(Update: announcement imminent) A new generation of thermal throttle - new intel laptop chip found

[williamcll]

Update: announcement is believed to be April 2nd

 

It is slightly faster than the previous generation.

Source translated from german:

Quote

Intel's confirmation that the 10th generation of core CPUs for the desktop PC will offer up to ten cores is still pending, but is considered an open secret . Despite further optimized 14 nm production, the power consumption should increase, so notebooks should remain with a maximum of eight cores. The editorial team had already expressed this assumption at CES 2020, and the first benchmark entries for a Core i9-10980HK now support it. The well-known for the accuracy of its information Apisak has on Twitter three Geekbench - results published, allegedly each created on a core i9-10980HK. And according to the benchmark, the processor has only eight cores and sixteen threads like the Core i9-9980H (K) based on Coffee Lake-H Refresh, which has been available since spring 2019 .

 

Most recently, the editors measured the Core i9-9880H in Geekbench with the Apple MacBook 16 inch (test) . Compared to this processor, which is allowed to operate in the MacBook with a permanently high power consumption of 60 to 65 watts, the performance of the CPU discovered by APISAK is on average three percent higher than the three results, by two (single-core) or nine (multi-core). A base clock of 3.1 GHz was read out, the maximum clock determined by Geekbench fluctuated between 4.4 and 5.0 GHz. However, such information should always be viewed with extreme caution.

Source:https://browser.geekbench.com/v4/cpu/15153782

https://www.computerbase.de/2020-01/intel-core-i9-10980hk-erste-benchmarks/

Thoughts: since this is probably an engineering sample I would assume the clock speed would be lower, however, it would be a tough fight against the upcoming ryzen 4800H which is only slightly lower in clocks, but considering how intel mobile cpus have rarely maintained their temperatures this is going to be an L for intel.

[Chen G]

I hate how people on the internet talk about "throttle" without knowing what it actually means.

[Fasauceome]

More performance

 

More heat 

 

more money for big boy Intel

[Zando_]

23 minutes ago, Chen G said:

I hate how people on the internet talk about "throttle" without knowing what it actually means.

^^^^

Hot =/= throttle. Dropping below base clocks is proper throttling AFAIK, though you could call dropping below reasonable clocks that as well. 

My work mac (base model 2019 13" MBP) can slap 97C for extended periods and still keep over 2.5Ghz boost consistently (1.4Ghz base clock, 3.9 max single core boost). 

[Bombastinator]

We shall see.  Apple did move away a bit with its 16” laptop from its heat problem, and 8/16 is a lot. Irrc it was stated once that the “tendy” stuff worked better on laptops than it did on desktops.  The combo might make for some fast machines.

[Mira Yurizaki]

5 minutes ago, Zando Bob said:

^^^^

Hot =/= throttle. Dropping below base clocks is proper throttling AFAIK, though you could call dropping below reasonable clocks that as well. 

My work mac (base model 2019 13" MBP) can slap 97C for extended periods and still keep over 2.5Ghz boost consistently (1.4Ghz base clock, 3.9 max single core boost). 

Though playing devil's advocate or tinfoil hat man, who's to say the base clock wasn't set that low deliberately to avoid hitting this definition?

[Zando_]

9 minutes ago, Mira Yurizaki said:

Though playing devil's advocate or tinfoil hat man, who's to say the base clock wasn't set that low deliberately to avoid hitting this definition?

It's set that low to allow it to clock very low when idle/under light load. It sits around 1.2-2.2 depending on workload, can stay in the 40s-70s under light to medium load without ever spinning the fans. 

Aka it's there to allow for a 0rpm fan mode, there are higher base clocked options if you want them (IIRC they spec up to a 2.4Ghz base clock i5, which is still a clock the cooling solution in these Macs can handle under load). 

[TempestCatto]

12 minutes ago, Bombastinator said:

Irrc it was stated once that the “tendy” stuff worked better

I too, love chicken tendies

 

[Chen G]

19 minutes ago, Mira Yurizaki said:

Though playing devil's advocate or tinfoil hat man, who's to say the base clock wasn't set that low deliberately to avoid hitting this definition?

Because it doesn't matter if it was.

You knew what the base clock is, they tell you right on the product descriptions, that's what you're buying, and there's no reason to complain when it clearly does reach that base clock. And there is certainly no reason to say it's "throttling" when it clearly isn't.

[MageTank]

56 minutes ago, Chen G said:

Because it doesn't matter if it was.

You knew what the base clock is, they tell you right on the product descriptions, that's what you're buying, and there's no reason to complain when it clearly does reach that base clock. And there is certainly no reason to say it's "throttling" when it clearly isn't.

How exactly would you define the reduction of clock speed as a result of higher thermals / power consumption? I've referred to it as a boost throttle in the past, and I do not see how that would be considered an inaccurate statement. Now, I wouldn't consider this false advertising or me not getting what I paid for, but I do need to call it something, and that makes the most logical sense. By your definition, we can't consider it "underclocking" as it's not falling under the base speeds, which is what we paid for. 

 

If we can't call it throttling, then we need to come up with a general term that everyone understands in relation to what exactly is happening when your boost clocks are reduced.

[Bombastinator]

7 minutes ago, MageTank said:

How exactly would you define the reduction of clock speed as a result of higher thermals / power consumption? I've referred to it as a boost throttle in the past, and I do not see how that would be considered an inaccurate statement. Now, I wouldn't consider this false advertising or me not getting what I paid for, but I do need to call it something, and that makes the most logical sense. By your definition, we can't consider it "underclocking" as it's not falling under the base speeds, which is what we paid for. 

 

If we can't call it throttling, then we need to come up with a general term that everyone understands in relation to what exactly is happening when your boost clocks are reduced.

Part of the problem is the concept of boost really isn’t taken into consideration completely when talking about overclocking. A boost is sort of a temporary automatic overclock.  “Under boosting” maybe?  “Boost throttling”?  CPUs didn’t vary in speed when overclocking became a thing.  Setting boost clock to all cores is basically an overclock, except it only sort of is.  The problem I think is that the term “overclock” has gotten muddied.

[Bombastinator]

1 hour ago, TempestCatto said:

I too, love chicken tendies

 

You turn green and your ears get huge.  That must hurt a lot.

[Mira Yurizaki]

Just now, Bombastinator said:

Part of the problem is the concept of boost really isn’t taken into consideration completely when talking about overclocking. A boost is sort of a temporary automatic overclock.  “Under boosting” maybe?  “Boost throttling”?  CPUs didn’t vary in speed when overclocking became a thing.  Setting boost clock to all cores is basically an overclock, except it only sort of is.  The problem I think is that the term “overclock” has gotten muddied.

The more technical sounding term for "turbo boosting" from what I've seen is "dynamic overclocking" Which to me makes sense. Under stock/default configurations, you're not asking the processor to work that fast all the time.

[TempestCatto]

27 minutes ago, Bombastinator said:

You turn green and your ears get huge.  That must hurt a lot.

But I'm loved and famous as hell. Feels good man.

[CephDigital]

Looks like my 8750H is gonna last a long time in the mobile space compared to newer mobile Intel chips (guess Im going AMD for my next laptop)

[Chen G]

3 hours ago, MageTank said:

How exactly would you define the reduction of clock speed as a result of higher thermals / power consumption? I've referred to it as a boost throttle in the past, and I do not see how that would be considered an inaccurate statement. Now, I wouldn't consider this false advertising or me not getting what I paid for, but I do need to call it something, and that makes the most logical sense. By your definition, we can't consider it "underclocking" as it's not falling under the base speeds, which is what we paid for. 

 

If we can't call it throttling, then we need to come up with a general term that everyone understands in relation to what exactly is happening when your boost clocks are reduced.

“Throttling” should refer to when it dips below base clock.

 

Not being able to sustain boost clock is to be expected and completely normal behavior. It doesn’t need a description. It’s expected even in regular desktop computers.

 

what is actually exceptional and not normal, is if you have so much cooling that it can boost all the way.

[Dabombinable]

It'd still beat the shit out of my previous laptop CPU. Considering even my tablet's M3-6Y30 does...

[mr moose]

I think half the problem here with nomenclature is that a good portion of today's enthusiast (active on forums that is) aren't familiar with technology from 20 years ago when variable clock rate was becoming a thing.   For those of us who had to set everything manually (no software to do it for you) then test then reset and retest and so on, Throttling has a different connotation.    Some people seem to be under the impression that throttling is the result of bad CPU design, or that throttling is bad full stop.    Neither of those are true, what is true is that throttling is the restriction of clock speed only.  Not just from a specific point only but any restriction.  As soon as the CPU gets to hot it throttles down (restricts itself) to reduce heat damage.

 

 

[Bombastinator]

3 hours ago, Chen G said:

“Throttling” should refer to when it dips below base clock.

 

Not being able to sustain boost clock is to be expected and completely normal behavior. It doesn’t need a description. It’s expected even in regular desktop computers.

 

what is actually exceptional and not normal, is if you have so much cooling that it can boost all the way.

Which is arguably a tad obnoxious in that what it does is base a chip’s listed stats on cooling ability.  If boost clock is variable by installation and that installation is pre done, as it is in a laptop, the rating of the device should reflect that. If a laptop has a cpu with a base of randomly 3.0 and a boost of randomly 4.0. It shouldn’t say boost of up to 4.0 if it’s only ever going to hit 3.5 because of the cooling in the machine.  It should say base 3.0 boost up to 3.5.

[NumLock21]

28 minutes ago, mr moose said:

As soon as the CPU gets to hot it throttles down (restricts itself) to reduce heat damage.

That’s what it was back then and it’s still is today. Old cpu didn’t have this feature, cpu overheats and it’s bye bye cpu. Thermal throttle is a great feature, save a bunch of enthusiast from having their cpus fried and spending more money. 

[mr moose]

3 minutes ago, NumLock21 said:

That’s what it was back then and it’s still is today. Old cpu didn’t have this feature, cpu overheats and it’s bye bye cpu. Thermal throttle is a great feature, save a bunch of enthusiast from having their cpus fried and spending more money. 

That was exactly my point.  People today don't understand why it's a thing because they have no experience where it has come from and why, hence they have drawn some rather arbitrary conclusions as to what it is and when it applies.

[Fnige]

10 minutes ago, Bombastinator said:

Which is arguably a tad obnoxious in that what it does is base a chip’s listed stats on cooling ability.  If boost clock is variable by installation and that installation is pre done, as it is in a laptop, the rating of the device should reflect that. If a laptop has a cpu with a base of randomly 3.0 and a boost of randomly 4.0. It shouldn’t say boost of up to 4.0 if it’s only ever going to hit 3.5 because of the cooling in the machine.  It should say base 3.0 boost up to 3.5.

because marketing

[Bombastinator]

Just now, CircleTech said:

you mean like this:

 

Those would be long and thin not large but I suspect that would be uncomfortable too.

[BiG StroOnZ]

10 hours ago, TempestCatto said:

I too, love chicken tendies

 

Spoiler

 

 

[NumLock21]

2 hours ago, mr moose said:

That was exactly my point.  People today don't understand why it's a thing because they have no experience where it has come from and why, hence they have drawn some rather arbitrary conclusions as to what it is and when it applies.

That’s because newcomer these days mostly just watch YouTube channels, and don’t really discover for themselves the true origin of why thermal throttling was invented in the first place.

Thermal throttling isn’t associated with performance, that’s what most users of today think it is. It’s a safety feature that prevented the cpu from physically destroying itself. Too bad some won’t believe that, and only believe in who they trust says, even when there are those who has been in the field way before them.