I

Discussion in 'Hardware' started by kurt corbin, Jun 4, 2019.

  1. kurt corbin

    kurt corbin Scribbler - Standard Member

    Messages:
    175
    Likes Received:
    55
    Trophy Points:
    41
    .
     
    Last edited: May 24, 2020
  2. Shogmaster

    Shogmaster Pen Pro - Senior Member Senior Member

    Messages:
    4,202
    Likes Received:
    1,466
    Trophy Points:
    231
    "Best chassis" would allow some room for M.2 slots...
     
    bloodycape likes this.
  3. desertlap

    desertlap Scribbler - Standard Member Senior Member

    Messages:
    2,069
    Likes Received:
    2,488
    Trophy Points:
    181
    "To me, it really looks like Dell are following the very worst plays from Apple's book: prioritizing quiet fans over lower temperatures and performance (XPS 13 9370),"

    There's a disconnect for consumers and a dilemma for manufacturers. MS for instance came to the conclusion after the surface pro 3, that consumers would rather have quiet devices over the absolute fastest ones. That is implemented very specifically in the current line of the Book 2, Go and Surface laptops.

    Apple has told us similar things, e.g. that in the past noise and heat were the most commonly cited reasons for consumer returns.

    The other thing we all need to remind ourselves here is by the self selecting nature of the type of user that posts here, we don't represent the average or majority of customers.

    More directly on topic, one thing that we have preliminarily discovered with our engineering sample is that graphics performance takes a big leap with ice lake, to the point where low end separate graphics cards make less sense.
     
  4. Steve S

    Steve S Pen Pro - Senior Member Super Moderator

    Messages:
    7,869
    Likes Received:
    3,368
    Trophy Points:
    331
  5. Steve S

    Steve S Pen Pro - Senior Member Super Moderator

    Messages:
    7,869
    Likes Received:
    3,368
    Trophy Points:
    331
    <<...Iris Plus outperforms MX230 due to thermal contraints…>>

    In that particular laptop design. Improvements to the thermal design might very well allow better discrete GPU performance...
     
  6. Steve S

    Steve S Pen Pro - Senior Member Super Moderator

    Messages:
    7,869
    Likes Received:
    3,368
    Trophy Points:
    331
    <<...Does anybody understand the TDP of the i-7 1065G7 ?..>>

    I have some thoughts on this that I'll be happy to share later today, but no time for a long posting now. I believe that your first comment (referencing the Vantage app) is the key. I've been spending the last few days digging into the BIOS on my ZBook X2, and, at least in HP's case, that seems to be where the X2's TDP is controlled...
     
  7. desertlap

    desertlap Scribbler - Standard Member Senior Member

    Messages:
    2,069
    Likes Received:
    2,488
    Trophy Points:
    181
    We are set for a deep dive in to both Ice lake and Comet lake late next month and I'm really interested what Intel will share with us.

    Purely anecdotally and based both on the very limited number of ice lake systems we've come across and other information floating around out there, but the 10th gen chips are different in some fundamental ways.

    We've observed for instance that the 10th gen is both quicker to Turbo up than 8th gen chips but also quicker to clock back down. I suspect that has a lot to do with thermal management which sort of reached it's apexes challenging with 8th gen chips and the thin and light systems which in the worst cases such as the HP Spectre Folio caused the processor to pogo between high and lower power state later firmware essentially just down clocked the entire system.

    As to why the variability or that someone like Razer can do what they do, I think the answer lies with Razer itself . They are possibly the only somewhat mainstream company that really tests their systems with the tools they release (not to mention the copious disclaimers about what you are risking when you push them).

    For the other companies it allows for greater flexibility in their system designs while allowing a core technology set. And because those companies don't test like Razer does, not making the tools available to the general public avoids a bunch of bricked systems.

    In the longest term I think this will have two impacts, one is that making a blanket assumption about basic performance based purely on the chip in the system will be less reliable than ever and thus those that are enthusiasts will have to examine reviews and tests even closer.

    The other thing this sets the stage for is lakefeild which will upend a lot of conventional wisdom about intel chipsets out of necessity for good or bad.

    My $.02 anyway

    EDIT: For clarity , the chip specifically in the Spectre Folio is a 7th gen chip but the issues still apply
     
    Last edited: Jan 21, 2020
  8. Steve S

    Steve S Pen Pro - Senior Member Super Moderator

    Messages:
    7,869
    Likes Received:
    3,368
    Trophy Points:
    331
    <<...Does anybody understand the TDP of the i-7 1065G7 ?..>>
    <<...I have some thoughts on this...>>

    I've had some similar questions in my mind about the 8th Gen i7-8650U processor in my HP ZBook X2:

    https://ark.intel.com/content/www/u...-8650u-processor-8m-cache-up-to-4-20-ghz.html

    The 8650 is rated for TDPs between 25 W and 10 W and, at least in terms of PassMark scores, appears to be just slightly superior to a 47 W Haswell i7-4770HQ.

    In a discussion thread elsewhere in the forum, another X2 owner mentioned that his X2 seemed to be underperforming on a benchmark test but that in the X2's BIOS, there was a setting to increase performance when on battery power. He subsequently commented that when the setting was enabled, the processor temperature shot up to 99° C, which is essentially the maximum on-die temperature permitted (100° C). So his X2 appeared to be temperature limiting.

    Those comments caused me to do some research on HP Sure Start, which is HP's combined security and BIOS application which controls start up and underlies the OS. Sure Start is one of the most complicated apps of its type that I have ever encountered, but I was fortunate to find that HP offers extensive documentation of the X2 and its BIOS. To cut to the chase, the BIOS contains several switches that control the processor's performance. Hyperthreading, multi-core operation and turbo mode can all be switched on and off, and turbo mode can be disabled when on battery power ( ! ). Taking a step backward, the factory default settings throughout the BIOS strongly suggested that HP's design team tried to balance processor power with battery operating time, but with a tilt towards battery operating time. Given the nominal power consumption of the Adobe RGB-capable display, CPU and discrete GPU, that tilt is probably understandable, but it does leave the impression that the 8650 is hobbled when not on the AC adapter.

    All of this is just a long way of explaining that, at least for the X2, the TDP of the 8650 appears to be dictated by the settings in the BIOS but also mitigated by the cooling. Once the on-die temperature hits the 100° C limit, the processor is throttled (or shut down). If the cooling is not optimal, which it might not be in the form factor of the X2, the processor will be throttled before it can reach or maintain a TDP of 25 W. I suspect that applications of the i7-1065G7 and its TDP will be structured in a similar way and will depend upon what options the BIOS offers...
     

Share This Page