Gaming on a Laptop while on Battery

Discussion in 'What Notebook Should I Buy?' started by PJPeter, Oct 11, 2018.

  1. MorphineX

    MorphineX Notebook Enthusiast

    Reputations:
    5
    Messages:
    19
    Likes Received:
    5
    Trophy Points:
    6
    Peter,

    Thanks for the follow up reply. I've since returned Alienware 51m in favor of m15 variant with i7 and GTX 2080 MQ. I'm sure I messed something up because notebookcheck showed alienware 51m with GTX2080 performing impressively with Firestrike while on battery.

    I'm not much of a benchmarker / gamer - just trying to get a good laptop :) But once i get the Alienware m15, I'll run some test and post my results!
     
    PJPeter and etern4l like this.
  2. etern4l

    etern4l Notebook Deity

    Reputations:
    251
    Messages:
    1,197
    Likes Received:
    457
    Trophy Points:
    101
    With Alienware m15 it's very simple on battery - you get 40W system power limit, so just over 2h of gaming with very reasonable performance. If you're playing a less graphics intensive game then switching on Windows Power Saver mode limits total system power to 20W, ergo over 4h of gaming time. The CPU undervolts very well. Plus you get on-site support unlike pretty much all the other brands.
     
  3. PJPeter

    PJPeter Notebook Deity

    Reputations:
    110
    Messages:
    1,122
    Likes Received:
    35
    Trophy Points:
    66
    Thanks for sharing that, looking forward to your results. Hope you don't stress over this too much though - if you're happy with the laptop and it does what you actually need to do, then that's all your really need and you can save some time and money in the long run - but up to you and I'm happy to share my knowledge and ideas.

    The only thing I could find for the m15 is for the last gen GTX 1070 MaxQ which is designed to work with about the same amount of power as the RTX2080 Max Q (80/90W):

    Running on battery power will impact CPU and GPU performance. A 3DMark 11 run on batteries returns Physics and Graphics scores of 9680 points and 6859 points, respectively, compared to 12273 points and 20573 points when on mains.

    It seems like from that review it should give you 33% of plugged in performance while on battery without any tweaks - which is disappointing. But undervolting and other things can be tried, and the RTX280MaxQ version might be better for this anyway.

    And there may be other hacks you can do to improve that as I posted before. And I don't know what battery level they did these at - as the voltage/amperage available drops, so does the max available power.

    I'm currently experimenting with my Razer Blade Base 1060 Max Q in maximizing battery life when working and not gaming - seeing if I can limit the noise when running heavy workloads (like encoding) while plugged in or on battery (so I can choose longer battery and quiet over performance if the situation warrants, which it did yesterday and does often). Since I removed the Synapse Software because it was causing high CPU usage and didn't work, I don't have other fan control. Using Throttle Stop to disable turbo and other tweaks I look to be getting well over 7 hours at 70% brightness on non battery saver mode.

    I like that I can have an almost all day battery when I just need to work, but can also game at almost 90% performance when I don't - I really value the versatility this laptop is giving me - quite grateful for it.

    All the best,
    Peter

    P.S. One more result to share - for a Razer 15 Advanced i7-9750H RTX2080 MaxQ 240Hz:

    Running on battery power will still reduce overall performance. A 3DMark Fire Strike run on batteries returns Physics and Graphics scores of 3535 and 11292 points, respectively, compared to 17038 and 20718 points when on mains.

    That translates to providing 54% of plugged in performance while gaming on battery.
     
    Last edited: May 23, 2019
    MorphineX likes this.
  4. etern4l

    etern4l Notebook Deity

    Reputations:
    251
    Messages:
    1,197
    Likes Received:
    457
    Trophy Points:
    101
    This is a ridiculous statement. If the laptop drew enough power to give 90% performance, battery life would be 40 min tops which is hardly practical for anyone.

    Err, performance on battery is 20% for physics and 54% for graphics, so it we weighted them equally as the benchmark probably does, it would be around 35%.

    My m15 with i7-8750H and 2070 Max-Q running in Better Battery mode manages 2500 in TimeSpy vs 7150 (or 7550 OCed) plugged in.
    A Blade 15 with the same CPU and GPU models managed just 6367 on AC power: https://www.notebookcheck.net/Razer...TX-2070-Max-Q-FHD-Laptop-Review.402631.0.html

    I can see why Razer users prefer to talk about percentages of plugged in performance, however, if we compare my optimized battery performance to Blade 15 plugged in performance, we get 39% :)

    I have not benchmarked in other battery modes - would likely be possible to squeeze out even more performance at the expense of battery life, however, I feel that just over 2h of battery life is where I would prefer to stay when running intensive tasks therefore I've been optimising in that direction.

    BTW the m15 review also states that the m15 is significantly more power efficient than earlier AW models and the Blade, just saying.

    For completeness, compared with the Blade, the m15 also benefits from an additional m.2 SSD slot, the superior Alienware Graphics Amplifier port, and on-site support.
     
    Last edited: May 23, 2019
  5. PJPeter

    PJPeter Notebook Deity

    Reputations:
    110
    Messages:
    1,122
    Likes Received:
    35
    Trophy Points:
    66
    Wow, I've never experienced (basically) being called a 'Razer fanboy' before. I actually have owned a lot of other laptops and I tested each and I'm reporting against them in a neutral fashion. I came very close to returning the Razer Blade Base for a ASUS GA502 w/GTX1660 but unfortunately it was too power limited when testing on battery for my use (see previous pages).

    "I like that I can have an almost all day battery when I just need to work, but can also game at almost 90% performance when I don't - I really value the versatility this laptop is giving me - quite grateful for it."​

    I reported my findings accurately, and it is giving 89% of the plugged in performance, and it lasts about 70 minutes doing so.

    When I intentionally throttle my machine in order to achieve longer battery life and reduced noise, then I can get almost all work day battery (7.5 hours).

    But I do not do it at the same time. I like being able to do it with the same machine though, especially since the in built Razer software was garbage, would freeze, caused the fans to kick up randomly and kept using up all my CPU (as I stated before).

    I have a Razer Blade Base 2018. It has a GTX 1060 MaxQ and both an m2 and SATA slot, with a smaller 65W 4 cell battery. I don't care about the brand. I care about being able to use the laptop for what I need. And when it comes to gaming, graphics is more the limiting factor. (I usually compared both scores, but I've been more focused on the graphics score in recent posts - I'll try to be more clear about that going forward, my apologies if I caused confusion)

    If you want to say your laptop or brand is better, I don't care. I really don't. Life is too short to have such an attitude. (Unless maybe you have investments in it or work for them?) My laptop is a tool that I can use for work and fun and eventually it'll wear down and I'll replace it with whatever meets my needs and budget best then (I've owned Dell, ASUS, Gigabyte, Lenovo and now Razer and others and I purchased many different brands as part of my IT job).

    My laptop on battery w/NVIDIA GeForce GTX 1060 Max-Q(1x) and Intel Core i7-8750H Processor returns a TimeSpy graphics score of 3110 vs 2500 on your self reported m15 i7-8750H and RTX2070 Max-Q (would love if you can share the actual comparison link btw).

    I don't care if it's a Razer, Dell (Alienware) or whatever - you do not need to defend yourself or your pride and joy. I don't care - I'm just reporting my experience - and I gave full battery life numbers, what % I tested against, etc.. earlier in the thread for all to see - including when I had a Lenovo laptop - and against a Razer Blade model which I said was inferior in this way and definitely not worth the price. If I hadn't gotten this Blade Base on clearance ($1699 CAD or approx $1260 USD) I wouldn't have bought it - Razers at full price are way too expensive for me.

    Please read the thread and I'd love if you want to contribute your findings further - I have no ill will towards you, I just want to help learn and share on this topic. :)

    All the best,
    Peter

    P.S. I think you took offense because you looked at that one post - the reason I did a P.S. and included the details of the battery performance for the new Blade w/RTX2080 Max Q (similar to the GPU in the laptop MorphineX ordered) is that it was just posted on May 21 by Notebookcheck (only a few hours before my post) so it was additional new data to all the previous posts I made about their Battery performance results. I wanted to find some for Acer, MSI and others to confirm if there is an association between OEM but there isn't enough data yet to determine so (some reviews they include them, some don't - and never with a battery % included, which can have a huge effect on performance as my testing on the Y740 showed). And the new insight I found about workarounds via the system settings adds more potential avenues of investigation as well. That's why it's great to keep collecting more data, thanks for contributing and please feel free to check out the entire thread for more context.
     
    Last edited: May 23, 2019
    4W4K3 likes this.
  6. etern4l

    etern4l Notebook Deity

    Reputations:
    251
    Messages:
    1,197
    Likes Received:
    457
    Trophy Points:
    101
    The reason for my initial reaction was that for a proper dGPU laptop (which I incorrectly assumed you were talking about), the 95% of performance for over an hour number would defy basic physics and maths. Let's take a look at my m15. On stock voltages, it will have a max power draw of about 200W (90W for the CPU, and 100W for the GPU, 10W for the remaining components). In a benchmark such as TimeSpy, the draw will be probably closer to 145W so let's take that number. Assuming linear relationship between power and performance, which is good enough for the purpose of this very rough conversation, it would need about 130W of power to deliver 90% performance, resulting in about 40 min of a 90WHr battery life. I would expect the numbers to be in a similar ballpark for all modern BGA gaming laptops.

    Having scanned back and I have found your "95% of performance on battery" claim is referring to a Surface Pro with integrated graphics :D Fair enough. I'm sure any modern laptop will achieve "performance ratio" in that ballpark using iGPU. This whole notion of performance ratio is nonsense. What matters is absolute performance on battery as well as the amount of time the laptop battery can maintain that performance. Yes, Surface Pro may have a better ratio but it doesn't matter since dGPU laptops will vastly outperform it on battery. The slower and less power hungry the laptop is that smaller delta between plugged in and battery performance, obviously.


    Assuming you mean TimeSpy score link? OK, I might re-run it later - this time on max performance settings (the 2500 score was in balanced mode and pretty aggressive battery saving settings, where the max power draw of the entire notebook was 40W (not 90W as in your 3300 combined score), affording me around 2h of battery life..
     
    Last edited: May 23, 2019
  7. ariyanbeton

    ariyanbeton Newbie

    Reputations:
    0
    Messages:
    1
    Likes Received:
    1
    Trophy Points:
    6
    The battery on the Aero 15X is 15.2V * 6.2Ah = 94.24 Wh. So the best potential for this - assuming 15.2V as nominal/close to average voltage, is itکشمش پلوییcould give a little more than the 91W minimum for about 1 hour. However the battery can't maintain that load continuously, so the OEM has set up various power restrictions/throttling to prevent it from dying so quickly.
     
    etern4l likes this.
  8. etern4l

    etern4l Notebook Deity

    Reputations:
    251
    Messages:
    1,197
    Likes Received:
    457
    Trophy Points:
    101
    That's an excellent point. Even if it's possible to discharge the battery fast, it's not a good idea from the battery health perspective.
     
  9. MorphineX

    MorphineX Notebook Enthusiast

    Reputations:
    5
    Messages:
    19
    Likes Received:
    5
    Trophy Points:
    6
    I need to figure out what to tweak. I am currently watching 3d mark Fire Strike benchmarking with 11 FPS while on battery. GeForce Experience, turned off whisper mode, turned off battery boost. Nvidia control panel - select for the RTX 2080 instead of the onboard intel. Windows 10, everything to max performance on battery. Fan control - it's running at high.

    It just finished... overall score 3413 v1.1.

    Graphic score 3656
    Physics sccore 14668
     
  10. MorphineX

    MorphineX Notebook Enthusiast

    Reputations:
    5
    Messages:
    19
    Likes Received:
    5
    Trophy Points:
    6
    Believe it or not - - Any RTX 20xx series will perform worse than GTX10xx equivalent series while on battery. It's as if soon as a RTX video card senses there's a limit to the power source, it shrinks up and performs as poorly as an integrated Intel gpu. It's ridiculous.

    Here's benchmark of Alienware m15 i7-8750h with RTX 2080 max-q

    When it's on mains:

    https://www.userbenchmark.com/UserRun/17328078

    TimeSpy: https://www.3dmark.com/spy/7325614

    FireStrike: https://www.3dmark.com/fs/19472003

    When it's on battery:
    https://www.userbenchmark.com/UserRun/17328177

    TimeSpy: https://www.3dmark.com/spy/7325968

    FireStrike: https://www.3dmark.com/fs/19469873
     
Loading...

Share This Page