• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

What controls Turbo Core in Xeons?

Page 22 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

sciff

Member
Mar 6, 2017
136
52
71
@sciff ,
>having no microcode loaded, causes certain applications to actually run faster
This has already been discussed - Intel changed 256-bit registers throttling during development. You can try to lower uncore clock speed and that you'll get higher base clock, as it's hit by TDP limits pretty hard.
I don't know how to do that, there's no such setting in the BIOS. Could you suggest anything?


I personally won't run any production enviroment without any microcode, and probably with anything earlier then the microcode from the times of the first QS released - Intel doesn't share what they've actually patched, only some really outstanding bits, like TSX bug.
Why not? In case of Corona renderer for 3ds Max for example, if we extrapolate from how Corona Benchmark works, no microcode is way better. Again, I'm not talking about 24/7 in turbo mode, only occasional use of the CPU's full power.

I haven't had any stability issues yet when there's no microcode installed. Only when I used 0x27 microcode, then there was one issue, with the video driver dropping out.
 

RamRock

Member
Apr 3, 2017
26
2
16
RamRock, I was told by TheSellHard, that dual CPU mainboards have weaker CPU power supply, therefore you can't boost a CPU or CPUs as much on C612 chipset.
I have 2x 7 power phases :)
I don't need cinebench break record resoluts... I don't want to make more than VRM can deliver.
My CPU at cinebench at 3,1GHz taking 500W of energy
When I working in my apps at stress it take "just" 400W... so I have 100W gap to make Intel limit.
So it's not power limit if I have 100W to the cinebench power consumption.

I can use only 10 cores of 18... it's still 3,1GHz.
Temperature is fine too 60 C degree

On my ASUS X99-E WS/USB 3.1 I've got just one CPU but two CPU power connectors. You see what I mean? In all core turbo my E5-2696 V3 starts to consume power like crazy, hence increased temperatures and cooler FAN RPM. I can even start hearing a slightly noticeable weird sound, as if my PC had a mini-turbine inside :D
You have 1x 8 phases. It's mainly marketing 2x EPS connector in your case-> it can do a lot if you will LN2 overclock.
But that's not a point!

It's unrealistic to expect the same results.

And why do you think x31 multiplier is bad? It's good enough, in my opinion, especially if you have two CPUs running at this speed.
My quastion is why it's x31 :) Can it be change?

P.S. Which revision is your CPU? Mine is R2.
Same
 
Last edited:

zzori

Junior Member
Apr 9, 2017
2
0
1
I'm back! :p

1. Used UBU/MMtool to remove the uCode from bios (Using MM to remove code and save unsigned bios, then use UBU to save it so it can be flashed)
Please Help me.
MMTool load bios after...
uCode is what mean? i can't find uCode T.T

or I want RAMPAGE V EDITION 10 BIOS 1503 modded bios ^^
Thanks a lot
 
Last edited:

Vuk Djordjevic

Junior Member
Mar 13, 2017
4
1
16
GTZ, interesting, I didn't even notice this difference!

By the way, it turns out having no microcode loaded, causes certain applications to actually run faster. Take Corona benchmark, which utilizes AVX and SSE4.2 instructions. Normally in my case it runs at a maximum of 2.8 GHz, whereas if you uninstall the microcode, it runs at 3.3 GHz


Hi Sciff can you explain how did you manage to bypass the AVX? Is there any downside in stability by having no microcode? I have seen your result on corona benchmark page my time with the same CPU is 1:20 the only thing I did was rising the BCLK to 105 but seems that Corona is not even responding to that since it says the render was done with 2.8ghz even though all cores for me work on 2.925 ghz.
2 weeks ago when I got the cpu delivered, here on the forum people were posting that the moded bios clock was not working with rendering applications that use AVX but now I see you managed to bypass that. I was also wondering since I have my system already up and running on the SSD will I have to reinstall windows and all if I decide to go with your route?
 

chris_osw

Junior Member
Apr 5, 2017
11
2
51
http://valid.x86.fr/8b4ykc





Manualy load every time diver V3.EFI and works ,but i cant add driver 4 automatic load before OS load win7 x64.
Problem is (load and create) driver on disk ,maybe because folder c:Windows\Boot\EFI\ is protect by OS ???


http://valid.x86.fr/cache/screenshot/8b4ykc.png

before this mod i have 14000pts in cpu-z ,now i have 16590pts
Power profile: BITSUM Process Lasso - BITSUM highest Perfmance.


 
Last edited:

ajschot

Member
Feb 16, 2017
65
2
16
Hi everyone, I want to mod my bios of my MSI X99S SLI Krait edition, what is the best methode or did somebody already mod that bios that I can get?

Also I can't find a download link for Ubu both are down
 
Last edited:

ajschot

Member
Feb 16, 2017
65
2
16
not possible with the MSI bios :-(
i am getting errors when saving... in UBU but also when i try with MMTool
 

ajschot

Member
Feb 16, 2017
65
2
16
@someEEguy

Great little compilation of steps to go through for the ones not wanting to read/go through everything posted up till now ^^

Edit:
Beware that using UBU only can cause your bios to not boot if your on one of those APTIO 5 biosses (UBU warns for that)

I had this a couple of times and had to use MMtool to remove the mirocodes, then just open the bios with UBU and just immediately press 0/Exit and save that way.
Worked every single time for me on both Extreme4 and Extreme6 Biosses :)
Hey how did you edit your bios with MMTool? I try to edit an MSI bios but i can't save the file (Aptio V)
 

ajschot

Member
Feb 16, 2017
65
2
16
@Ionstream ,
>this is a SuperMicro board

MMTool won't work with it. I've successfully modded X10DRi-T by an algorithm described at the beginning of this thread, when you manually modify 2 bytes of the microcode signature so it is no longer recognized as 306F2. There're four f2 06 03 00 sequences in this bios, and you need to change 2nd and 3rd ones (but always re-check bios validity with all tools available after this, yours may have different structure).
Can you please tell me where i can not find it... i want to edit a MSI bios
 

ajschot

Member
Feb 16, 2017
65
2
16
@fromax Sometimes it's enough to just modify the microcode header so it isn't usable. For instance we can change the CPUID of 0306F2 to 0506F2 AND 0106F2. The reason for changing by both +2 and -2 is to keep the simple checksum the same. Seems there are two microcode locations in BIOS E7885IMS.1D0. We could try changing the following offsets

F0C69E 03 -> 05
F0C6E6 03 -> 01

F1C29E 03 -> 05
F1C2E6 03 -> 01

Of course if the MSI BIOS has been written to halt on no microcode or uses more than just the usual simple checksum or uses a function that needs a microcode fix then it will not work.
Hey how to find the right offset? I have a MSI board too but it seems like those are not the right offsets for me
 

sciff

Member
Mar 6, 2017
136
52
71
WOW!

When I see this, I assume that my 75Watt TDP 2618L v3 will reach 34x on 8 Cores instead o only 25x stock! This is soo cool...


I am just not brave enough to flash my BIOS - everything else is prepared.... I really shouldl try it!

Cheers Andy
Yes, Andy, you should ;)
Your motherboard has a BIOS Flashback button anyway. So if you've modified the bios file incorrectly, it just won't flash. It will blink for like 5 seconds and then just light up steady, which means there's a problem with a BIOS file and nothing was flashed.

I had this probably 5 times until I figured out that it's not the USB stick's file system which is problematic (I tried to reformat it in different ways to make the flashing work), but that I used the method intended for MSI and Gigabyte motherboard BIOSes, rather than for ASUS¹, without understanding it and therefore missing the crucial detail. But I didn't fugure out that detail back then, because a friend ( Anrentren ) provided me the same BIOS modded by himself using standard method.

¹ Instead of opening .CAP file in UEFITool, extracting 1BA0062E-C779-4582-8566-336AE8F78F09 part of it, changing a value inside it with a hex-editor (from 0C to 08) end then replacing it back into the .CAP file (two instances of 1BA0062E-C779-4582-8566-336AE8F78F09 inside it, by the way), I opened the .CAP file itself in hex editor and followed this instruction.

It wasn't a critique per se, but rather suggestions to improve it or avoid mistakes. For your particular case, item nr. 3 was the problem. To have UBU support, you just have to remove FFS_ATTRIB_FIXED (useless on Volume Top Files), in other words, change file attribute from 0C to 08. You could do this with UEFITool, by extracting 1BA0062E-C779-4582-8566-336AE8F78F09 and in that file change offset 0x13 from value 0C to 08, save the file, then replace it. The problem is that UEFITool will also remove the trampoline for recovery (CodeRush can offer more details). To avoid this, open your E7883IMS.110 file in hex editor and change offset FFE0C8 from 95 to 99, then change offset FFE0CB from 0C to 08. You can double check the result with UEFITool. After this change, your file will work in UBU for microcode update. See the picture bellow, you need to change only the two values, but keep the top red line unchanged.
But I didn't have a second value of 95 (which had to be changed to 99), it was 14 instead in my file, so I didn't touch it. But I had to change it, because as I understand now, for checksum to be correct, after subtracting a 4 in one place (when you change 0C from 08), you have to also add 4 in another place. Therefore I just had to go ahead and also change my 14 to 18. And then, if we assume that I'm correct, the flashback would have worked.

P.S. By the way, does anyone know what recovery lordkag was talking about?
 
  • Like
Reactions: itou31

sciff

Member
Mar 6, 2017
136
52
71
Hi Sciff can you explain how did you manage to bypass the AVX? Is there any downside in stability by having no microcode? I have seen your result on corona benchmark page my time with the same CPU is 1:20 the only thing I did was rising the BCLK to 105 but seems that Corona is not even responding to that since it says the render was done with 2.8ghz even though all cores for me work on 2.925 ghz.
2 weeks ago when I got the cpu delivered, here on the forum people were posting that the moded bios clock was not working with rendering applications that use AVX but now I see you managed to bypass that. I was also wondering since I have my system already up and running on the SSD will I have to reinstall windows and all if I decide to go with your route?
Vuk, I don't have any stability issues. I even managed to successfully run prime95's "torture test", although that was pretty scary, because the CPU's TDP rose to crazy 340 W with temperatures also skyrocketing.





But that made me certain that it will run much less demanding Corona render just fine.



Linpack turned out to be even less demanding:



However, I can't guarantee that there will be no downsides, because I don't know.

And I don't know exactly how I bypassed it because I was given ready BIOS settings by a friend who has an identical mobo but a slightly different CPU (QS E5-2686 V3). After I applied them, it all worked! And of course no µcode, because otherwise Corona Benchmark caps at 2.8 GHz.

But I assume that the most important settings are the following:

  • BCLK Frequency slightly higher than 100, in my case 100.2. For some reason, it makes a difference.
  • CPU Current Capability setting set at its maximum of 140%
  • Turbo Mode enabled
  • Long and Short Duration Package Power Limits at 300 (Watts)
  • CPU Integrated VR Fault Management set as Disabled
  • CPU Integrated VR Efficiency Management set as High Performance
  • Limit CPUID Maximum set as Disabled
 

Vuk Djordjevic

Junior Member
Mar 13, 2017
4
1
16
Thnx for the info man I don't have the Asus mobo I took the Asrock Taichi as the people on the forum but I will try to dig out something.
 
  • Like
Reactions: sciff

ajschot

Member
Feb 16, 2017
65
2
16

Hey i am trying to remove microcode with UEFItool i am stuck....
please i need some help.




i need help with this i found the key i now how to extract the body but i don't know how to take out the microcode.... how do i find the values to change.
 

sciff

Member
Mar 6, 2017
136
52
71
RamRock, well I don't know why it works that way, if that's what you're asking. Sorry

http://valid.x86.fr/8b4ykc





Manualy load every time diver V3.EFI and works ,but i cant add driver 4 automatic load before OS load win7 x64.
Problem is (load and create) driver on disk ,maybe because folder c:Windows\Boot\EFI\ is protect by OS ???


http://valid.x86.fr/cache/screenshot/8b4ykc.png

before this mod i have 14000pts in cpu-z ,now i have 16590pts
Power profile: BITSUM Process Lasso - BITSUM highest Perfmance.


Chris, I'm no expert of course, especially when it comes to ASRock motherboards, but still, do you have some kind of "Fast boot" option in your BIOS? I'm asking because I do have it in my BIOS (ASUS), and before I disabled it, the V3.EFI driver also wouldn't load at every system startup.
 
Last edited:

AMDScooter

Senior member
Jan 30, 2001
303
3
81
http://valid.x86.fr/8b4ykc





Manualy load every time diver V3.EFI and works ,but i cant add driver 4 automatic load before OS load win7 x64.
Problem is (load and create) driver on disk ,maybe because folder c:Windows\Boot\EFI\ is protect by OS ???


http://valid.x86.fr/cache/screenshot/8b4ykc.png

before this mod i have 14000pts in cpu-z ,now i have 16590pts
Power profile: BITSUM Process Lasso - BITSUM highest Perfmance.


Honestly have not read the entire thread so excuse me for jumping in the middle. I just put together (like completed 2 hours ago) a system for my wife with that same MOBO (X99m Extreme4) and a E5-2630v4 ES that runs all cores at 23x. Think that modded bios you have would work for me or is it processor specific?
 

rottwag

Member
Apr 2, 2017
77
11
41
...a E5-2630v4 ES that runs all cores at 23x. Think that modded bios you have would work for me or is it processor specific?
somebody correct me if I am wrong, but this "all cores on 1 core turbo" bug only exist in the E5 V3 Xeons (Haswell). Sorry...
 
  • Like
Reactions: sciff

ajschot

Member
Feb 16, 2017
65
2
16
http://valid.x86.fr/8b4ykc





Manualy load every time diver V3.EFI and works ,but i cant add driver 4 automatic load before OS load win7 x64.
Problem is (load and create) driver on disk ,maybe because folder c:Windows\Boot\EFI\ is protect by OS ???


http://valid.x86.fr/cache/screenshot/8b4ykc.png

before this mod i have 14000pts in cpu-z ,now i have 16590pts
Power profile: BITSUM Process Lasso - BITSUM highest Perfmance.


How did you got your memory on 3GHz? I have G SKill Trident ddr4 3200MHZ but mine always keep running at 2133 with E5-2683v3... on the same ASRock X99 Extreme4 i loaded the speed profile in the BIOS but i can't get it to work for a reason always on 2133 and with E5-2658v4 on 2400. How did you do this?
 

ASK THE COMMUNITY