What controls Turbo Core in Xeons?

Page 41 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

rinseaid

Junior Member
Jul 6, 2017
3
0
6
traderjay,

Here are the modded BIOS file to unlock the turbo. I tested it on my X10SRL-f motherboard last night and it worked fine.

mod_X10SRL-f.zip

I included EFI files also. So you should load one of them before getting into OS.

Before and After

Cinebench R15 score: Before: 1500, After: 1725.
CPU-Z (v1.79) multi-score: Before: 3934, After 4552.

Again, I am currently using E5-2658 v3 QS chip on this motherboard so there is no different on single core performance, but multi-threads performance is improved from 2.5Ghz all cores to 2.9 all cores after mod.

And, here is my BIOS settings. You may tweak them if it is needed.

CPU Configuration
---------------------------
Hyper-Threading (ALL) - [Enable]
Cores Enabled - [0]

CPU Configuration -> Advanced
---------------------------
Power Technology - [Custom]
Energy Performance Tuning - [Disable]
Energy Performance BIAS setting - [Performance]
Energy Efficient Turbo - [Disable]

Advanced -> CPU P State Control
--------------------------------
EIST (P-States) - [Enable]
Turbo Mode - [Enable]
P-State Coordination - [HW_ALL]

Advanced -> CPU C State Control
--------------------------------
Package C State Limit - [C0/C1 state]
CPU C3 Report - [Disable]
CPU C6 Report - [Enable]
Enhanced Halt State (C1E) - [Disable]

CPU T State Control
-------------------
ACPI T-States - [Enable]

Last, Windows OS Power Options - High Performance


Let me know if you need anything.

Hi James, is there any possibility you could write a quite guide on how you applied the changes to the BIOS? I'm trying to do the same for a Supermicro X10DRi but got a little confused when opening UEFI tool.

Thanks!
 

James Kim

Junior Member
Jun 3, 2017
17
3
41
Hi James, is there any possibility you could write a quite guide on how you applied the changes to the BIOS? I'm trying to do the same for a Supermicro X10DRi but got a little confused when opening UEFI tool.

Thanks!

Hi rinseaid,

When you said that "apply changes to the BIOS", I am confused if you are asking me how to change the BIOS file or how to apply a EFI file in the boot process.

Did you already mod the X10DRL-i bios and upload it? I believe that the latest version is 2.0b on May 2017. If you don't have modded BIOS file, I can give you mine (2.0b).

Download X10DRL7.505.UEFI.zip (2.0b)

The link has both original and modded BIOS files.

Please let me know which steps are confused. Thanks,
 

rinseaid

Junior Member
Jul 6, 2017
3
0
6
Hi rinseaid,

When you said that "apply changes to the BIOS", I am confused if you are asking me how to change the BIOS file or how to apply a EFI file in the boot process.

Did you already mod the X10DRL-i bios and upload it? I believe that the latest version is 2.0b on May 2017. If you don't have modded BIOS file, I can give you mine (2.0b).

Download X10DRL7.505.UEFI.zip (2.0b)

The link has both original and modded BIOS files.

Please let me know which steps are confused. Thanks,

Thanks James - sorry that I did not explain myself well. It's the changing of the BIOS file that I'm having trouble with. The board I have is actually a different board, the X10DRi-f (http://www.supermicro.com/products/motherboard/Xeon/C600/X10DRi.cfm) - latest BIOS version appears to be 2.1a.

Thanks again, I appreciate your help!
 

James Kim

Junior Member
Jun 3, 2017
17
3
41
Thanks James - sorry that I did not explain myself well. It's the changing of the BIOS file that I'm having trouble with. The board I have is actually a different board, the X10DRi-f (http://www.supermicro.com/products/motherboard/Xeon/C600/X10DRi.cfm) - latest BIOS version appears to be 2.1a.

Thanks again, I appreciate your help!

Well, I don't have the board and I know that the BIOS file can't be modded by using UBU and mmtool. X10SRL-f and X10DRL-i BIOS files can be done by those tools without issue. However, other Supermicro BIOS files will have issue with them.. However, I modded manually using Hex tool. However, I can't guarantee if this will work or not because I haven't tested it myself on actual board. If you want to take that risk, here is the file that you can try.

UEFI BIOS mod: X10DRi17.505 (the original and mod)

Microcode removed: screenshot

Again, I couldn't test it because I don't have the board so please don't blame me if something goes wrong.

Good luck.
 
  • Like
Reactions: rinseaid

rinseaid

Junior Member
Jul 6, 2017
3
0
6
Well, I don't have the board and I know that the BIOS file can't be modded by using UBU and mmtool. X10SRL-f and X10DRL-i BIOS files can be done by those tools without issue. However, other Supermicro BIOS files will have issue with them.. However, I modded manually using Hex tool. However, I can't guarantee if this will work or not because I haven't tested it myself on actual board. If you want to take that risk, here is the file that you can try.

UEFI BIOS mod: X10DRi17.505 (the original and mod)

Microcode removed: screenshot

Again, I couldn't test it because I don't have the board so please don't blame me if something goes wrong.

Good luck.

Thanks so much! I really appreciate it. I'll let you know how it goes. And I won't blame you if it fails :)
 

Sean J Sauve

Junior Member
Jun 23, 2017
9
0
11
OK using this all the time but why is is cup39 better than the -50 one; using a XEON E5 2696 v3. Or better yet is there a list somewhere in the forum as to what each revision does?

Finally why do I get a VCC error everyday I boot up. I am using the FIVR vcc additions
 

CANONKONG

Member
Jul 11, 2017
98
62
46
I want to make a V3.EFI with CoreVOffset -70mv &
CacheVOffset -50mv
HELP ME,Please!@Dufus
Code:
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; V3.EFI                          ;;
;; Voltage in millivolts           ;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

format PE64 dll efiboot at 0 on 'nul'
include 'efi3.inc'
;-------------------------------------------------------------------------------------------------------

 CoreVOffset            = -70
 CoreVStatic            = 0          
 CoreMulti              = 0          
 CoreFixed              = 0          

 if       CoreVOffset   > 999
          CoreVOffset   = 999
 else if  CoreVOffset   < -1000
          CoreVOffset   = -1000
 end if

 CoreAdjVOffset         = ((CoreVOffset shl 15)/1000) and 0ffe0h
 CoreAdjVStatic         = ((CoreVStatic shl 18)/1000) and 0fff00h
 CoreVoltage            = ((CoreAdjVOffset shl 16) or CoreAdjVStatic) or CoreMulti or (CoreFixed shl 20)
;-------------------------------------------------------------------------------------------------------

 CacheVOffset           = -50
 CacheVStatic           = 0
 CacheMulti             = 0
 CacheFixed             = 0

 if      CacheVOffset   > 999
         CacheVOffset   = 999
 else if CacheVOffset   < -1000
         CacheVOffset   = -1000
 end if

 CacheAdjVOffset        = ((CacheVOffset shl 15)/1000) and 0ffe0h
 CacheAdjVStatic        = ((CacheVStatic shl 18)/1000) and 0fff00h
 CacheVoltage           = ((CacheAdjVOffset shl 16) or CacheAdjVStatic) or CacheMulti or (CacheFixed shl 20)
;-------------------------------------------------------------------------------------------------------

section '.text' code readable executable

entry $
        push    rbx
        push    rcx
        push    rdx
        push    rdi
        push    rsi
        push    r15
        push    r14
        push    r13
        push    r12
        push    rbp

        sub     rsp,21h*8
        xor     r12,r12
        mov     r14,rcx                 ; ImageHandle
        mov     r15,rdx                 ; pointer to SystemTable


        mov     ebp,80000000h

        xor     eax,eax
        xor     edx,edx
        mov     ecx,8bh
        wrmsr
        xor     ecx,ecx
        inc     eax
        cpuid
        cmp     eax,306f2h
        jne     WrongCPU

        mov     ecx,8bh
        rdmsr
        cmp     edx,0
        jne     MicroCodePresent

        mov     ecx,194h
        rdmsr
        bt      eax,20
        jc      OC_Locked

        mov     ecx,150h                ; Get Core OC ratio and capabilities
        mov     edx,ebp
        inc     edx
        xor     eax,eax
        wrmsr
        rdmsr
        cmp     dl,0
        jne     MailBoxError
        mov     [TopCore],al

        mov     edx,ebp                 ; Get Cache OC ratio and capabilities
        mov     dx,201h
        xor     eax,eax
        wrmsr
        rdmsr
        cmp     dl,0
        jne     MailBoxError
        mov     [TopCache],al

;        mov     edx,ebp
;        mov     dl,10h
;        xor     eax,eax
;        wrmsr
;        rdmsr
;        cmp     dl,0
;        jne     MailBoxError

        mov     eax,CoreVoltage
        mov     al,[TopCore]
        mov     edx,ebp                 ; Set Core OC ratio
        mov     dl,11h
        wrmsr
        rdmsr
        cmp     dl,0
        jne     MailBoxError


;        mov     edx,ebp
;        mov     dx,210h                 ; Set Cache OC ratio
;        xor     eax,eax
;        wrmsr
;        rdmsr
;        cmp     dl,0
;        jne     MailBoxError

        mov     eax,CacheVoltage
        mov     al,[TopCache]
        mov     edx,ebp
        mov     dx,211h
        wrmsr
        rdmsr
        cmp     dl,0
        jne     MailBoxError

        mov     al,[TopCore]            ; Set turbo ratio's
        mov     ah,al
        mov     dx,ax
        shl     eax,16
        mov     ax,dx
        mov     edx,eax
        mov     ecx,1adh
        wrmsr
        inc     ecx
        wrmsr
        inc     ecx
        or      edx,ebp
        wrmsr


        mov     ecx,620h                ; Set Cache Min/Max Ratios
        rdmsr
        mov     al,[TopCache]
        mov     ah,al
        wrmsr


        mov     ecx,194h                ; Lock OC
        rdmsr
        bts     eax,20
        wrmsr

        lea     rdx,[_Success]
        call    TextOut
        jmp     @f

  WrongCPU:
        lea     rdx,[_WrongCPU]
        call    TextOut
        xor     rax,rax
        jmp     @f

  MicroCodePresent:
        lea     rdx,[_MicroCodePresent]
        call    TextOut
        xor     rax,rax
        jmp     @f

  OC_Locked:
        lea     rdx,[_OC_Locked]
        call    TextOut
        xor     rax,rax
        jmp     @f

  MailBoxError:
        lea     rdx,[_MailBoxError]
        call    TextOut
        xor     rax,rax
        jmp     @f


  @@:

        xor     rax,rax
        add     rsp,21h*8
        pop     rbp
        pop     r12
        pop     r13
        pop     r14
        pop     r15
        pop     rsi
        pop     rdi
        pop     rdx
        pop     rcx
        pop     rbx

        ret

  TextOut:
        sub     rsp,5*8
        mov     rcx, [r15 + EFI_SYSTEM_TABLE.ConOut]
        call    [rcx + SIMPLE_TEXT_OUTPUT_INTERFACE.OutputString]
        add     rsp,5*8
        retn


section '.data' data readable writeable

   TopCore              db ?
   TopCache             db ?
  _WrongCPU             du 13,10,'V3 - Wrong CPU.',13,10,0
  _MicroCodePresent     du 13,10,'V3 - Microcode present.',13,10,0
  _OC_Locked            du 13,10,'V3 - Overclocking Locked.',13,10,0
  _MailBoxError         du 13,10,'V3 - Mailbox Error.',13,10,0
  _Success              du 13,10,'V3 - All Turbo Set',13,10,0
 
  • Like
Reactions: txgy

CANONKONG

Member
Jul 11, 2017
98
62
46
I've made my own V3x2.efi
https://ufile.io/541182
( Unlock boost for E5-26xx-v3 multiprocessor systems, based on original V3.efi ).

PS: It is my first experience with uefi programming... it works, on my PC at least... but why I have got x28 on the first and x30 on the second CPU... I'll check tomorrow.
Can you make a V3.EFI with CoreVOffset -70mv & CacheVOffset -50mv,I don't know to editor EFI file. V3.EFI default CoreVOffset -20mv & CacheVOffset -20mv, and I use it ,my E5-2696V3 can 18cores@ 3.2GHz on X99E-ITX/AC ,if i use X99-M WS with no MC,stock max radio
&make CoreVOffset -70mv in the bios,it can 18cores@3.5GHz.Due to Asrock mb can not make CoreVOffset effective, so I want a V3.EFI with CoreVOffset -70mv & CacheVOffset -50mv
 

DOC_Zenith

Junior Member
Jun 21, 2017
5
1
41
OK, Im here to report my success. I gave it a shot with E5-2696v3 and Asus X99-A/USB3.1 (because I had good experiences with X99-A that I used to own before.

My final result is 33x on all cores full load. Regardless if HT is enabled or not. It holds full 38x if 10 or less cores are loaded, 37x if 12, 34x if 16.
Whats the best, I do not even have to use the EFI boot. It works too, but not necessary with this board, I just set all the power limits to the maximum in bios, the turbo to all core 38x and uncore manual min/max to 30 (otherwise the system is not stable), C states all ON, and load the microcode when windows boot trough the Wmware driver (I use microcode 39). No EFI boot necessary (i use legacy win7). I was lucky even with the FSB, as my CPU is stable at 105mhz, making it 3460mhz on full load and 4Ghz on 10 and less turbo.

Thanks to the fact that power limits are moved, the system does not have the issue I had with broadwell-E 2696v4, ala the turbo there worked fine only on CPU loads, when GPU was loaded (thus NB as well), the turbo would drop to base turbo of 28x even if just 1/2 threads were loaded. Here this is not the case. 8+ full loaded and CPU still on full turbo.

Here are some results.

Corona:
http://i.imgur.com/zjTM0Ik.jpg

Hwbot H265:
http://i.imgur.com/lfq61Qi.jpg
http://i.imgur.com/THYL6AZ.jpg

The AVX underclock is to 29-30x fluctuating.

Cinebench R15:
http://i.imgur.com/GJaM98f.jpg

The 2 CPU's above, are first E5-2679v4, the fastest LGA2011v3 CPU (20cores all turbo at 3,2Ghz stock), and above are my old two Haswell-E enginnering samples. They are two 12 cores 30MB L3, run at 3,1Ghz full load (the turbo is only to 32x and has to be disabled with Cstates otherwise they are unstable those are old samples), but at the same time these do not have any AVX delta, aka run 31x even in full AVX load. And this works fine and uncore does not throttle. These use some ancient microcode with TSX enabled but there definitely is a microcode cause there are no uncore issues.

Now, this brings to me an idea. The 2696v3 can either be ran without microcode and C-states so there is no AVX underclock, but the uncore throttles down, or with C-states enabled so uncore is fine and the CPU generally performs better out of AVX. But the thing is, the 12 core ES could do both. But I was not able to find anywhere on the internet a valid uncore of that kind. Aka pre-launch uncore with TSX and no uncore throttle, no AVX throttle.

If you go to Asrock X99 extreme 4:

http://www.asrock.com/mb/Intel/X99 Extreme4/?cat=Download&os=BIOS

The very first bios 1.2 uses the old TSX microcode even for retail Haswells. The bios also has the 2 microcodes for the engineering samples. These would be amazing if someone was able to extract them or find them somewhere so we could use them. I'm not skilled enough to extract them from the bios. And I was unable to find them. The first microcode Intel released for Haswell-E is already post launch with all the errata fixes.
 
  • Like
Reactions: GTZ

sciff

Member
Mar 6, 2017
136
52
71
DOC_Zenith, I have a E5-2696v3 too, on ASUS X99-E WS/USB 3.1.

Mine works at 3.4GHz all-core turbo, including AVX applications, with C-states enabled. Normally I don't observe the uncore multiplier drop under full load. I currently use v3x2_50_vcc.efi, but before it was made, I used V3x2_cup.efi with 3.3GHz all-core turbo, also no uncore drop. I don't load microcode patch anywhere, including VMWare driver.



By the way, it has already been observed that UEFI boot is completely unnecessary for this exploit to work. My Windows 10 is installed on an MBR disk, loaded in legacy mode and everything works fine!
 

mike999

Junior Member
Jul 13, 2017
2
0
6
hello
I have a E5 2686 v3 too, on ASUS X99-E WS
Ubu does not work: error in replacing file.

I tried updating the microcode manually:
With uefitool I replaced cpu000306F2_plat6F_ver00000038_date#28-03-2016
With mmtool I checked the modified bios cab and it is ok

I ask you for help
I can not load the bios into the asus:
It does not work through usb
It does not work through bios
It does not work through afu
It does not work through afudos

how can I do?
thank you
 

sciff

Member
Mar 6, 2017
136
52
71
mike999, you have to edit the original BIOS file in a hex editor first and only then pass it through UBU.

Use search function inside it and search for "2E06A0" hex value and then edit the file by adding 4 right below "2E" and subracting the same value three to the right. This way securecode feature of the BIOS file is disabled. But be careful, the searched part should be towards the end of the file and there can be two instances (in case of ASUS BIOS there usually are two), edit both. If you also find "2E06A0" somewhere in the beginning or in the middle of the BIOS file, ignore it, it's just a coincidence.


In this case, as a result I got 18 instead of 14 and 08 instead of 0C
P.S. don't forget to agree to rename the .CAP file for ASUS Flashback, when you get asked by UBU at the very end.
 

GTZ

Member
Jan 25, 2017
36
12
41
OK, Im here to report my success. I gave it a shot with E5-2696v3 and Asus X99-A/USB3.1 (because I had good experiences with X99-A that I used to own before.

My final result is 33x on all cores full load. Regardless if HT is enabled or not. It holds full 38x if 10 or less cores are loaded, 37x if 12, 34x if 16.
Whats the best, I do not even have to use the EFI boot. It works too, but not necessary with this board, I just set all the power limits to the maximum in bios, the turbo to all core 38x and uncore manual min/max to 30 (otherwise the system is not stable), C states all ON, and load the microcode when windows boot trough the Wmware driver (I use microcode 39). No EFI boot necessary (i use legacy win7). I was lucky even with the FSB, as my CPU is stable at 105mhz, making it 3460mhz on full load and 4Ghz on 10 and less turbo.

Thanks to the fact that power limits are moved, the system does not have the issue I had with broadwell-E 2696v4, ala the turbo there worked fine only on CPU loads, when GPU was loaded (thus NB as well), the turbo would drop to base turbo of 28x even if just 1/2 threads were loaded. Here this is not the case. 8+ full loaded and CPU still on full turbo.

Here are some results.

Corona:
http://i.imgur.com/zjTM0Ik.jpg

Hwbot H265:
http://i.imgur.com/lfq61Qi.jpg
http://i.imgur.com/THYL6AZ.jpg

The AVX underclock is to 29-30x fluctuating.

Cinebench R15:
http://i.imgur.com/GJaM98f.jpg

The 2 CPU's above, are first E5-2679v4, the fastest LGA2011v3 CPU (20cores all turbo at 3,2Ghz stock), and above are my old two Haswell-E enginnering samples. They are two 12 cores 30MB L3, run at 3,1Ghz full load (the turbo is only to 32x and has to be disabled with Cstates otherwise they are unstable those are old samples), but at the same time these do not have any AVX delta, aka run 31x even in full AVX load. And this works fine and uncore does not throttle. These use some ancient microcode with TSX enabled but there definitely is a microcode cause there are no uncore issues.

Now, this brings to me an idea. The 2696v3 can either be ran without microcode and C-states so there is no AVX underclock, but the uncore throttles down, or with C-states enabled so uncore is fine and the CPU generally performs better out of AVX. But the thing is, the 12 core ES could do both. But I was not able to find anywhere on the internet a valid uncore of that kind. Aka pre-launch uncore with TSX and no uncore throttle, no AVX throttle.

If you go to Asrock X99 extreme 4:

http://www.asrock.com/mb/Intel/X99 Extreme4/?cat=Download&os=BIOS

The very first bios 1.2 uses the old TSX microcode even for retail Haswells. The bios also has the 2 microcodes for the engineering samples. These would be amazing if someone was able to extract them or find them somewhere so we could use them. I'm not skilled enough to extract them from the bios. And I was unable to find them. The first microcode Intel released for Haswell-E is already post launch with all the errata fixes.

Thank you for your findings!!!! It is also awesome the Asus board lets you do all the changes without the use of an EFI driver.

I too noticed the uncore throttling on my 2696V3 with no ucode and posted my findings a few pages back. I tried searching for the first batch of ucode but unable to find them.

Then you mentioned the original AsRock X99 Extreme 4 BIOS might have it. I have this board and will try it. The problem with the older BIOS (I have had this board since launch with my 5820K). They are quirky, outdated and missing a lot of memory support (think the initial Ryzen issues). The original BIOS that came with my board could not even run quad channel. But for the sake of science, I shall try. With no you ucode I can run an all core turbo of 35 but the uncore throttle to 2200 (or 1200 don't remember) and whatever advantage I gained got negated with the uncore underclock.
 

sciff

Member
Mar 6, 2017
136
52
71
Thank you for your findings!!!! It is also awesome the Asus board lets you do all the changes without the use of an EFI driver.

I too noticed the uncore throttling on my 2696V3 with no ucode and posted my findings a few pages back. I tried searching for the first batch of ucode but unable to find them
I'm sure he meant legacy windows boot instead of UEFI boot. You can't unlock v3 xeon without EFI hack.

As for uncore throttling down to x23 or whatever, with no microcode patch loaded one way or another. It has been mentioned that you can prevent it by using 'v3x2_cup.efi' or 'v3x2_50_vcc.efi', like I already said a couple posts back.
 

GTZ

Member
Jan 25, 2017
36
12
41
I'm sure he meant legacy windows boot instead of UEFI boot. You can't unlock v3 xeon without EFI hack.

As for uncore throttling down to x23 or whatever, with no microcode patch loaded one way or another. It has been mentioned that you can prevent it by using 'v3x2_cup.efi' or 'v3x2_50_vcc.efi', like I already said a couple posts back.

From what I got from re reading his message is, all he did was delete the microcode and set all the settings in the BIOS and it worked without a EFI driver.

As far as the no uncore throttling each X99 handles it slightly different. I currently use v3x2_50_vcc.efi and the throttling still occurs because I have to disable some c-states or else my system becomes unstable. I cannot enable them, the system will eventually hardlock. It can take a few seconds or hours inside the windows environment. So it is my fault for not explaining. Finding another way is helpful for my case.
 

CANONKONG

Member
Jul 11, 2017
98
62
46
I'm sure he meant legacy windows boot instead of UEFI boot. You can't unlock v3 xeon without EFI hack.

As for uncore throttling down to x23 or whatever, with no microcode patch loaded one way or another. It has been mentioned that you can prevent it by using 'v3x2_cup.efi' or 'v3x2_50_vcc.efi', like I already said a couple posts back.
ASUS&GIGABYTE can use a bios delete the microcode and set all the settings in the BIOS(set max turbo radio,CoreVOffset -70mv & CacheVOffset -50mv) and it worked very well without a EFI driver.
 

CANONKONG

Member
Jul 11, 2017
98
62
46
From what I got from re reading his message is, all he did was delete the microcode and set all the settings in the BIOS and it worked without a EFI driver.

As far as the no uncore throttling each X99 handles it slightly different. I currently use v3x2_50_vcc.efi and the throttling still occurs because I have to disable some c-states or else my system becomes unstable. I cannot enable them, the system will eventually hardlock. It can take a few seconds or hours inside the windows environment. So it is my fault for not explaining. Finding another way is helpful for my case.
Yes,ASUS&GIGABYTE can use a bios delete the microcode and set all the settings in the BIOS(set max turbo radio,CoreVOffset -70mv & CacheVOffset -50mv) and it worked very well without a EFI driver.
 

CANONKONG

Member
Jul 11, 2017
98
62
46
ASUS&GIGABYTE can use a bios delete the microcode and set all the settings in the BIOS(set max turbo radio,CoreVOffset -70mv & CacheVOffset -50mv) and it worked higher frequency
very well without a EFI driver. But Asrock&MSI can not! So i make a bios delete the microcode and insert V3.EFI (V3.EFI transform V3.ffs)into the bios and it also worked very well without a EFI driver.Due to Asrock&MSI can not have an effective CoreVOffset,the frequency will lower than ASUS&GIGABYTE mainboard.My 2696v3 can worked 3.5GHz on ASUS X99-M WS,but 3.2GHz on Asrock X99M-Killer 3.1,so i want to get two V3.efi with CoreVOffset -50mv & CacheVOffset -50mv and V3.efi with CoreVOffset -70mv & CacheVOffset -50mv,and transform them insert into the bios.I think Asrock&MSI can also work higher frequency.
 

mike999

Junior Member
Jul 13, 2017
2
0
6
sciff, Thanks, it works. You saved me.

I tried various microcode and it seems to be the best on 14 (opz 7,1, 14 - mmtool)
Overclock a 3120 mhz (30 x 104)
with cinebench 2120 points
with prime 95 stress test
The clock drops to 2800 ghz

can it be solved?
thank you
 

DOC_Zenith

Junior Member
Jun 21, 2017
5
1
41
Sciff - Nope. I just tested it again to be really sure, and it does not work for me. "v3x2_50_vcc.efi" gives me only 31x on all cores + uncore does not hold 30x, and throttles to 22-24x under some loads. "v3x2_cup.efi" basically the same. + Both are unstable when all C states are enabled. They both achieve worse results than if I just go with bios.

GTZ - Yes, I have achieved the best result with just setting the multiplier to 38x all in bios, uncore 30-30, Cstates all ON and loading the EFI with the wmware driver during OS load. I do not load any EFI boot partition/efi file. It works as well, for me the best results were with V3x2_cup39.efi, but since I can achieve that with just bios settings it became redundant.

Well, all in all, it looks like things are very board specific. I was expecting same results as Sciff since I also have Asus X99 board, but for me it behaves completely different than his.
 

DRM8626

Junior Member
Jul 14, 2017
3
0
6
Hey guys,

Ive got a 2683 v3 14 core, 2.0 ghz, 3.0 max turbo cpu. I also have an asrock taichi x99 motherboard for it. I also am using corsair vengeance lpx 2133 ram, and i am on windows ten. I am also using an h105 on the cpu, and my temps usually stay around 30c.

I was wondering if one of you guys could hook me up with an efi and bios to squeeze the absolute most performance out of this processor. Id like to hit 3.0 ghz on all cores. I really dont care about heat,as long as i wont thermal throttle or hurt the chip. And im having trouble trying to figure out which efi and bios to use for this cpu and bios. If anyone could point me in the right direction id greatly appreciate it.
 

txgy

Member
Apr 24, 2017
29
1
41
I'm sure he meant legacy windows boot instead of UEFI boot. You can't unlock v3 xeon without EFI hack.

As for uncore throttling down to x23 or whatever, with no microcode patch loaded one way or another. It has been mentioned that you can prevent it by using 'v3x2_cup.efi' or 'v3x2_50_vcc.efi', like I already said a couple posts back.
hi sciff
When I run R15,the uncore drop to 23X with no microcode. How do you make it ?Have you changed some options in BIOS?

BTW, I find something strange:
v3x2_50_vcc.efi linpack score 760GFLOPS
v3x2_50_39_vcc.efi linpack score 725GFLOPS
 
Last edited:

txgy

Member
Apr 24, 2017
29
1
41
ASUS&GIGABYTE can use a bios delete the microcode and set all the settings in the BIOS(set max turbo radio,CoreVOffset -70mv & CacheVOffset -50mv) and it worked higher frequency
very well without a EFI driver. But Asrock&MSI can not! So i make a bios delete the microcode and insert V3.EFI (V3.EFI transform V3.ffs)into the bios and it also worked very well without a EFI driver.Due to Asrock&MSI can not have an effective CoreVOffset,the frequency will lower than ASUS&GIGABYTE mainboard.My 2696v3 can worked 3.5GHz on ASUS X99-M WS,but 3.2GHz on Asrock X99M-Killer 3.1,so i want to get two V3.efi with CoreVOffset -50mv & CacheVOffset -50mv and V3.efi with CoreVOffset -70mv & CacheVOffset -50mv,and transform them insert into the bios.I think Asrock&MSI can also work higher frequency.
Great idea!
waiting for you work!