Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Other Performance hit after Spectre patch (11.2.2)

jav6454

jav6454

macrumors Core
Original poster
Nov 14, 2007
22,303
6,264
1 Geostationary Tower Plaza
  • Like
Reactions: AppleMacFinder
jav6454 said:
Apparently many users are reporting that the patch to mitigate Spectre and Meltdown in ARM based CPUs is having an unwanted, yet expected, side effect.

Up to 45% performance drop due to the patch. Can anyone confirm on their end?

Forbes has even written an article with detailed benchmarks that back their claims.

https://www.forbes.com/sites/gordon...ife-iphone-performance-slowdown-throttle/amp/
Click to expand...
Many users? Where are the complaints about it all here (where the tiniest drops in framerates of some animations get multiple dedicated threads)?
 
  • Like
Reactions: Yun0, Newtons Apple and Vermifuge
bopajuice said:
Apparently the patch was implemented at 10.13.2. I have not felt any slowdown whatsoever. Why are you concerned?
Click to expand...

Does OS X use ARM? No.
[doublepost=1515641354][/doublepost]
joeblow7777 said:
Worry about it if it actually affects you. Otherwise, what does it matter what Forbes or anyone else says?
Click to expand...

Compound the battery throttling and you’ve already got less performance... think it through before saying “it won’t affect you”.
 
  • Like
Reactions: AppleMacFinder
bopajuice said:
Please explain. You Seem to know more than I do. The patch only affected ARM chips?
Click to expand...

You used 10.13.xxx that’s an OS X version. iPhones run on iPhone OS (iOS) which is at 11.2.1 before the Spectre patch. After the patch we have 11.2.2

Spectre affects Intel, AMD and ARM based CPUs. Guess what your iPhone runs?
[doublepost=1515641654][/doublepost]
C DM said:
Many users? Where are the complaints about it all here (where the tiniest drops in framerates of some animations get multiple dedicated threads)?
Click to expand...

Considering metal interacts with the kernel quite often, and also metal allows a more direct app approach to the silicone... the impact is not a “tiniest drops in framerstes”.

It will depend on the app. But benchmark apps are already seeing that performance hit.
 
jav6454 said:
You used 10.13.xxx that’s an OS X version. iPhones run on iPhone OS (iOS) which is at 11.2.1 before the Spectre patch. After the patch we have 11.2.2

Spectre affects Intel, AMD and ARM based CPUs. Guess what your iPhone runs?
[doublepost=1515641654][/doublepost]

Considering metal interacts with the kernel quite often, and also metal allows a more direct app approach to the silicone... the impact is not a “tiniest drops in framerstes”.

It will depend on the app. But benchmark apps are already seeing that performance hit.
Click to expand...

Please back up your statement that "benchmark apps are already seeing that performance hit" on iOS.

According to Apple's release note 11.2.2 only provided a patch for Safari's vulnerabilities to Spectre, so there would be no system wide performance hit. My own personal (quite extensive) benchmarking indicates performance is basically identical between 11.2.1 and 11.2.2.
 
  • Like
Reactions: Starship67
jav6454 said:
Does OS X use ARM? No.
[doublepost=1515641354][/doublepost]

Compound the battery throttling and you’ve already got less performance... think it through before saying “it won’t affect you”.
Click to expand...

I didn’t say it won’t affect you. I said worry about it if it does.
 
jav6454 said:
You used 10.13.xxx that’s an OS X version. iPhones run on iPhone OS (iOS) which is at 11.2.1 before the Spectre patch. After the patch we have 11.2.2

Spectre affects Intel, AMD and ARM based CPUs. Guess what your iPhone runs?
[doublepost=1515641654][/doublepost]

Considering metal interacts with the kernel quite often, and also metal allows a more direct app approach to the silicone... the impact is not a “tiniest drops in framerstes”.

It will depend on the app. But benchmark apps are already seeing that performance hit.
Click to expand...
You missed the point I was making with that example--something minute like tiny frame drops get noticed and discussed ad naseum here, so it would follow that soemtbing like this would already be discussed, especially if truly "many users" are experiencing and reporting it.
 
Last edited:
same performance on my phone and MacBook pro after the patch, no noticeable difference and I would have noticed it for sure at least on the laptop since I use it all day for work since i will not use the laptop provided by my work since it is windows based. I will only use my mac
 
BugeyeSTI said:
Fake news. Geekbench score went up on my Air2, 6S and X. No difference in performance on any of my devices
Click to expand...

Yup, scores went up slightly on my 8+.

5b0aa7d7-a097-4386-bd93-d400c6be5eda-png.745962
 
  • Like
Reactions: tonybarnaby and AlphaGuitarist
jav6454 said:
You used 10.13.xxx that’s an OS X version. iPhones run on iPhone OS (iOS) which is at 11.2.1 before the Spectre patch. After the patch we have 11.2.2

Spectre affects Intel, AMD and ARM based CPUs. Guess what your iPhone runs?
[doublepost=1515641654][/doublepost]

Considering metal interacts with the kernel quite often, and also metal allows a more direct app approach to the silicone... the impact is not a “tiniest drops in framerstes”.

It will depend on the app. But benchmark apps are already seeing that performance hit.
Click to expand...

My mistake, thanks for your understanding...
 
Not seeing any performance drop on my 7 or 8.

estabya said:
Please back up your statement that "benchmark apps are already seeing that performance hit" on iOS.

According to Apple's release note 11.2.2 only provided a patch for Safari's vulnerabilities to Spectre, so there would be no system wide performance hit. My own personal (quite extensive) benchmarking indicates performance is basically identical between 11.2.1 and 11.2.2.
Click to expand...
 
For those stating "Please back up your claim", pleas read the article I linked in the first post. It is an article to Forbes. Although Forbes is not known as a technical media outlet, they have a reputation to keep from false or fake news.
 
  • Like
Reactions: AppleMacFinder
jav6454 said:
For those stating "Please back up your claim", pleas read the article I linked in the first post. It is an article to Forbes. Although Forbes is not known as a technical media outlet, they have a reputation to keep from false or fake news.
Click to expand...

Interesting, but I still don't buy it. The other article that the Forbes page cites includes a single case of slowdown on one iPhone 6. You can't draw any conclusions from a test case that includes just a single device.

Like I said, unless Apple is being misleading in their release notes, this patch ONLY affects Spectre vulnerabilities in Safari. The OS level vulnerabilities that could result in system wide slowdown have yet to be patched.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back