Spectre patch tested on the iPhone 6 - 40% DROP in both single and multi core performance

[ScratchCat]

This seems to be in the same ballpark as the battery slow down issue. Probably the tester installed the update on battery such that it was drained, leading to a later slowdown when benchmarked. 

 

The update has been out for a while, if the performance loss was so great then people would notice , given that after the battery fiasco many people will be testing their phones to check for a slowdown.

[Tech_Dreamer]

so it's just the same as when an iphone gets a new release? *grabs salty pops*

[mr moose]

yeah, I'm going to wait for more consistent testing form a few more sources before making a judgment.

[Hunter259]

1 hour ago, ScratchCat said:

This seems to be in the same ballpark as the battery slow down issue. Probably the tester installed the update on battery such that it was drained, leading to a later slowdown when benchmarked. 

 

The update has been out for a while, if the performance loss was so great then people would notice , given that after the battery fiasco many people will be testing their phones to check for a slowdown.

That's not how that works. The battery slow down is caused by lower voltage than expected. Battery percentage is based on the amp hours left in the battery.

[79wjd]

2 minutes ago, Hunter259 said:

That's not how that works. The battery slow down is caused by lower voltage than expected. Battery percentage is based on the amp hours left in the battery.

A lower battery charge will affect performance too. My 6s+ at 50% charge scores about 20% lower than when it's at 80%+.

[Hunter259]

2 minutes ago, djdwosk97 said:

A lower battery charge will affect performance too. My 6s+ at 50% charge scores about 20% lower than when it's at 80%+.

That's different than what he was talking about. The battery slow down is due to voltage. Obviously the percentage is just normal everyday battery management.

[mynameisjuan]

10 minutes ago, Hunter259 said:

That's not how that works. The battery slow down is caused by lower voltage than expected. Battery percentage is based on the amp hours left in the battery.

As a battery drains it slowly loses voltage, hence the slow down.

[leadeater]

10 minutes ago, Hunter259 said:

That's not how that works. The battery slow down is caused by lower voltage than expected. Battery percentage is based on the amp hours left in the battery.

As the battery drains the performance is reduced, that is part of Apples battery management they introduced. As the battery discharges the peak current output decreases.

[Hunter259]

1 minute ago, leadeater said:

As the battery drains the performance is reduced, that is part of Apples battery management they introduced. As the battery discharges the peak current output decreases.

The quote the voltage not being able to be supplied as the reason why you have maximum performance drop. With my old battery I had a performance maximum drop combined with the already in place performance management for daily use.

[Valentyn]

5 hours ago, GatioH said:

The fact that my iPhone 7 is a slug after the update really concerns me, It’s just over a year old...

 

3 hours ago, Whiskers said:

Not sure I trust the validity of this claim, at least not yet.

 

I just ran a Geekbench run on my iPhone 7 Plus, and scored a single core score of 3495 and multicore of 5923; within margin of error of my previous test results. I'm currently on iOS 11.2.5 beta 5, too.

 

No Issues on the iphone 7 Plus here either. Within margin of error on iPhone 7, 8, and X here.
The Reddit that brought this up, also has people with iPhone 6's not seeing any drops like that at all.

 

 

 

 

 

Dat massive 40% performance drop in action folks! Poor iPhone 6

[Zodiark1593]

12 minutes ago, Hunter259 said:

That's not how that works. The battery slow down is caused by lower voltage than expected. Battery percentage is based on the amp hours left in the battery.

As the volts gradually lower over the battery's runtime, more current needs to be drawn for regulation to maintain stable voltage. This in turn brings down voltage even further due to the sag from the additional current.

 

A battery meter based solely on amp hours would exhibit a faster drain toward the last portion of capacity.

[leadeater]

2 minutes ago, Hunter259 said:

The quote the voltage not being able to be supplied as the reason why you have maximum performance drop. With my old battery I had a performance maximum drop combined with the already in place performance management for daily use.

 

1 minute ago, Zodiark1593 said:

As the volts gradually lower over the battery's runtime, more current needs to be drawn for regulation to maintain stable voltage. This in turn brings down voltage even further due to the sag from the additional current.

 

A battery meter based solely on amp hours would exhibit a faster drain toward the last portion of capacity.

Nice, beat me to it. Good explanation 

[ScratchCat]

37 minutes ago, Hunter259 said:

That's not how that works. The battery slow down is caused by lower voltage than expected. Battery percentage is based on the amp hours left in the battery.

Edit: already explained, ignore.

 

The voltage of a battery can vary with current capacity. Most Lithium ion batteries have a charged voltage of 4.2V and a discharged state of 3.5-3.7V. 

Capacity does mean the remaining amp hours , however the terminal voltage does change over the capacity of battery.

 

(In this context capacity means current capacity from full)

Edited January 13, 2018 by ScratchCat

[Bit_Guardian]

3 hours ago, Trixanity said:

Well, it isn't exactly BS. Intel is the company most affected and the company that has done the most shady things before, during and after. So to call it BS is more misleading than anything else.

Yes, pretty much any modern processor is vulnerable to Spectre to some degree. That's pretty clear. That doesn't detract from Intel being the biggest culprit in this case.

The most shady things? Do tell.

 

Chief contributor to the Linux and BSD kernels since it was informed, release of BIOS microcode patch component the day after public announcement, pretty quick diagnosis of their whole fleet of processors going clear back to the Pentium II...

 

The only thing you can remotely shake a stick at is Krzanich's share sale. Unfortunately for the tinfoil hat people, he signed an agreement to sell clear back in April, before Google informed Intel in June. So, if anyone's shady, it's the board of directors in this case, and they're not executives. They're the chief share holders for the company.

[IceCold008]

Thank god i dont touch crapple stuff. 

[Trixanity]

1 hour ago, Bit_Guardian said:

The most shady things? Do tell.

 

Chief contributor to the Linux and BSD kernels since it was informed, release of BIOS microcode patch component the day after public announcement, pretty quick diagnosis of their whole fleet of processors going clear back to the Pentium II...

 

The only thing you can remotely shake a stick at is Krzanich's share sale. Unfortunately for the tinfoil hat people, he signed an agreement to sell clear back in April, before Google informed Intel in June. So, if anyone's shady, it's the board of directors in this case, and they're not executives. They're the chief share holders for the company.

[DocSwag]

7 hours ago, Alexzz_ said:

i didn't know iphones used intel cpu's i always assumed they just use some qualcomm or apple specific processor.

They do use an Apple designed CPU.

7 hours ago, tjcater said:

They don't, the spectre flaw effects a very wide range of processors. What you are thinking of is meltdown which was exclusive to intel.

Edit: well it seems that some processors outside of Intel were impacted.

 

6 hours ago, Alexzz_ said:

ah, that must be it, i forgot that there where multiple flaws and that only meltdown was intel specific.

Meltdown actually affects Apple as well.

https://support.apple.com/en-us/HT208394

Quote

Meltdown is a name given to an exploitation technique known as CVE-2017-5754 or "rogue data cache load." The Meltdown technique can enable a user process to read kernel memory. Our analysis suggests that it has the most potential to be exploited. Apple released mitigations for Meltdown in iOS 11.2, macOS 10.13.2, and tvOS 11.2. watchOS did not require mitigation. Our testing with public benchmarks has shown that the changes in the December 2017 updates resulted in no measurable reduction in the performance of macOS and iOS as measured by the GeekBench 4 benchmark, or in common Web browsing benchmarks such as Speedometer, JetStream, and ARES-6.

 

[RagnarokDel]

2 hours ago, Valentyn said:

 

 

No Issues on the iphone 7 Plus here either. Within margin of error on iPhone 7, 8, and X here.
The Reddit that brought this up, also has people with iPhone 6's not seeing any drops like that at all.

 

 

 

 

 

Dat massive 40% performance drop in action folks! Poor iPhone 6

That's not how you do benchmarks...

[leadeater]

11 minutes ago, RagnarokDel said:

That's not how you do benchmarks...

True but even with the improper testing method if there was 40% drop you'd see it as that's far greater than per unit performance variant.

[Brooksie359]

6 hours ago, RadiatingLight said:

FYI, you need both the windows update and BIOS update to be protected from meltdown/spectre, so people testing just the windows update will not see nearly as big a drop as people testing with the windows update + the BIOS update

Wait I thought you only needed to us are the bios for meltdown not specter? Maybe I am wrong but that's what I had thought others had said.

[captain_to_fire]

8 hours ago, Alexzz_ said:

i didn't know iphones used intel cpu's i always assumed they just use some qualcomm or apple specific processor.

Wut? 

[Bit_Guardian]

30 minutes ago, Brooksie359 said:

Wait I thought you only needed to us are the bios for meltdown not specter? Maybe I am wrong but that's what I had thought others had said.

Microcode updates can fix one variant of Spectre outright, but apps still need to be recoded to defend against the other.

[Zodiark1593]

2 hours ago, Bit_Guardian said:

Microcode updates can fix one variant of Spectre outright, but apps still need to be recoded to defend against the other.

Unfortunately, one would have to rely on multiple app developers to acually code and issue the updates, some may not be keen on doing.

[Bit_Guardian]

3 minutes ago, Zodiark1593 said:

Unfortunately, one would have to rely on multiple app developers to acually code and issue the updates, some may not be keen on doing.

It didn't take long to get Clang and GCC updated. Lazy-ass development houses...

 

Actually it probably is difficult for big pieces of software just to assess where the boundaries are, so let's wait and see.

[System Error Message]

unlike x86 CPUs that laugh at ARM in performance (both math and logic), many apps have a bloated UI that makes use of the CPU. With geekbench scores much lower, for sure your UI responsiveness is going to be lower. For instance in many mobile apps, the UI itself has so much code to track your touch and activities of the phone itself (like networking).

 

So phones are going to be sluggish. If your phone uses a generic ARM CPU, the impact is not so bad, snapdragons with custom qualcomm chips will suffer a bigger performance hit than generic arm but will still perform better.than generic ARM.