Cube i9

First Intel Skylake tablet of 2016. First kickstand tablet.
Follow Price check



User Avg

Cube i9 Benchmarks and Gaming Test (Video)

Cube i9 Benchmarks and Gaming Test (Video)

A few people have been after this, some popular titles like Counter Strike Global Offensive, League of legends and Dota 2 tested on the Cube i9, as well as other titles like Call of Duty Modern Warfare 3, Dirt 3 and Half-Life 2 Lost Coast.

Another clip that covers benchmarks on the tablet like Geekbench 3, PCMark and 3Dmark. Overall, the Cube i9 is the fastest Chinese brand tablet out now, but not quite as fast as the Surface Pro 4 which is very obvious in this Battlefield 4 test. On the Surface Pro 4 M3 Core M-6Y30 I was able to get a higher frame rate due to better thermals and a higher power limit. Be interesting to see if the Teclast X3 Pro has similar thermals and performance of the Cube i9 or Teclast has pushed the Core M3-6Y30 higher with better thermals and power limits.

As it stands the Cube i9 does have a completely unlocked bios. Tweaks with power limits should be possible as long as we have thermal headroom of course and don’t run into thermal throttling. But I have no idea where to start within the advanced CPU settings in the bios…

Video tech reviewer and tech blogger. I have a huge interest in the latest tech, tablets, laptops, mobiles, drones, and even e-scooters. Active in the tech community since 2008 days of the Omnia i900 Windows phone., founder from way back.


  1. Hi Chris!
    Could you please add War Thunder to your benchmarks? Thanks a lot, keep doing cool stuff!

  2. Hmm. Chris i have now seen numerous videos of you with this device, and are a little baffle that you keep comparing oranges and apples, both in this and the previous video, the scores you compared it to was mainly atoms scores and the newest surface.

    this aint an 900US surface or an 200US atom.

    this is an coreM in the cheap-clone-segment around the 400US-line, so why not compare it against its alternatives..

    like the X2 or there coreM-i7, and alternatives in this segment CoreM / M2SSD and machines in the same price range-..

    I got the X2 and I do recall you also got it and actually reviewed it.

    Have you totally forgotten that these scores your now publicing with an sky lake is actually lower then X2 and last years coreM on numerous scores.?

    let me give you an exsempel.

    cloud gate i9 3724
    X2 pro 4503 (4193)

    Ice storm 1.2 i9 35684
    X2 pro 46183.

    and so, but my point is that 30% on GPU and +20% overall like your impression on the coreM 6gen vs 5 gen. doesn’t seem to hold on bench an paper, and that interesting why not’ and what is the reason..

    some of the reason is that Teclast is really pushing the max out of the x2 pro, and perhaps to much, ref heat and lack of fan, but I also think that the evolution of the 6gen sky lake …is not so massive as it was stated an ofcourse wants the consumer to think and that would be an very interesting point to make..

    x2 thread

    no critisme just an point and suppressed that you didn’t took notice the massive difference and how the i9 6y+515 compares against last years x2 pro’s 5y+5300.. since i have just seen your review on the x2 pro and most seems to be about that tacky seller not giving you an licensed win10 and keep pointing out the good wifi in x2 pro. if you still have the x2 pro, perhaps test some of the same test on the x2 pro and compare 5gen vs 6y coreM

    there is signs that Cube doesn’t really have got the full effect out of the skylake, and actually comes short against last years socket..(and that would be very intersesting to see and if there is throtling & limiting-problems in the i9!.

    anyway just an notice.. the values speaks for it self.

    anyway thx Chris cherise your videos, huge thumps up, but I also dont get it, why you keep having these cheap fabric-screenprotectors on when benching tempertures on an china-branded-coreM, since its like a pillow for heat. theres quite an difference how good heat can exhaust on glas versus plastic.. and comes down to the hole structure of glass, that are like an active-molecule-structure.. (you see it in big windows, where the bottom part of the windows actually gets bigger as time goes, and theres a fluid structure that moves hence gravity and is active over the years) and i would reckon the heat would be quite easyer to exhaust into the air on this tempted glass then with an pillow of plasticprotector, and partically on these 90degrees-hot devices, like the coreM.

    glass surface versus platic on glass, quite an difference how the surface-molecyles interact with air, a shame that it can’t be microscoped up, it would be obvious how pillow-like an cheap plastic protectors acts, and partially on cheap-china-clones-CoreM where you dancing with heat in an ongoing battle. 🙂 partially when benching..

    • It’s indeed caused by thermal limitations, disabling some of the inefficient throttling mechanisms and improving dissipation by using better thermal pads / thermal paste will increase performance.
      In similar circumstances the m3-6Y30 should always score ~25% higher because of the slightly better architecture.

    • Hi, thanks for the comment. I’m comparing apples to apples as the only other 6Y30 I have used is the Surface Pro 4, yes it’s twice the price of the Cube i9. Not fair? Is it too much to ask that the Core M3 in the i9 should perform just as well as the Surface Pro 4. In my Surface it was a large step up in terms of performance over the 5Y10. But here not so obvious, power limits Cube have in place.

      I’m not comparing it side by side with the X2 Pro as I sold that machine and my Cube i7 Stylus. Which is why the last layer of the screen protector is on the tablets for resale. I cannot afford to keep every tablet I buy. It’s a huge cost. I have the server to pay, a mortgage and mouths to feed etc etc I spends thousands some months on tablets. Last month alone over 2k, and now this Cube i9 I have is a huge paper weight as I bricked it looking in the bios flash settings. 500+ euros down the drain there…

      And 3Dmark scores (GPU scores) are only one factor. The Cube i9 / 6Y30 beats all the other things like Geekbench 3 PCMark etc. again power limits Cube must have set are limiting it. But it will be reviewed if I can finish it as stock, with this limit.

      If remove the plastic screen protector (remember these aren’t scratch resistant screens or I would) this would have very little effect on the temps if any.

    • Hi Jakob,

      For your information, I checked some other tests.
      Actually, the cube i9 has very close results with the Surface pro 4 m3. I guess (really as i am not specialist at all) that Teclast overclock its products to get such results.
      In the end I do not feel screwed to buy a product that hqs similqr performance to the top market product (surface pro) but slightly less than an overclocked product from prior generation.
      Anyway, not to say that you are wrong or right just to say that you scared me (I purchased it westerday) and really thought that the performance is bad.
      For other people like me, I did the check and you will get similar performance to surface pro 4 m3, which is probably what you are looking for !

      For others; wait for Teclast x3 Pro to be out on the market and see how it défends (Attack ?).


      • After examining the score between both the X2 & Cube i9, it seem that:

        a) The scores are the same because of a small difference. The 5Y10 is using the older 5300 graphics but has 24 pipes. While the 6Y30 with the 515 has 12 pipes but has a more aggressive clocked.

        So each “pipeline” of the 515 is pushing more data ( because of the better infrastructure / higher mhz ). The 5Y10 is using more pipelines and because of that its physics scores are about 20% better.

        So in short:

        6Y30: 10 a 15% better graphics but loses in the physic scores
        5Y10: 20% better physics scores but loses in the graphics scores

        This is why both seem to have a almost identical 46.000 score.


        Now, it does open some questions: Was the X2 equipped with the same thermal mod? From what i read the X2 did overheat much more easy ( higher thermal limits + no proper copper / heat mod ).

        Also i expect that the Cube i9 is not pushing to the max for several reasons. I do not remember if it was Chris or the other Cube i9 owner but it was mentioned that the GPU never ran up to 900mhz constantly. Mostly clocking in between 600 a 700mhz. This seems to indicate still a power / heat issue. The more important detail i find is: Chris is benchmarking the tablet while it laying flat on the table. There is no way that the heat can easily move ( wood is a thermal isolator ). This is also a factor in his results… I suspect also a more better heat spreader like in the Surface Pro 4 can push the tablet really higher. In other words, there is maybe another 20 a 30% of power not being used…

        Now from a upgrade point of view… It am starting to wonder if its just not better to get a X2 & better thermal mod it, compared to the massive price jump on the X3. From the results in several youtube movies it seem that with a proper thermal mod, the X2 is right on the I9 its heels ( both with thermal mods ).

        • Almost forgot to mention: I think when the Cube i9 can properly access the 900mhz range, it will probably beat the X2’s 5Y10 by a 40% margin. Just as the PR material state for the 5Y -> 6Y graphical improvements. Its just not utilizing its maximal range while i suspect the X2 with thermal mod’s is hitting its maximal range.

          So you end up with a 5Y hitting max range, vs a 6Y not hitting its max range. So both seem to be closer then they are supposed to be. It mostly seems to impact the GPU because the CPU clearly shows the improvement in architecture.

  3. Battery Bar Pro, it’s only an approximate battery figure. But it’s useful nevertheless and better than Windows own estimate.

  4. Do you think that the thermals will be a problem in longevity of the tablet? I fear that at those temps your getting, plus the sd card slot problems mean I might have to look into getting another tablet… I thought this one would be promising and it might just be early days but I was excited about the prospect of buying one of these. Keep up the good work, love your videos.

    • Shouldn’t be an issue I think, it doesn’t get that hot on the Surface of the tablet. And the TJunction max is 105c on these.

  5. hi criss.which program do you use for getting battery information?

Comments are now closed for this post.

Lost Password