched. There is a bug either at Intel’s side (FWUpdate tool) or the OEM’s side (BIOS). I don’t understand how BIOS engineers or Intel do not see this issue.

Based on previous tests and experiments not even an UPD image works. It will show another unexplained error. Still, do try it. I have attached the latest Consumer PCH-H 11.0.0.1173 UPD image.

11.0.0.1173_CON_H_PRD_UPD.rar (1.15 MB)