Help needed from new XPS owners: Report serious BIOS bug to Dell

I am not sure if this is a bug that also applies to the new XPS 13 but I think XPS 13 users should check to see if they have it. I would strongly suggest users here spend the time to check for this since it is in our best interests to get Dell to fix this if it is a real issue. I am confident it is a problem with the XPS 15.

Essentially the problem has to do with the CPU frequency getting capped at a much lower speed than what the processor is actually rated for. This happens every time on my XPS 15 when you unplug your AC connector and re-plug it. This will greatly reduce the speed of your computer and reduce your user experience.

To see if you have it, do the following:
Connect your laptop to the power and reboot. I have found that if you start out connected to the power and you reboot, everything should work properly.

Set your power policy to "High performance" and/or run your favorite benchmark like Geekbench, CPU-Z, etc
This is just meant to stress the CPU to get the frequency up. You can also use the benchmark to get an idea of your current performance. While this is going on, open up the task manager and look at the reported CPU speed. On my machine this tends to be about 3.3-3.5 GHz.

Now unplug the power after a few moments you should see the CPU frequency level out at some level that is lower than what you previously saw (about 2.57 Ghz on my machine). This is to be expected as the machine wants to conserve battery life.

Now plug your system back into the power, now you would expect things to return to what you originally saw but instead the CPU will get stuck at some much lower frequency and won't be able to return to normal unless you sleep and wake up the system or reboot it all together. I find in this situation my CPU is stuck at 1.67 GHz, dramatically lower than where it should be.

This is an unacceptable bug with the Dell BIOS/Firmware since we are paying money to get a high end Intel CPU but we are getting the performance of a much lower processor. If other users can confirm this, I think we need to report it to Dell. Hopefully they already know.

Evidence of the fact that the firmware is limiting the CPU EVEN while on AC power can also be found in the event viewer in Windows:
"The speed of processor 6 in group 0 is being limited by system firmware. The processor has been in this reduced performance state for 70 seconds since the last report."

 

Similar BIOS issues exist with the Latitude E7440 laptops in my experience. It isn't a constant and pressing issue in that situation; the machine must be heavily stressed and get quite hot normally for the permathrottle to be introduced. However the issue has never been completely resolved in the instances where I've encountered it. Some of my clients received board replacements from Dell and it fixed their problem.

Hopefully in this situation a board replacement will not be necessary, and perhaps a mere BIOS update would suffice.

Can you run ThrottleStop and force it back to normal Turbo frequencies?

 

Yep ThrottleStop can be used to bring the CPU back up to any supported frequency but it no longer will scale with load like it usually does, so this would kill your battery pretty quickly at high frequency.

FYI all you have to do to induce this "perma throttle" is to unplug the power and plug it back in, it is very easy to induce.

I am almost certain that this is simply a bug in the BIOS and can be fixed with a bios update, my main concern is whether or not Dell knows about it. Dell potentially won't do anything about it unless we users complain about it and this is the kind of problem that less technical and less attentive users may not even realize exists. I have had a very similar problem in a previous HP laptop and HP never issued a single bios update for that laptop, I really don't want Dell to do that. If Dell does not acknowledge the problem soon, I might consider returning this laptop even though I love everything else about it just because I don't think we should have to put up with such problems after paying such a large amount for a premium product.

 

Was able to successfully reproduce the error, don't know why this isn't getting any attention. Where/who should I file a report to to make sure this gets the most attention?

 

Yeah I've been complaining about this in the other thread but not sure why most users aren't taking it seriously. We are all paying for these expensive Intel i5s and i7s and are sometimes getting sub-i3 performance, it is unacceptable for Dell to leave things in this state but I'm worried they will ignore it unless some people complain about it. I also noticed that if you repeatedly unplug the system and plug it back in all while keeping the CPU loaded, you can induce some system lock ups (which seem to last about 10+ seconds and then eventually recovers). I have already tested these problems on two separate XPS 15s and am fairly confident that the source of the problems is indeed the BIOS, which needs an update.

There is obviously some work to be done as far as fixing the BIOS and Dell has released it in an incomplete state. I am trying to figure out what the best way to inform Dell is, but I'm not sure. I think talking to their tech support would be a waste of time since they are unlikely to understand this very well. Anyone have suggestions?

 

I will test that asap, by the way, what is your BIOS version ?
Have you install the v1.0.3 from here ?

Have you tried another OS ?

 

I will mention this to my colleagues at Notebookcheck to try and ensure it receives attention/exposure if it still exists.

 

Are you using the Dell Power Companion? I remember there being a similar issue on the XPS 13, but this only occurred when using the Power Companion.

 

Yes I have flashed that BIOS from the dell support page, it is actually an older bios than what shipped with the laptop. The actual timestamp on this BIOS (not the date listed on the website) is 9/28/2015. Basically I ended up downgrading and I have no way to upgrade until Dell posts the latest BIOS update on their website.

Unfortunately the issue exists in both BIOS versions. I have not tried another OS.

 

No, I am not using it. Just a direct AC connection to the wall using the provided power adapter.

Much appreciated, thank you.

 

FYI, I was able to contact Dell support and informed them of this issue, they have told me they will convey the information to their engineering team to look into. I would advice others to do this along with maybe posting about this on the official Dell support forums.

 

I just posted a complaint on Dell's forums and will do the same here to up the count of affected cases. Running 1.0.0.5 on xps 15-9550. Restarting the computer fixes the problem temporarily until the next time that I unplug/plug the power adapter.

 

Checked, indeed...
Before unplugging


after unplugging


But I don't know if is not the chipset that is not reporting the frequency correctly, as I tested while playing GTA V, and didn't notice an FPS drop. Ok maybe it is because CPU doesn't influence that much on FPS, a cpu benchmark may be more accurate.
The fan run a lot quieter, that said.

 

It is actually an accurate reading of CPU frequency. I used the CPU-Z benchmark and found an approximately 50% drop in CPU performance! You can use any tool that will benchmark CPU performance and you will find that we are indeed losing half our CPU capacity.

 

My 9550 arrived today and noticed this bug straight away. Between this and the worlds worst laptop speakers which crackle when I scroll in web pages and just sound plain awful, I am tempted to send it back

 

i notice the problem too. i have confirmed it on the thread on the dell forum that you posted. hopefully dell will look at it more seriously. i am on the latest firmware/bios posted today 1.0.0.7...
the only thing is,. when the power is removed, it drops to 17x, and the climbs to 26x on battery................ then when powered back on AC - drops to 17x now.

 

I guess the speakers don't bother me too much. I consider all laptop speakers to be pretty worthless, I find these to be generally passable.

 

Thanks for posting on the Dell thread, hopefully Dell will fix it soon. It is disappointing that they have claimed to not be able to reproduce it even though so many others have confirmed the bug now. It isn't exactly hard to reproduce it.

I too am on BIOS 1.0.7 and it did not fix the issue of course. I wish Dell would at least provide changelogs to the BIOS updates so we can see what is being fixed.

 

I agree that UEFI/BIOS firmware is faulty. On a XPS 13 9333 my BIOS start up time was 1.5 seconds. Typically I was waiting 13 seconds for the BIOS to load. You can check this in task manager and the start-up tab. Otherwise I'm enjoying this computer.

 

I was NOT able to reproduce the bug on UEFI Ubuntu 15.10 kernel 4.2.0-18 with BIOS 01.00.05 (factory).
While doing several plugging / unplugging, did this:

while displaying clock speed with

It yield (when on battery and on AC):

and

So cpu frequency was continuously at 3.1GHz while stressing, whether I was on battery or on AC (just a short downclock to 800MHz when plugging AC) and when I stopped the stress test, clock went down between 800-2600MHz.

 

Annihil, Your results look like what happens when you disable SpeedStep in the BIOS. Did you happen to do that? When on battery, your max CPU should be about 2.6GHz.

 

Didn' t change anything in the BIOS except RAID On to AHCI.
So my BIOS perfomance settings are:
MultiCore Support: All
Intel℗ SpeedStep™: Enable
C-State Control: Enable
Intel℗ TurboBoost™: Enable
HyperThread control: Enable

I don't know why clock is capped @ 3.1GHz instead of 3.3-3.5GHz on AC, and 2.6GHz on battery though.

 

I can answer the first part of this. If you Google around, you'll find that the i7 has different turbo speeds depending on the number of cores that are active. You get 3.5 GHz with 1 core active, and 3.1 GHz with 4 cores active.
Point being that if your primary application is not multi-threaded, then you would do well to set "MultiCore" to 1 (or less than 4 anyway).

 

Just found another BIOS bug, when I disable Thunderbolt the laptop will either boot to a black screen or a corrupted screen. The Bios screen only shows about an inch of the screen if I hit F2.

Fished around the screen blindly and reset the BIOS to defaults and laptop boots normally. Did this three times to verify that it was disabling thunderbolt that was the cause.

The laptop takes way too long to boot, something has to be off in the BIOS.

 

Just replicated this issue. Am calling dell to report it. Where can I find the thread on Dell to post this issue?

 

It's here: http://en.community.dell.com/support-forums/laptop/f/3518/t/19660225?pi239031352=2

This worries me:

Can it really be this difficult for their engineering department to obtain a release version of the laptop....?

 

Aren't us who have the XPS 15 pre-release engineering version and them who have the XPS 15 release version ?

 

Touché.

 

So does anyone with the 6300HQ experience this too?

 

Yep. It seems to always happen after the laptop goes into hibernation mode, but when it's in sleep mode the clocks sometimes return to normal.

 

Same problem here. Top end model (i7, 256gb ssd, 4k). On battery, the max frequency is capped at 2.6 ghz, if I plug in the ac adapter it limits it to 1.69 ghz. If I reboot with ac plugged in, then it goes up to 3.3 ghz or so. Called Dell and asked them to report this to their engineering department and follow up with me. Everyone that has this problem should be doing the same thing, this won't get fixed if we don't make some noise.

 

My 14 day return period expires from BestBuy expires in two days. Today, my spacebar has started creaking. CPU throttling bug still not fixed. Fans run at odd times. Keyboard backlight times out so fast it's annoying - and timing can't be adjusted. Tempting to dump it while the dumping is good.
Problem is that I can't find anything else that comes close in performance to a well-optioned 9550 with an i7/nVidia and large SSD right now. Even the Macbook Pro is stuck on Broadwell.

 

The HP ZBook Studio is looking like a potential XPS 15 alternative. Not sure about price (HP is usually higher) since not fully released yet. Better ports, less battery, bigger (not infinity-style screen), Quadro instead of GeForce (so more comparable to the Precision 5510), Iris Pro CPU option in 2016, and maybe better QC (could it be worse?). I'm waiting to see how both are reviewed by owners.

 

I think there is actually a fix for the backlight: https://www.reddit.com/r/Dell/comments/3slh3t/xps_13_9530_xps_15_9550_keyboard_backlight/?

Spacebar sucks on mine too.. getting a replacement though.

Could always just return it and then purchase another one if/when the bugs get sorted out.

 

No news?

I have the 9350 i5 / FHD / 256Go version, I'll try it tonight and give you my feedback.

 

I believe we are still waiting on Dell to update us...

It is kind of shocking how incompetent this so called engineering team is that they couldn't initially reproduce the problem even after so many of us described it in great detail. It isn't exactly rocket science to find this bug. Hoping they will have the fix soon.

 

I reproduced what you did and both on battery or pluged in, i'm stuck at 2,66Ghz. So it's fine to me!

 

Just an FYI all. Dell has acknowledged the problem and are now working towards a fix.

More info at:

http://en.community.dell.com/support-forums/laptop/f/3518/t/19660225?pi239031352=4

 

You did something wrong if you are stuck @2.66GHz on AC

Good news.
But as I said, I don't know if it is the BIOS fault or Windows fault, as this bug does not affect Ubuntu, so...
It can be a Windows power management issue, as power management can change core multiplier (for example, high performance will put all cores to max multiplier, so power management can indeed interfere with core multiplier).

 

It is possible that Windows might be at fault, but I doubt it. This isn't an issue that is occurring on other Skylake Windows machines as far as I know. Maybe the CPU governor in Ubuntu is forcing different CPU multipliers and ignores the frequency that the firmware is trying to enforce? Using ThrottleStop in Windows (the beta version for Skylake) I too am able to force any supported multiplier even when the bug kicks in. I was able to get back to 3Ghz+ using this after being stuck at 1.69 but I believe there is still a bug at the BIOS level, this is just a way of masking it.

Starting with Windows build 10586, Skylake systems all support Intel SpeedShift, which allows the CPU to regulate it's own voltage and frequency. In order for it to work, the BIOS must support it. I wonder if the Dell one supports it? It should lead to improvements in system responsiveness, performance and battery life (small improvement).

 

The fault is most likely in BIOS.

I've had the exact same issue with my old Dell Studio 15 laptop and it has never been resolved to this date.
This is after several Windows installations (7, 8, 8.1 and now 10).

 

Maybe, as the multiplier can reach x31 on battery and should be x26, but at least the multiplier can go from x8 to x31-x34 as you can see:

 

This has nothing to do with Windows.

I just hope this is sorted out before mine ships. I have bought laptops and have been left hanging with a major bug in the past from other manufacturers and don't want to cross Dell off my list either :/

 

You can get around the issue by hibernating your laptop instead of standard sleep.
It will take longer to resume, sure, but beats having to reboot every time.

 

Will keep that in mind for sure till it gets a perm fix, thanks!

 

End of this page.

Looking forward to BIOS 01.01.11 !

PS: When this BIOS will be released, please post a message here to notify everyone, thanks

 

For the record: I received today the "small" 45w power adapter. With that adapter my XPS (i7 skylake, bios 1.07, 16Gb, 1tb m2 ssd/AHCI with samsung driver) does nog cap max speed of the CPU.

 

Quote (Haris163) I was just able to update the BIOS to 01.01.15 using the update driver software option from the device manager.

Quote: earlier post of me in an other thread:
Nice find, thanks for the update.

I noticed 4 firmware update:
1) system bios update (I think that was the text, since I started writing things down after I noticed their were 4 updates);
2) Intel ME firmware update;
3) Main TI PC 0
4) EC Bios

Anyone knows what 1..4 means??

I can confirm speedcapping issue is gone (i7, 4k, 1tb ssd, 16gb).
Processor SST is (still) not available.
TPM still disabled.

Geekbench gives same results running on battery and AC, both ways peaks multicore at 3,3 Ghz and 3,5 Ghz singlecore.

Crystaldiskmark score (PM951, AHCI, Samsung NVMe drivers, software bitlocker) are same as before fw update.

 

Please report this to Dell for the latest 1.1.15 bios bugs.

1. Intel Speed Shift Technology (SST) is still not enabled.

2.Failure to load back into Windows after doing a soft system restart. Hangs at the Dell logo. This is serious bug, for example after you do updates, it will hang, fail to finish and may compromise system integrity.

 

I tried restarting a few times, to test if it happens on my machine (connected to AC and just on battery). The restart bug doesn't happen with my XPS 15 (i7/512/16GB/FHD). Does it always happen for you?