4
lxmcf
5y

More an opinion here, feel free to disagree and explain why...

The Surface Pro X has me excited but really angry at the same time for the same reason. The fact we are seeing a flagship consumer grade ARM based computer being pushed out by such a massive company is amazing, I personally hope that ARM takes over x86 and hope this might encourage people to look at ARM as a legitimate platform to develop for and hope it also bleeds into the Linux world (sick of only finding distributions designed for servers).

But I am annoyed because it's being treated as just another computer and no one is actually looking at the bigger picture with an open mind, plus it's also more than likely going to be treated as an after thought rather than looked at like a legitimate project.

Comments
  • 2
    I personally don't think ARM is any better to take x86's place. RISC-V is a much better candidate since it does not have a restrictive license and will allow for innovation at the ISA level.
  • 1
    I don think there are ARM cpus that are as powerful as x86
  • 2
    Well for my eyes, it looks like (no offense intended) bag of potatoes.

    Surely, the battery usage etc, but come on, it's like hoping that everyone will run intel atom on their laptops.

    I have i7 7820hk in mine and still sometimes it struggles. I had to deal with low-power device before and that was one big fat hell.

    That's my take, I hope ARM will not take over, someone mentioned RISC-V. It would be same story. One will need text editor and 2 tabs in chrome for stack overflow. Other will need to run JB IDE, yt for music in background and tens of other tabs in chrome, discord, slack, something else in tray while checking on VM if build works on other OS.

    I personally consider such machine as potato, and for some people that will be enough. Not for everyone, I even doubt it will be enough for most. That's why.. No, I hope that it will be far from taking over.

    Consider that nowadays there is more and more inefficient high-level code and nobody wants to write low-level for performance.
  • 0
    2/2

    And Im not joking, my machine is right now pretty much chilling, not doing anything in particular.. Just having stuff open that I use anyway, so reopening it would be PITA.

    I think best snapdragon out there would choke already
  • 0
    @DubbaThony If it ran at all...
    I dont think there is a build of windows for arm yet... (oh wait my bad, the surface pro mentioned by OP...)

    We would have to compare them using some kind of benchmark, arm and x86 are different and freq and cores arent the only things that affect performance.
  • 0
    @DubbaThony you have a point, but you are comparing a desktop class CPU with dedicated cooling and the chip itself physically dwarfs any snapdragon.

    It's the same mentality a lot of people took to the Radeon 7 having the same performance as the 1080 TI (I think it was the TI), but no one acknowledged that the entire fabrication process was comparing 7nm to 14nm.

    Too be fair however, ARM was never designed with the desktop in mind so obviously would have major teething issues being scaled so much though.

    @beegC0de and you are correct that RISC-V would be better, however this is maybe my bias talking but I feel companies would simply take RISC-V, create there own instruction set and then boom, we have 40 variations that are incompatible with each other
  • 1
    @lxmcf I dont think nm matter. Clock, instruction sets, cache size, precaching, good cooling, matters
    nm just allow better parameters.
  • 0
    @Gregozor2121

    Yes, very true, and Im perfectly aware. Actually if that only mattered my comment would not exist.

    The mentioned modded snapdragon would be 60-70% of my CPU if only that mattered.

    @lxmcf
    And, by the way that's laptop-grade CPU from performance sector.
Add Comment