Intel slowdown

[wynand32]

So far, I'm finding battery use to be roughly equivalent to Edge. Nothing scientific, but the average mW drain is similar (using BatteryBar to eyeball things). Even if it were worse, I'd still continue to use Firefox for the performance benefits. Not only is browsing much faster, but the touchpad/keyboard lag is significantly less and general performance is better.

That said, I have to imagine this is all related to the funky Intel microcode use in the patch/firmware update, and that once Intel releases its new (hopefully improved) version then Microsoft will update and things will get better.

 

[GreyFox7]

So far, I'm finding battery use to be roughly equivalent to Edge. Nothing scientific, but the average mW drain is similar (using BatteryBar to eyeball things). Even if it were worse, I'd still continue to use Firefox for the performance benefits. Not only is browsing much faster, but the touchpad/keyboard lag is significantly less and general performance is better.

That said, I have to imagine this is all related to the funky Intel microcode use in the patch/firmware update, and that once Intel releases its new (hopefully improved) version then Microsoft will update and things will get better.

Yeah, my expectation is there will be several months of fine tuning after they get the major problems solved. So.. A couple rounds of Intel Firmware updates and some OS and Browser tweaking to better deal with the new realities of Spectresized CPU behavior.

Intel vows to release new "fixed" CPUs this year (it was previously speculated this could take much longer) but it's unclear what that means and what the performance of those CPUs will be or which generation CPUs will be fixed (8th, 9th, 10th). Best case, Intel is sitting on (holding back) performance increases they can use to offset the adverse effects of the "new secure architecture" otherwise new CPUs may not offer a performance increase and possibly even a performance decrease from previous Pre-Spectre performance levels.

 

[wynand32]

Yeah, my expectation is there will be several months of fine tuning after they get the major problems solved. So.. A couple rounds of Intel Firmware updates and some OS and Browser tweaking to better deal with the new realities of Spectresized CPU behavior.

Intel vows to release new "fixed" CPUs this year (it was previously speculated this could take much longer) but it's unclear what that means and what the performance of those CPUs will be or which generation CPUs will be fixed (8th, 9th, 10th). Best case, Intel is sitting on (holding back) performance increases they can use to offset the adverse effects of the "new secure architecture" otherwise new CPUs may not offer a performance increase and possibly even a performance decrease from previous Pre-Spectre performance levels.

It's a great time to be AMD.

 

[wynand32]

So, Microsoft pushed out some methods for mitigating the Intel Spectre Variant 2 patch that's suspected as causing so many issues. I'm pretty convinced the patch/firmware update is the root cause of the recent slowdown my SB2 15" has experienced -- I've applied the patch, but haven't been able to determine yet if the performance is truly better.

Here's the info though if anyone wants to apply Microsoft's fix: Grab this Microsoft update now if your PC was affected by the Spectre fixes

I suspect Patch Tuesday will have some additional fixes, and I'm hoping a firmware update that truly resolves the performance issues. I'd love to get my speedy SB2 back.

 

[GreyFox7]

I'd love to get my speedy SB2 back.

I don't think the same speed will ever be achieved with mitigation in place. AFAIK the old speed was the result of a cheat and now that they got caught they will have to do it straight or er ah as close to straight as possible. The new CPUs post Spectre will be left to find another way as this was not an implementation error rather it was a design/architectural fault and it was intentional. Either they never foresaw the ramifications or they ignored them. They and everyone else in the industry that used the same architecture ala ARM, AMD, IBM, Sun/Oracle, Apple, etc.

 

[wynand32]

I don't think the same speed will ever be achieved with mitigation in place. AFAIK the old speed was the result of a cheat and now that they got caught they will have to do it straight or er ah as close to straight as possible. The new CPUs post Spectre will be left to find another way as this was not an implementation error rather it was a design/architectural fault and it was intentional.

I do believe that Intel's first attempt to fix this was wonky, and that the next fix won't have so profound an impact. I expect some slowdown, but not what I've been experiencing. The touchpad and keyboard shouldn't be lagging the way it does, for example. And at worst, I expect the SB2 to perform no worse than any other system with equivalent components, which doesn't seem to be the case right now.

If the reality is that every system is slowed down somewhat, and the SB2 no more or less, than we'll all just have to live with it. I suspect that right now, this isn't the case -- because I think Microsoft was very quick to implement the bad Intel microcode.

 

[GreyFox7]

I do believe that Intel's first attempt to fix this was wonky, and that the next fix won't have so profound an impact. I expect some slowdown, but not what I've been experiencing. The touchpad and keyboard shouldn't be lagging the way it does, for example. And at worst, I expect the SB2 to perform no worse than any other system with equivalent components, which doesn't seem to be the case right now.

If the reality is that every system is slowed down somewhat, and the SB2 no more or less, than we'll all just have to live with it. I suspect that right now, this isn't the case -- because I think Microsoft was very quick to implement the bad Intel microcode.

Yes, it will improve from the current level but won't be like b4.

 

[GreyFox7]

With the Spectre variant 2 fix disabled per Microsoft's recent new update the Passmark performance score is returned to 2045.9 as it was before the Firmware update.

 

[wynand32]

With the Spectre variant 2 fix disabled per Microsoft's recent new update the Passmark performance score is returned to 2045.9 as it was before the Firmware update.

Awesome, thanks for the report. By the way, I ran the update but it gave zero feedback -- did yours give any kind of indication that it worked when you ran it?

Also, I went ahead and implemented the manual registry update as well. I'm thinking it probably just duplicated whatever the update did, and I've not seen any bad effects.

 

[GreyFox7]

Awesome, thanks for the report. By the way, I ran the update but it gave zero feedback -- did yours give any kind of indication that it worked when you ran it?

Also, I went ahead and implemented the manual registry update as well. I'm thinking it probably just duplicated whatever the update did, and I've not seen any bad effects.

I just ran the Powershell commands, did the registry changes, and rebooted so it would take effect.

 

[wynand32]

I just ran the Powershell commands, did the registry changes, and rebooted so it would take effect.

Those are actually one step, right? Running the commands creates the new registry keys... just want to make sure I did the right thing.


Sent from my iPad using Tapatalk

 

[GreyFox7]

Those are actually one step, right? Running the commands creates the new registry keys... just want to make sure I did the right thing.


Sent from my iPad using Tapatalk

Go to Speculation Control Validation PowerShell Script

Download SpeculationControl.zip to a local folder.

Extract the contents to a local folder, for example C:\ADV180002

Run the PowerShell module to validate the protections are enabled or disabled

Start PowerShell, then (using the example above) copy and run the following commands:

Save the current execution policy so it can be reset

PS> $SaveExecutionPolicy = Get-ExecutionPolicy

PS> Set-ExecutionPolicy RemoteSigned -Scope Currentuser

PS> CD C:\ADV180002\SpeculationControl

PS> Import-Module .\SpeculationControl.psd1

PS> Get-SpeculationControlSettings

The output of this PowerShell script will resemble the following.

Speculation control settings for CVE-2017-5715 [branch target injection]
For more information about the output below, please refer to https://support.microsoft.com/en-in/help/4074629
Hardware support for branch target injection mitigation is present: True
Windows OS support for branch target injection mitigation is present: True
*Windows OS support for branch target injection mitigation is enabled: False
*Windows OS support for branch target injection mitigation is disabled by system policy: True
*Windows OS support for branch target injection mitigation is disabled by absence of hardware support: False
Speculation control settings for CVE-2017-5754 [rogue data cache load]
Hardware requires kernel VA shadowing: True
Windows OS support for kernel VA shadow is present: True
Windows OS support for kernel VA shadow is enabled: True
Windows OS support for PCID performance optimization is enabled: True [not required for security]

* = changed lines
PS> # Reset the execution policy to the original state

PS> Set-ExecutionPolicy $SaveExecutionPolicy -Scope Currentuser

SORRY for the crappy editing... using the FCU OSK = Biggest POS ever.