EVGA recently released a BIOS update (Revision SZ2Z) that enables HPET support in the system's BIOS.
What does this mean? It's quite simple, it means that people with EVGA's X58 motherboards can now have vanilla speedstep.
WARNING: This guide involves flashing your system's BIOS with a newer revision. ALWAYS Follow the onscreen instructions during the process. DO NOT remove the floppy or CD-ROM until it says it is safe to do so.
First, visit this thread to grab the update, and for flashing instructions.
After you've completed the process, follow these steps to enable speedstep:
First, your BIOS settings will be wiped out during the update. The first things you will want to do is:
1. Go to the Integrated Peripherals menu
1a. Go to Onboard PATA/SATA Devices
1b. Set the SATA Mode to AHCI
1c. (optional) Set the JMB363 mode to AHCI + IDE (this will allow OS X to natively detect one of the two JMicron controllers)
2. Go back to the main menu, and go into Power Management Setup
2a. Enable HPET support
2b. Set HPET Mode to 64-bit
3. Go back to the main menu once again, and this time go into Frequency/Voltage Control
3a. Go into CPU Feature
3b. Enable Intel Speedstep
3c. Set the CxE Function to C1E
After doing this, save your changes and reboot to OS X.
Go into your Extras/Extensions folder and remove Disabler.kext and NullCPUPowerManagement.kext (and any equivalents), and reboot using verbose mode (optional, but always useful for watching for any issues).
After the system has rebooted, you should now be using Apple's native SpeedStep drivers with no issues!
18 replies to this topic
#1
Posted 04 December 2009 - 11:50 PM
#2
Posted 07 December 2009 - 06:16 PM
Thanks for the great info!!
Just one question from a noobo!
If you update or refresh a new bios, do we have to reset or reconfigure DSDT patch?
I'm on the snow, everything works beautifully except the wakeup from the deep sleep.
Does your info get it right?
Thanks
Just one question from a noobo!
If you update or refresh a new bios, do we have to reset or reconfigure DSDT patch?
I'm on the snow, everything works beautifully except the wakeup from the deep sleep.
Does your info get it right?
Thanks
#3
Posted 11 December 2009 - 04:36 AM
PS: This appears to be working, but still getting some P-State errors regarding turbo mode. I still have no sleep, but I'm not getting C-state errors anymore...
#4
Posted 11 December 2009 - 10:39 AM
Bigfat, on Dec 11 2009, 05:36 AM, said:
In theory this should get us one step closer to Sleep/Wake support, are you still getting the C-state errors?
PS: This appears to be working, but still getting some P-State errors regarding turbo mode. I still have no sleep, but I'm not getting C-state errors anymore...
PS: This appears to be working, but still getting some P-State errors regarding turbo mode. I still have no sleep, but I'm not getting C-state errors anymore...
#5
Posted 14 December 2009 - 07:35 PM
Thanks Geenz! I feel like we're so close to getting Sleep/Wake to work on this board, but without someone with a lot of DSDT know how we're stuck.
#6
Posted 13 September 2010 - 08:25 PM
Hello Everyone
I have an EVGA x58 SLI board and I would like to enable native speedstep. however anytime I remove nullcpupower.....kext I get a KP.
I am running 10.6.4
chameleon rc5
BIOS is set up as listed in post # 1 with the diff. that its verion 73 of the bios
Any thoughts on how I can remove nullcpupower without causing KP?
thanks
I have an EVGA x58 SLI board and I would like to enable native speedstep. however anytime I remove nullcpupower.....kext I get a KP.
I am running 10.6.4
chameleon rc5
BIOS is set up as listed in post # 1 with the diff. that its verion 73 of the bios
Any thoughts on how I can remove nullcpupower without causing KP?
thanks
#7
Posted 16 September 2010 - 10:19 AM
geenz, you missed one important thing - enabling HPET in Mac OS. The HPET section in DSDT should be modified for HPET to work. So if your section looks like this
change it with this
this fix is from gigabyte boards, it works for my DFI (both HPET sections are the same) and may work for eVga, because AFAIK, our DSDTs are almost the same.
Device (HPET)
{
Name (_HID, EisaId ("PNP0103"))
Name (ATT3, ResourceTemplate ()
{
IRQNoFlags ()
{0}
IRQNoFlags ()
{8}
Memory32Fixed (ReadWrite,
0xFED00000, // Address Base
0x00000400, // Address Length
)
})
Name (ATT4, ResourceTemplate ()
{
})
Method (_STA, 0, NotSerialized)
{
If (LGreaterEqual (OSFX, 0x03))
{
If (HPTF)
{
Return (0x0F)
}
Else
{
Return (0x00)
}
}
Else
{
Return (0x00)
}
}
Method (_CRS, 0, NotSerialized)
{
If (LGreaterEqual (OSFX, 0x03))
{
If (HPTF)
{
Return (ATT3)
}
Else
{
Return (ATT4)
}
}
Else
{
Return (ATT4)
}
}
}
change it with this
Device (HPET)
{
Name (_HID, EisaId ("PNP0103"))
Name (ATT3, ResourceTemplate ()
{
IRQNoFlags ()
{0}
IRQNoFlags ()
{8}
Memory32Fixed (ReadWrite,
0xFED00000, // Address Base
0x00000400, // Address Length
)
})
Name (ATT4, ResourceTemplate ()
{
})
Method (_STA, 0, NotSerialized)
{
Return (0x0F)
}
Method (_CRS, 0, NotSerialized)
{
Return (ATT3)
}
}
this fix is from gigabyte boards, it works for my DFI (both HPET sections are the same) and may work for eVga, because AFAIK, our DSDTs are almost the same.
#8
Posted 17 September 2010 - 06:44 PM
Hey guys,
I just wanted to chime in and mention that your instructions to enable vanilla Speedstepping did not work for me either. My settings were already exactly like those posted in the first post to begin with, I have the e75864 bios (which is newer than the one mentioned) and my DSDT already has the mentioned DSDT edit for the HPET section.
Removing NullCPUPowerManagement.kext from /Extra still gives me kernel panics.
I just wanted to chime in and mention that your instructions to enable vanilla Speedstepping did not work for me either. My settings were already exactly like those posted in the first post to begin with, I have the e75864 bios (which is newer than the one mentioned) and my DSDT already has the mentioned DSDT edit for the HPET section.
Removing NullCPUPowerManagement.kext from /Extra still gives me kernel panics.
#9
Posted 24 September 2010 - 09:57 PM
I'm in the same boat; I get KPs too. Just got the new EVGA X58 SLI 3 board and an i7 950.
#10
Posted 25 September 2010 - 03:32 AM
Whenever you add/delete kexts its a good idea to boot with -f to re-generate your kext caches.
#11
Posted 25 September 2010 - 05:26 AM
Thanks STLVNUB, but I already know that's not going to help ME. I did some research, and it looks like the issue has to do with your bios version. On InfiniteMac Airwalk posted that he uses bios version E758_SZ2Z because any of the newer ones add support for 6 and 8 core chips, and that apparently causes the AICPUPM KP. I think I'm probably screwed because I just bought the SLI 3, which just came out, and there are no new or older BIOSes for it. I guess I'll just have to wait for a fix of some kind.
EDIT: I've sent EVGA a request to see if I can get an experimental BIOS that doesn't support 6 cores for my SLI 3.
EDIT: I've sent EVGA a request to see if I can get an experimental BIOS that doesn't support 6 cores for my SLI 3.
#12
Posted 25 September 2010 - 09:16 AM
#13
Posted 25 September 2010 - 05:33 PM
EVGA said they won't give me an experimental bios for my SLI 3. I wonder if I just used E758_SZ2Z. On the other hand, last night I realized that there's a possibility that since the new 6 and 12 core mac pros came out after 10.6.4, that they have a newer build which includes a newer power management kext that can handle 6 core support, and so possibly 10.6.5 will also include this. If that's the case then perhaps I'm just better off running the disabler for now, and then trying power management again when I update to 10.6.5.
#14
Posted 11 November 2010 - 04:46 AM
So I just updated to 10.6.5, and I still get a KP if I try to do native PM. Is there anyone out there with an EVGA X58 board with a bios newer than sz2z that has native PM working?
On this thread on osx86.co, they claim the following:
So what is Gigabyte doing differently for 6 core/32nm support that makes their bioses continue to work with native PM?
On this thread on osx86.co, they claim the following:
Quote
First thing we found was BIOS version is very important if you want to continue any further you must be running BIOS E758_SZ2Z. If you don't know how to flash your bios don't bother reading this post any more. After flashing don't load bios setting from memory enter the settings manually. BIos settings should be similar to what you had before the flash but you should enable HPET 64bit and speedster and CxE function to C1E. Don't forget the AHCI needs to be enabled. reload to make sure it works correctly before continuing before adding anything else.
The reason for this version is because anything later supports the 6 core CPU and that is giving a KP when using native power management.
The reason for this version is because anything later supports the 6 core CPU and that is giving a KP when using native power management.
So what is Gigabyte doing differently for 6 core/32nm support that makes their bioses continue to work with native PM?
#15
Posted 23 December 2010 - 09:51 AM
On EVGA's website, they list these differences between bios 44, which breaks AppleIntelCPUPM support, and SZ2Z, which enables it:
To me, this suggests three possible causes.
1. The CPU microcode
2. The "Turbo Performance" option
3. Notice how the naming conventions of the bioses change from SZxx to just two digits. Does this mean that perhaps they did a complete rewrite and that the cause of our problem could be something unrelated to the added features in version 44?
I'm wondering if I replaced the CPU microcode in bios version 77 from EVGA with the microcode from a Gigabyte X58 board's bios if that would fix it, but I'm not sure if that's really the problem.
Quote
44 Updates:
Supports 32nm Desktop CPUs
Added "Turbo Performance" option (may boost performance if enabled)
Fixes 1866MHz Memory Ratio
SZ2Z Updates:
Added "Reserve RC for Bridge" - Enables resources for 5 VGA's
Added High Precision Event Timer
Disables PS2 keyboard in device manager if no PS2 keyboard present
Supports 32nm Desktop CPUs
Added "Turbo Performance" option (may boost performance if enabled)
Fixes 1866MHz Memory Ratio
SZ2Z Updates:
Added "Reserve RC for Bridge" - Enables resources for 5 VGA's
Added High Precision Event Timer
Disables PS2 keyboard in device manager if no PS2 keyboard present
To me, this suggests three possible causes.
1. The CPU microcode
2. The "Turbo Performance" option
3. Notice how the naming conventions of the bioses change from SZxx to just two digits. Does this mean that perhaps they did a complete rewrite and that the cause of our problem could be something unrelated to the added features in version 44?
I'm wondering if I replaced the CPU microcode in bios version 77 from EVGA with the microcode from a Gigabyte X58 board's bios if that would fix it, but I'm not sure if that's really the problem.
#16
Posted 05 April 2011 - 03:57 AM
banini_jeque, on Dec 23 2010, 09:51 AM, said:
On EVGA's website, they list these differences between bios 44, which breaks AppleIntelCPUPM support, and SZ2Z, which enables it:
To me, this suggests three possible causes.
1. The CPU microcode
2. The "Turbo Performance" option
3. Notice how the naming conventions of the bioses change from SZxx to just two digits. Does this mean that perhaps they did a complete rewrite and that the cause of our problem could be something unrelated to the added features in version 44?
I'm wondering if I replaced the CPU microcode in bios version 77 from EVGA with the microcode from a Gigabyte X58 board's bios if that would fix it, but I'm not sure if that's really the problem.
To me, this suggests three possible causes.
1. The CPU microcode
2. The "Turbo Performance" option
3. Notice how the naming conventions of the bioses change from SZxx to just two digits. Does this mean that perhaps they did a complete rewrite and that the cause of our problem could be something unrelated to the added features in version 44?
I'm wondering if I replaced the CPU microcode in bios version 77 from EVGA with the microcode from a Gigabyte X58 board's bios if that would fix it, but I'm not sure if that's really the problem.
#17
Posted 07 April 2011 - 06:31 AM
@kdtt
What boot loader did you use for Lion?
And is your sleep working?
What boot loader did you use for Lion?
And is your sleep working?
#18
Posted 07 April 2011 - 06:02 PM
Hangten, on Apr 7 2011, 07:31 AM, said:
@kdtt
What boot loader did you use for Lion?
And is your sleep working?
What boot loader did you use for Lion?
And is your sleep working?
#19
Posted 27 February 2012 - 06:53 PM
Is there anybody who gets x22 turbo in msrtools?
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users



Sign In
Create Account









