Supplemental update 2 not working?!

[Billy Shears]

Guys,
I'm kinda losing my mind here. Yesterday I finally got my new i7 2.6, 32GB, 560X, 1TB MacBook Pro 15 inch. I installed all the updates that were needed, BUT, using Intel's Power Gadget, I see that CPU clock fluctuations are not fixed. Am I missing something obvious? Is my MacBook defective?

Thanks

 

[IngerMan]

What program\App were you running for the snap shot? Geekbench 4?

 

[Billy Shears]

Yes, Geekbench 4

 

[IngerMan]

Well it looks normal. I ran it and have a 2017 15" MBP T1 chip

 

[Billy Shears]

Shouldn't it be a lot more stable, without big dips and spikes in performance (GHz)?

 

[IngerMan]

I would say it's the application running not the machine.
[doublepost=1538515240][/doublepost]Run this test for full CPU test turbo. My stats below

https://corona-renderer.com/benchmark

 

[Billy Shears]

I would say it's the application running not the machine.
[doublepost=1538515240][/doublepost]Run this test for full CPU test turbo. My stats below

https://corona-renderer.com/benchmark

View attachment 791505 View attachment 791506

Yes, it does indeed seem like it's normal behaviour. In panic I formatted SSD and then again installed MacOS, only to find same results. Will formatting cause any damage in the long term? Also, is running many benchmarks harmful for the computer?

 

[IngerMan]

Yes, it does indeed seem like it's normal behaviour. In panic I formatted SSD and then again installed MacOS, only to find same results. Will formatting cause any damage in the long term? Also, is running many benchmarks harmful for the computer?

Your machine will be well outdated before you can erase and write to your SSD to cause an issue. I would not worry about benchmarks and heat. They are designed to sustain the heat and they have safeguards to keep them under the limit.

 

[leman]

Its perfectly fine. Geekbench is running a series of short duration benchmarks, so your clock will shoot up when one of these benchmarks is running and then drop back to idle speeds in between. Which is again a rather illustrative case of people who lack technical knowledge (no offence intended) interpreting things they really shouldn't.

 

[Billy Shears]

Its perfectly fine. Geekbench is running a series of short duration benchmarks, so your clock will shoot up when one of these benchmarks is running and then drop back to idle speeds in between. Which is again a rather illustrative case of people who lack technical knowledge (no offence intended) interpreting things they really shouldn't.

Well, this forum's purpose is to help people, no? I know you said no offence, but I couldn't help but feel a slightly condescending tone. I have my reasons for asking, and I, by all means, do not lack technical knowledge. But OK. Thanks for the help

 

[leman]

Well, this forum's purpose is to help people, no? I know you said no offence, but I couldn't help but feel a slightly condescending tone. I have my reasons for asking, and I, by all means, do not lack technical knowledge. But OK. Thanks for the help

No, you are completely right. It's simply that I have been observing a trend lately that makes me very worried. Especially the latest discussion on throttling has had a lot of people — who clearly don't understand how CPUs/OS work or how to interpret the diagnostic tools output — throwing around their opinion as if they were experts while at the same time showing no interest in learning. I am aware of Dunning–Kruger effect and its prevalence in online discussion, but over the last few years, I feel that this has been accelerating in an alarming rate (applied to MR). People are not doubting their opinion anymore, they do not discuss, they simply assert, no matter how little actual understanding they might have in the topic — and they get directly aggressive when their opinion is challenged. Note that I don't count you among such people — you are asking a question. What usually happens though is that someone posts a graph like this and goes "Apple is lying to us, MBP is crap, its throttling like crazy, bring back Steve", and then you have bunch of other people chiming in and reinforcing this perception. I wish we here on the forums would focus more on discussing, sharing, learning from each other, and discovering things together.

I can totally imagine that I sound condescending, but my political skills are not high enough to avoid it I really meant no offence. But at the same time, saying that you have the technical knowledge and then asking this kind of question simply doesn't fit. If you know and understand this stuff, you shouldn't be alarmed by the PowerGadget readouts to begin with...

 

[Billy Shears]

But at the same time, saying that you have the technical knowledge and then asking this kind of question simply doesn't fit. If you know and understand this stuff, you shouldn't be alarmed by the PowerGadget readouts to begin with...

The thing that puzzles me is why are you labeling this question as unnecesarry and as something someone uneducated would ask. You see, you can't judge if you don't know the full picture. For years I had Windows machines, and before buying this one I read lost of articles and reviews where they analised thermal throttling of thesr machines, and almost every test had Power Gadget lines that fluctuated. Then, when fix came out, they showed how line is now stable.
I'm not familiar with how Geekbench works. I just saw lines that were spiking and droping all the time. Add to that the fact that my machine is constatly getting under 1000 score in Cinebench, and under 21000 in Geekbench, and now hopefully you realise the nature of my concern.
I'm still not completely sure that my machine is working as it should, because of lower performance when compared to other same spec'd machines.

 

[IngerMan]

You will only break a 1000 on cinebench from a very cold start. I have seen posts that it does drop linear from repeated times ran due to heat and throttling. Also if you want the highest score possible (Geek or Cine) make sure nothing else is running, check your activity monitor to verify. Its not realistic scores but if your looking for numbers, close everything and let your machine sleep for 30 plus minutes to cool down, wake and give it another 2 or 3 minutes then start the test.