Dell 5400 TPL Help

[neomerc]

Hello
Could anyone with a Dell latitude 5400 with i5 8365u, please share a ss of their TPL window from inside throttlestop?
I wanted to know if the PP0 power limits are locked on this laptop like some 7000 lattidue series (Latitude 7490 for example)

Thank you taking the time to read this post
Cheers

 

[unclewebb]

PP0 power limits

Most laptops leave the PP0 limits unlocked. If unlocked, I set them to 0 so they are ignored. If they are locked by the BIOS, there is nothing you can do.

 

[neomerc]

Most laptops leave the PP0 limits unlocked. If unlocked, I set them to 0 so they are ignored. If they are locked by the BIOS, there is nothing you can do.

Yes that's why I was asking regd this specific model
A 7490 in the family got has it locked, power throttles incessantly, refusing to adhere to the pl1 set in ts, while a friend's latitude 3400 with the 8365u has it unlocked (no lock icon shown in tpl), hence I was asking regd this latitude 5400 I was considering to purchase, better be safe than sorry this time

 

[unclewebb]

refusing to adhere to the pl1 set in ts

Many Dell laptops use an embedded controller which is used to enforce turbo power limits. This can cause power limit throttling at way less than the 15W TDP. An EC power limit cannot be modified by ThrottleStop.

It is usually the EC power limit and not the PP0 power limit that is the main problem. I have had better luck and less throttling with Lenovo laptops.

 

[neomerc]

Many Dell laptops use an embedded controller which is used to enforce turbo power limits. This can cause power limit throttling at way less than the 15W TDP. An EC power limit cannot be modified by ThrottleStop.

It is usually the EC power limit and not the PP0 power limit that is the main problem. I have had better luck and less throttling with Lenovo laptops.

Right, is there a way then, to check for whether or not a laptop is gonna power throttle with load put on it, using data displayed by TS or any other software, to check for the EC power limit as you mentioned?
For ex, the latitude 3400 I mentioned, has the PP0 unlocked (no lock sign), and faithfully adheres to the PL1/2 values set in ts, while benching, only downclocking with temp based throttling (when the HOT symbol appears in TS)
However, the 7490 in family, it refuses to listen to the PL1/2 set in TS, always power throttling, to 22W on prolonged bench loads, with EDP and PL1 flashing in red when that and also shows its PP0 limits as locked with the lock sign, all other factors remaining the same between the two laptops

Really would help if there was a way to check for whether there is an EC limit, as it'd potentially save me from getting another laptop with this power throttle cripple
Sidenote: I am also looking at the Lenovo t480, however that one isn't avlbl locally, only able to get it via Amazon refurb, which I've got little trust over wrt actually delivering on listed specs, with a great deal of variance

 

[PHVM_BR]

Right, is there a way then, to check for whether or not a laptop is gonna power throttle with load put on it, using data displayed by TS or any other software, to check for the EC power limit as you mentioned?
For ex, the latitude 3400 I mentioned, has the PP0 unlocked (no lock sign), and faithfully adheres to the PL1/2 values set in ts, while benching, only downclocking with temp based throttling (when the HOT symbol appears in TS)
However, the 7490 in family, it refuses to listen to the PL1/2 set in TS, always power throttling, to 22W on prolonged bench loads, with EDP and PL1 flashing in red when that and also shows its PP0 limits as locked with the lock sign, all other factors remaining the same between the two laptops

Really would help if there was a way to check for whether there is an EC limit, as it'd potentially save me from getting another laptop with this power throttle cripple
Sidenote: I am also looking at the Lenovo t480, however that one isn't avlbl locally, only able to get it via Amazon refurb, which I've got little trust over wrt actually delivering on listed specs, with a great deal of variance

Have you tried changing the power limits and checking Lock MMIO?

 

[neomerc]

Have you tried changing the power limits and checking Lock MMIO?

Yeah of course I did, the above data is after applying those options only

 

[PHVM_BR]

Yeah of course I did, the above data is after applying those options only

You can change the IMON slope so that your power limits can go higher.

 

[neomerc]

You can change the IMON slope so that your power limits can go higher.

I am aware of that method, used it on my Acer helios, however that entailed unlocking the bios and digging through the settings there
Do you know how to do it via ts here, cause I'm unable to find it for this latitude 7490

 

[PHVM_BR]

I am aware of that method, used it on my Acer helios, however that entailed unlocking the bios and digging through the settings there
Do you know how to do it via ts here, cause I'm unable to find it for this latitude 7490

If there is no specific guide for the Latitude 7490, you just need to extract the variables and find which ones are related to the IMON slope.

Via Throttlestop it will not be possible.

Perhaps a palliative alternative could work.

On my Precision 7540 I discovered that increasing the power limits and checking Lock MMIO after returning from sleep the limits are applied.

These higher limits may run all day or may return to default in a matter of minutes depending on the duration and intensity of the load applied to the CPU.

For example, a benchmark that lasts several minutes forcing the CPU to 90°C or more will cause the original limits to return after about 3 minutes.

 

[neomerc]

Y

If there is no specific guide for the Latitude 7490, you just need to extract the variables and find which ones are related to the IMON slope.

Via Throttlestop it will not be possible.

Perhaps a palliative alternative could work.

On my Precision 7540 I discovered that increasing the power limits and checking Lock MMIO after returning from sleep the limits are applied.

These higher limits may run all day or may return to default in a matter of minutes depending on the duration and intensity of the load applied to the CPU.

For example, a benchmark that lasts several minutes forcing the CPU to 90°C or more will cause the original limits to return after about 3 minutes.

Yes I'd tried that too, putting it to sleep after applying power limits and then coming back, however it power throttled to below 22w again, despite whatever pl1/2 it'd been set to, even after checking sync MMIO and locking it
Also for me, wrt the intensity of load, even if it's below 90, the 22w power throttle happens after some 1-2 mins =/= f(temp, power-draw)

On my Precision 7540 I discovered that increasing the power limits and checking Lock MMIO after returning from sleep the limits are applied.

These higher limits may run all day or may return to default in a matter of minutes depending on the duration and intensity of the load applied to the CPU.

For example, a benchmark that lasts several minutes forcing the CPU to 90°C or more will cause the original limits to return after about 3 minutes.

Also I found something interesting
I got my hands on a 5400 to test, and with the same processor as the 3400, it does power throttle while on battery running a cinebench r23 all core load (aka there is a red colour power showing in the main window), however it hovers around the 25w mark while doing that, netting me a higher cinebench score compared to the 3400 which power throttles to around 15w for the same load on battery, ending up with a lower score on r23 (~600 points) . On charging, it also does throttle like the 7490 (which throttles both on charge and battery to around 22w), to around the 25w mark irrespective of pl1/2, sync and lock MMIO and all
However on the 3400, it power throttles on battery while running that benchmark, but not on charge, same settings as the 5400.
Im confused as all hell
@unclewebb