ATTENTION! Potential Flaw with ALL XPS 15 9560 - GPU Throttling!

Discussion in 'Dell XPS and Studio XPS' started by Pete Light, Mar 17, 2017.

  1. adobepro

    adobepro Notebook Enthusiast

    Reputations:
    0
    Messages:
    30
    Likes Received:
    2
    Trophy Points:
    16
    Sorry, to be clear, I meant the laptop I just ordered from the Dell Outlet, it's not new, it's used and based on posts here, they run hot due to insufficient cooling on the VRMs and GPU VRAM. I have no idea how that proc was used. Not sure if LLC is exposed in the bios and they jacked it up, over volted, etc...but even if they didn't/couldn't, just having turbo enabled with auto voltage set, can cause electron migration over time (degrading the silicon) if the cpu is maxed out the majority of the time in turbo with max voltage + peak voltage with vdroop/voffset peak voltage. Like someone using handbrake encoding lots of videos almost daily, let's say the person who had this laptop was a youtuber/vimoer, when between encodes, the voltage drops, then peaks and the settles at load voltage, over time, that will cause degradation, slowly, but it will, but that will speed up when overclocked/overvolted/LLC too high.

    My understanding of EM comes from:
     
  2. pressing

    pressing Notebook Deity

    Reputations:
    208
    Messages:
    1,574
    Likes Received:
    692
    Trophy Points:
    131
    Highly unlikely a used XPS was undervolted/repasted/repadded. If you disassemble your laptop you can see a repaste/repad pretty clearly.

    Since the vrm throttles at about 76*c (you can see that on one of the ambient temp sensors in HWiNFO64), the CPU and GPU don't usually get the chance to run this laptop very hot in normal use.

    undervolt/repaste/repad allow one to run these laptops at higher temps but that 76*c vrm limit is generally in effect for most users, except the few who disabled some Intel firmware.

    There is a ton of undervolting data in these forums for the 6300HQ and 6700HQ and you can see where your max undervolt compares to those to get some view of your CPUs condition. But each chip varies so I don't think that will tell you too much unless your max stable undervolt is 100mv hahaha
     
  3. Papusan

    Papusan JOKEBOOKS = That sucks! Dont wast your $$ on FILTH

    Reputations:
    14,966
    Messages:
    19,662
    Likes Received:
    30,571
    Trophy Points:
    931
    The risks of overclocking
    [​IMG]
     
    Vasudev and pressing like this.
  4. Vasudev

    Vasudev Notebook Nobel Laureate

    Reputations:
    3,955
    Messages:
    7,061
    Likes Received:
    4,711
    Trophy Points:
    431
    Aren't most BGA not overclockable? Are they saying Turbo boost can result in CPU failure if used extensively for longer duration?
    I have no problem using a cpu/PC for 10 yrs as long as I get microcode and other fixes through UEFI BIOS that don't introduce additional lagorithms to throttle the PC and force users to switch to new PC.
     
  5. MrBuzzkill

    MrBuzzkill Notebook Consultant

    Reputations:
    27
    Messages:
    149
    Likes Received:
    91
    Trophy Points:
    41
    But when a CPU 'degrades' it will simply stop working alltogether. There is not intermittent performance loss. And, for what it is worth, the CPUs performance generally is not worth it long after there may be a chance of failing (averagely speaking of course, CPUs sometimes do randomly fail). I put my old laptop through immense heavy workloads for over 7 years, and in the end, it wasn't the CPU that failed. The CPU after 7 years of rendering and whatnot was still as good as the first day I got it (while keeping the laptop free of dust and repasting every once in a while) in terms of speed / benchmarks.

    In hindsight, I should have replaced the laptop way before that, as the performance jump from a i7-720QM and the current 7700HQ was immense. The pure speed increase in the generational leaps count for way more than trying to prevent any form of degradation.
     
    Vasudev likes this.
  6. Vasudev

    Vasudev Notebook Nobel Laureate

    Reputations:
    3,955
    Messages:
    7,061
    Likes Received:
    4,711
    Trophy Points:
    431
    Maybe @Freitz can share his experience with CPU failing early.
     
  7. Papusan

    Papusan JOKEBOOKS = That sucks! Dont wast your $$ on FILTH

    Reputations:
    14,966
    Messages:
    19,662
    Likes Received:
    30,571
    Trophy Points:
    931
    The low entry models of Skylake/Kaby i7 BGA is the only full locked chips. The others from same line can obtain 2-4 extra bins.

    As long cooling work as normal and Voltage is within Intel’s own specs... None should have problems running them max 24/7.
     
    pressing and Vasudev like this.
  8. adobepro

    adobepro Notebook Enthusiast

    Reputations:
    0
    Messages:
    30
    Likes Received:
    2
    Trophy Points:
    16
    Excellent photo showing that, thanks!
     
  9. adobepro

    adobepro Notebook Enthusiast

    Reputations:
    0
    Messages:
    30
    Likes Received:
    2
    Trophy Points:
    16
    Thanks MrBuzzKill and Pressing,

    Quick question for both of you. I just got my XPS 15 9550 Scratch and Dent from Dell today, there are some minor scratches on the bottom, but no where else, some of the standoff rubber had been worn down too. I don't mind the scratches, as like a new car stressing about with the first scratch, now that it's scratched to begin with, I don't / won't have that stress moment :) Anyways, on to the question: On the keyboard (which I really don't like, it's a joke compared to the Dell M4400 esp. since I paid double+ for the XPS compared to the M4400):

    1. The "B" key is unusually very stiff, like something is pressing against it from the bottom up.
    2 The Backspace key barely moves a fraction of a mm. Literally, there is no motion to the Backspace key, it registers with a space, but no motion to it, it feels like when you try to press it, you're pushing into fabric about 1mm inward, but that's it, no push down/up motion. Is this the issue where some tape is holding onto it that I read somewhere?
    3. The other keys look off...hard to explain, like when you know something is off, but it's difficult to explain what, my impression is that the keys all look out of uniform orientation, like some are fractionally angled left/right.

    Other stuff:
    1. Display is near perfect, no imperfections except for when looking top down, the left side of the panel about 1/2 inch from the left, has a slight purplish tinge. But straight on, it's white (with a white screen)
    2. Will setup throttlestop and push this thing during the return period. Will not let Win10 download any crap to interfere with the stability of the machine (that includes installing candy crush...) If all is well, I'll install Win 8.1/Centos.
    3. Will open it up and check for anything unusual and check if the battery is part of the recall in case if they missed that.
    4. Will consider buying another if all goes well and return this for a certified refurb unit if they wont send me a new keyboard and let me do it myself, just send the KB and I'll return the defective one in the same box (I don't want a tech messing with the machine as getting to the keyboard looks to be a PITA and tech may not be patient and def. wont treat it like it's his own.)
    5. I miss native ethernet. This laptop could've been thicker, no reason not too, esp. to include ethernet/better cooling. This laptop from the side looks like a crappy anorexic ultrabook. If I could, I'd just replace the M4400 panel with this one and call it perfect.
     
  10. adobepro

    adobepro Notebook Enthusiast

    Reputations:
    0
    Messages:
    30
    Likes Received:
    2
    Trophy Points:
    16
    BTW, this was the reference to the red shift panel of the XPS 9550, but it was a 2015 model, newer panels probably fixed the issue. Mine doesn't do this, except for the left side of the panel, when looking top down or top right side looking at the left side.

     
Loading...

Share This Page