PSA: Two variants of the Nvidia MX150 Graphics Chip Exist

scrimshaw on 25/03/2018 - 22:54
Last edited 25/03/2018 - 22:56

Nvidia has been up to some sneaky business lately and if you're in the market for some new laptops in the recent spate of tech deals, you may want to do some extra research on what kind of GPU your laptop will have. Specifically, there are actually 2 versions of the popular MX150 GPU that's floating in the market right now. The underclocked variant performs between 20~25 percent slower than the vanilla version.

https://www.notebookcheck.net/Nvidia-has-been-sneaking-in-sl…

Quoted from Notebookcheck

TL;DR: We've discovered two distinct versions of the GeForce MX150 with wide performance differences and power demands. The second version is notably slower and less demanding than the "standard" MX150 with underclocked clock rates, Boost rates, and VRAM not unlike a Max-Q GPU. This slower "MX150 Max-Q" can only be found on 13-inch Ultrabooks so far. We recommend being cautious if purchasing a notebook with the MX150 GPU as neither Nvidia nor the manufacturers have been explicitly advertising the slower GPU version.

Please see our dedicated page on the GeForce MX150 for more technical information on the two versions of the GPU.

Nvidia's response was that it's ultimately up to the notebook's manufacturer to choose between the 'slow' and 'normal' version of the graphics chip. Best to do your research before you buy.

Computing

Related Stores

NVIDIA
NVIDIA

Comments

  • Diji1 on 25/03/2018 - 23:20
    -5

    That's just how things work not Nvidia being sneaky …

  • Riczter on 26/03/2018 - 00:28
    +3

    Nah thats bullshit, if they are different by that much they should be labelled differently, like MX150u or something.
    Cheers for the heads up.

    • Diji1 on 26/03/2018 - 10:25
      +1

      You seem blissfully unaware that exactly the same mobile GPU behaves differently in various devices as in the performance is all over the place because of throttling - as Notebookcheck is well aware through all their benchmarking.

      So complaining that a GPU has different clock speed is not required, it would have a different clockspeed if it was the same GPU in a different device. What's the difference other than user self delusions?

      • tonyjzx on 26/03/2018 - 11:59
        +1

        Just because its par for the course for Nvidia doesnt mean its not terrible practice.

        Be that as it may, we're talking about rubbish 64-bit crippled gpus… its either 'poor' vs. 'v.poor' performance so people defending this is kind of funny.

      • Riczter on 26/03/2018 - 14:21

        Do other cards have different TDP as well? I know different desktop graphics cards have different overclocks which give them minor performance differences, and thermal throttling causes minor differences in notebooks. But here we are talking about two very distinct versions of the card, one with 10W TDP, and one with 25w TDP. 36% Base Clock speed difference, 17% Memory difference, 32% Boost difference which is 20-25% performance difference. I expect differences on desktop graphics cards because they are clearly labelled "OC" or similar, and in notebooks if the cooling is not up to par thats an external factor, this seems very deceptive on NVIDIAs part.

  • KaTst3R on 26/03/2018 - 02:10

    Seems like Nvidia did a stealth nerf

    • Diji1 on 26/03/2018 - 10:29

      No it seems that people do not understand that having the same mobile GPU means wildly differing performance in different devices due to throttling.

      They're claiming that getting this particular GPU means you were getting a particular, lower level of performance when it meant no such thing. Throttling could mean this slower clocked GPU runs faster than a faster clocked GPU. There is no baseline with mobile GPUs.

      • Kangal on 26/03/2018 - 10:49

        There's a 20-25% difference in performance, which is huge.
        However, the difference in Clockspeed is 35-40% between the two.

        They both have the same Cuda Cores and VRAM, so they are actually the same hardware: GT 1030 (aka MX 150).
        Nvidia has very little control here on how the OEMs will implement their hardware in their product, especially when it comes to the clockspeed and thermal profile. So I would say Nvidia in this matter are Not Guilty, but that doesn't mean they're innocent.

        For one, when the details emerged about Intel's new Iris iGPU's, and the new Intel-AMD partnership iGPU, and the Ryzen Raven-Ridge GPU's….. I didn't know about the MX150 disparity. I compared everything to Nvidia's GT 1030 (desktop) which performs equal to the MX150 in the regular (not-underclocked) version. So overall I was a little disappointed in Mobile-Vega as my conclusion was:
        Intel-U + Iris = Good Performance, Best Efficiency
        Ryzen + Vega = Great Performance, Okay Efficiency
        Intel-HQ + Vega = Great Performance, Low Efficiency
        Intel-MaxQ + MX150 = Best Performance, Good Efficiency

        …so it seems I overvalued Pascal, and my data points were skewed as I was thinking the 25W card was really drawing out 10W as explained.

  • AdamFury on 28/03/2018 - 16:14

    Just checked my Xiaomi Air (2018) with GPU Z. It appears to have the slower variant for anyone wondering.

Login or Join to leave a comment
Login Join