Dell on Laptops and Throttling

Dell on Laptops and Throttling

Direct2Dell

The Official Dell Corporate Blog

Dell on Laptops and Throttling

Recently, there's been a lot of speculation about CPU throttling on sites like Engadget, ZDNet and others. This is not about system lockups, freezes, data loss or a design issue. It is about the degree of throttling, which varies under different usage models and ambient conditions.

Throttling is a power management methodology used throughout the industry to balance system performance, component temperature and user experience. Throttling optimizes performance, regulates component temperatures and skin temperature (the amount of heat you feel at external touch points) while using a laptop.

Under normal conditions and use (i.e. a typical office environment and running a typical set of applications), customers won't see any issue at all. At this point, we've only heard from a small number of customers who have reported issues related to throttling. Those issues arose under more extreme thermal and usage models. These customers report more throttling than expected, plus they tend to experience a prolonged recovery time that sometimes requires a reboot to recover from the throttled state. In those scenarios, users may see slower system performance.

What we learned from the customers we've talked to is that we could improve thermal algorithms that dictate throttling thresholds on our mainstream business-class product line. Previous BIOS revisions for some platforms were not optimized for certain extreme operating conditions. That's why we've recently introduced BIOS revisions for the following systems:

  • Latitude E4200 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E4300 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E5400 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E5500 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E6400 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E6500 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E6400 ATG <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Latitude E6400 XFR <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Dell Precision M2400 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Dell Precision M4400 <ADMIN NOTE: Broken link has been removed from this post by Dell>
  • Dell Precision M6400 <ADMIN NOTE: Broken link has been removed from this post by Dell>

One other point I wanted to make is that we have not banned any users from our forum for discussing this issue.

0

To post a comment login or create an account

Comment Reminder

Unrelated comments or requests for service will be unpublished. Please post your technical questions in the Support Forums or for direct assistance contact Dell Customer Service or Dell Technical Support.. All comments must adhere to the Dell Community Terms of Use.

  • I've not tested other systems, but the new BIOS for the E6500 (A18) does not resolve the problem. It provides an improvement, but the system will still throttle under heavy usage at room temperature when docked. The throttling behavior is essentially the same - it only occurs at 4 or 5 degrees C higher than before. This is not acceptable, especially for a system that is marketed as "pure business performance designed for the most demanding applications" (p. 27, October 2009 printed catalog key code 65051). What "component temperature" considerations could there be? The CPU and GPU don't even approach their rated maximum temps before the E6500 cripples itself severely by throttling. What "skin temperature" considerations would there be when the system is docked? I spent about $2000 on a top-of-the-line corporate-class docking desktop replacement and the last thing I would expect (or tolerate) is that Dell would deliberately design this system to cripple itself into utter uselessness (down to less than 5% of processing capacity - even crashing with a bluescreen in the worst circumstances) and, to top it all off, to do so with no indication whatsoever. Dell has no business secretly crippling my system (and thus, my productivity) - I paid $2000 to get 2.2GHz ALL the time, not just some of the time and ESPECIALLY not just some of the time with Dell not telling me when. I'm afraid the incalculable hours of frustration and troubleshooting caused by this problem will unfortunately result in an even further erosion of Dell's global PC market share (once #1, now #3 and falling).

    My original 59-page report outlining this issue in excruciating technical detail (which contains test results valid for E6500 BIOS versions through A16) is at:

    http://www.sigmirror.com/files/44490_iweoz/throttlegate.pdf

    Randall Cotton (AKA tinkerdude)

  • I have Dell 1537 and trying to find out the difference between windows vista and windows 7, which is better? I am also having issues with this thing always wanting to repair almost every time I start it !!!Need help!

  • I do think that Windows should pop up a notification rather than just post an event log when the machine starts throttling.

    Initial testing with A18 on my E6500 is looking okay, but the weather is quite cool at the moment here in Brisbane, Australia.

  • What is throttling??????

  • tinkerdude: Thanks for taking the time to share more of your thoughts. Lots of us here appreciate the kind of time and effort you've already put into this. Would you be open to discussing this in more detail with members from the Dell team?

  • dadiomcallister: Regarding the Windows 7 vs. Vista. I'd have to give Windows 7 a clear advantage. Regarding the problems with your 1537, I'll send you an e-mail to get more information from you to get things moving.  

  • Mr. Menchaca writes:

    "Thanks for taking the time to share more of your thoughts. Lots of us here appreciate the kind of time and effort you've already put into this. Would you be open to discussing this in more detail with members from the Dell team?"


    If you check the record, you'll see I've been in constant contact with Dell on this issue since June 24, 2009. I've talked to roughly 20 different Dell tech support staff, finally getting escalated to a Team Lead at Dell's Resolution Expert Center in Round Rock. I would consider talking directly with Dell Engineering (only because I can only do that now through my current contact). Failing that, I intend to continue following up with my current contact.

  • I'm a bit curious why you did all the tests with a room temperature of 27.5°C-28.5°C?

     

    Not your typical office in any way shape or form...

  • Why has nothing been mentioned about the Studio XPS 1645 and it's related power and throttling problems? All 1645 systems are limited to 931MHz performance while plugged into the wall outlet for power, and can only achieve 1.6GHz performance while running on the battery and unsupported 130-watt adapters. Users have found that exchanging the standard 90-watt power adapter with a 130-watt one fixes the problem, but tech support won't send them out for most. What is Dell's slowness to respond to this? It was mentioned in several of the same articles as the throttlegate issue.

  • what about the sXps1645s???

     

    thats the big issueee

  • I purchased my E6400 in October 2008. All was well until the weather warmed up last spring. Since then my system has slowed to an unusable crawl multiple times per day every day. I did not report this because I could see that people like tinkerdude were on the case and I didn't expect it to take this long to get resolved.

    I flashed A19 on my system as soon as I saw it this morning (I use RSS to watch for updates). Lots of things that I do trigger throttling, but the most reliable is to play a video so I played an 8 minute video after flashing the BIOS today. At the 3 minute mark, the CPU stepped down and CPU use started to ramp up. At 6 minutes CPU hit 100% and my system essentially locked up. It took 5 minutes to stop the video. CPU remained at 100% after that as the CPU was still throttled. It took another several minutes to start RMClock. As soon as RMClock was running, CPU fell to nearly 0%.

    I have been using RMClock to somewhat manage this problem for several months. It is an improvement, but not a solution. Before I found tinkerdude's analysis on the Dell community forum several months ago, I thought I was going crazy. I plan to report this to Dell tomorrow so that they know there are users with this problem who have been silently and patiently waiting for a fix.

    I run docked with two external monitors most of the time. This problem happens even when undocked. I agree with tinkerdude's assessment 100%. This is a serious problem. This is my 4th Dell laptop and it will be my last unless this problem is corrected.

  • http://forum.notebookreview.com/showthread.php?t=437800

     

    Here's a post with details I found when I tried it:

     

    http://forum.notebookreview.com/showpost.php?p=5577220&postcount=11

  • Mr. Menchaca writes:

    "Thanks for taking the time to share more of your thoughts. Lots of us here appreciate the kind of time and effort you've already put into this. Would you be open to discussing this in more detail with members from the Dell team?"


    If you check the record, you'll see I've been in constant contact with Dell on this issue since June 24, 2009. I've talked to roughly 20 different Dell tech support staff, finally getting escalated to a Team Lead at Dell's Resolution Expert Center in Round Rock. I would consider talking directly with Dell Engineering (only because I can only do that now through my current contact). Failing that, I intend to continue following up with my current contact.

  • thedead writes:

    "I'm a bit curious why you did all the tests with a room temperature of 27.5°C-28°C?

    Not your typical office in any way shape or form..."

    So is it the case then that E6500 should only be used in a "typical office environment" or otherwise you should have no expectations of normal operation? People who work at home without air conditioning should not expect their E6500 to function normally? How about people who work on their porch in the summer. Or in the park on a sunny day? Or how about people who use it wherever they feel like when they're not working? Are these all unreasonable environments to use the E6500? It certainly wasn't advertised that way. The tech specs say the E6500's specified operating range tops out at 95 degrees F (see p. 31 at http://support.dell.com/support/edocs/systems/late6500/en/sqrg/pdf/U084C0MR.pdf ). Is it not reasonable to expect that the system shouldn't secretly cripple itself, no matter what you run, so long as you remain within its specified operating range and the system is not somehow physically damaged?

    27.5-28C happens to be the typical summertime temperature of my home office, by preference.

    If you would like to see test results showing severe throttling (for both E6500 BIOS A16 and A18) at 72 degrees Fahrenheit, check out http://en.community.dell.com/forums/p/19247293/19601699.aspx#19601699 . I just happened to be in that environment earlier today when testing the new E6500 BIOSA18.

    But I maintain that this is a legitimate beef even it only happened at 90F or even 95F.

  • Another user with a "troubled" XPS 1645 with the underpowered 90W adapter. :(

    Let's get the ball  rolling, eh, Dell?

    You can make up the huge time delay with the Latitudes by showing loyal customers that you can fix problems efficiently. I'm only on day 25 of the system and was one of the first to get it.


    ~Ibrahim~

  • tinkerdude: Members of our engineering and performance teams are indeed who I was referring to.

  • Lionel:

    I am experiencing the same issues with my E6500 (purchased Fall of 2008) as Tinkderdude and have been trying to get answers from Dell as well. First and foremost, while I can appreciate your need to perform damage control on this issue, I have to disagree with your comment about this not being a problem for typical office users. I am not using my laptop in an "exotic" environment; I am using it in an typical ~70 degree office! I will try the BIOS update provided today, but I'm not convinced that this is a BIOS issue alone. As evidence I submit the following:

    Like Tinkderdue, I have spent a great deal of time trying to diagnose this issue on my laptop. This issue appears to be re-appearing every 4-6 months and I first reported it to Tech Support back in May of 2009 after about a month of my own troubleshooting efforts. Tech Support has twice replaced the motherboard + heat sink on my E6500 "attempting" to fix this issue. What I noticed after each replacement was that my heat sink was blocked by a thin layer of dust and lint... While the layer of dust was thin, it was evenly distributed over the heatsink fins and was VERY CLEARLY blocking most of the airflow. My feeling is that the fins on the heatsink are simply spaced too close together, making them susceptible to dust buildup. Since the fan appears to rarely go into its 'high' speed, the end result is that as more and more dust buildup occurs, the processor runs at warmer and warmer average temperatures (even in office environments!), and eventually, BIOS thermal throttling kicks in, rendering the laptop useless.

    To be clear, I am in no dustier an environment than an average user. I work in an typical (non-industrial) office environment, and also use my laptop from home. Here is a link to an image of my heatsink which I took for Dell Tech Support for the May '09 case (Honestly, I don't recall if I ever sent the image to them or not, but I'm pretty sure we talked about it).

    [IMG]http://i48.tinypic.com/2luwduv.jpg[/IMG]

    The dust build-up on my most recent replacement was even worse than the image above. Obviously it would take time for dust to build up on the heatsink fins, and so it makes sense that this problem took a few months after original shipment and heat-sink replacement to re-appear.

    I hope Dell continues to investigate this issue. What I have now is a not a high performance laptop, but a $3000 heated paper weight.... And a hot one at that.

    -James

  • If Dell Engineering (that is, the group at Dell directly responsible for providing technical fixes including BIOS fixes) would like to contact me directly, that's fine. My e-mail address is published  [admin note: e-mail removed for privacy].

    But you all should have had my contact info, including my phone number, since June 24. 2009 when I first called Dell regarding this problem. And Engineering has had my 59-page report for months. They've seen my emails on this. They could have contacted me at any time. And they still can.

    And you know what, I really kind of resent that you write as if I hadn't been talking to Dell all along and was just shooting from the hip on this or something. I talked to well over a dozen tech support staff from Tampa to El Salvador to India before this was even given any real attention. That was back in July and August. And now you're asking me if I would kindly be willing to talk to Engineering as if I haven't been trying to do that myself for the last 5 months. I'm drowning in irony.

    So yeah, people at Dell who can actually fix this problem themselves (but apparently won't for whatever reason) are welcome to email or call me. And I will hopefully be in a better mood about this than I am now (sorry).

    Randall Cotton (AKA tinkerdude)

  • we dont even know what it is yet

     

    if its entirely the lack of power, some say the bios makes the mobo cap out at 90 and doesnt allow it to draw anymore

     

    sooo annoying

     

    we want an official statement about the 1645's!

     

     

  • hey lionel what about the xps 1645s maaan

  • w0g and Ryan: I've asked the team to review the details on the Studio XPS 16 as well. When I have more details to share, I will.

  • Hello, finally some official words on this. Please understand that many people including myself think, that this is just taking too long now. The thermal table update might solve it for some - but the problem lies here:

    The design of the cooling is quite poor. The CPU heats up the chipset/GPU and the heat of the GPU/Chipset must pass the CPU using the heat pipe. The fan is constantly running when docked - even when IDLE or only office work. After a couple of weeks, the fins where the fan blows through are covered with dust and need treatment. 

    Hopefully, the next generation of Latitudes will have a better cooling design.

    Beside this, it's a great laptop!JB

  • We're all waiting for an official statement on the sXPS 1645, I didn't shell out 2,300 for a stylish paperweight.  I bought a gaming machine, for gaming.  Which I cannot do.  It would be nice for something to come from dell other than "we're looking into it".

  • I ordered a Studio XPS™16 (N00X6003) a few days ago. Now I have read severla times of the heating problems!!!! Is the problem solved meanwhile? ... The technical chat support agreed with that but i do not trust them! So is the problem really solved completely? Maybe i should cancel my order...

    Regards, Michael  

  • I've experienced these exact throttling issues on my XPS M1330 too. The CPU is pegged to 100% but when I look at the Resource Monitor the CPU frequency drops 10%. Unfortunately disabling speedstep in the BIOS doesn't fix the problem. Will you be releasing a BIOS update for that model?