Announcement

Collapse
No announcement yet.

Benchmarking The $25 Raspberry Pi 3 Model A+ Performance

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #21
    Odroid XU4 and Pentium J5005 added: https://openbenchmarking.org/result/...RA-1811244RA22

    Comment


    • #22
      Originally posted by Danny3 View Post
      I wonder if they call removing the better and more secure networking type an improvement.
      For me, buying a desktop or server computer without ethernet seems very stupid.
      Not a server nor desktop computer, comparison invalid.

      Comment


      • #23
        A17, A53 uarch are quite old already, more than 3 years old. Why are the SoCs on our SBCs stuck in the past like that?

        Comment


        • #24
          Originally posted by mbello View Post
          A17, A53 uarch are quite old already, more than 3 years old. Why are the SoCs on our SBCs stuck in the past like that?
          embedded stuff has a pretty long support life, usually because its purpose does not change or require more processing power as the time goes on.

          It's not uncommon for a vendor to keep selling the same hardware for 5+ years, and support it (with his own customers, aka OEMs making hardware with it) for a decade or more.

          This is true for most SBC or even Arduino/microcontroller boards too, they are still perfectly capable of doing the tasks they were able to do from day 1, they are still cheap, and people still need something cheap to do these tasks.

          Comment


          • #25
            hello michael and phoronix members

            For on-device deep learning benchmark, you may interest in the ncnn opensource project.
            ncnn is a high-performance neural network inference framework optimized for the mobile platform.
            It is a heavily hand-optimized and widely-used NN inference framework. It makes extensive use of NEON assembly instructions.

            ncnn is a high-performance neural network inference framework optimized for the mobile platform - Tencent/ncnn


            Many companies and individual developers use caffe/mxnet/tensorflow/... for training and use ncnn for deploying on arm devices.

            ncnn is a high-performance neural network inference framework optimized for the mobile platform - Tencent/ncnn


            There is a tool bundled for easy benchmarking, and results for some CNN models posted

            ncnn is a high-performance neural network inference framework optimized for the mobile platform - Tencent/ncnn

            Comment


            • #26
              Originally posted by starshipeleven View Post
              Not a server nor desktop computer, comparison invalid.
              I've been using as a server for years and now you tell me that it's not a server?
              I wouldn't have used it as server if it hadn't had a wired connection.

              Comment


              • #27
                Originally posted by Danny3 View Post
                I've been using as a server for years and now you tell me that it's not a server?
                It's not a server and you make no sense.

                This is a hardware refresh of the Raspberry A model from years ago which never had ethernet (it only had a single USB port) and you claim:
                -that you would have not bought it if it didn't have ethernet connection
                -that you used it as a server for years


                I'd say you've been mixing it up with the Raspi B model (that has the USB hub and the ethernt dongle bolted on) I guess, which still does NOT qualify as a server, but at least has "integrated" ethernet running like shit I guess, so you could press it into service as a (very) light server.

                Raspi A and B are two different product lines sharing common parts (basically the A lacks the USB hub and usb-ethernet chipset and is smaller as a result), A does not replace B.
                Last edited by starshipeleven; 18 December 2018, 12:26 PM.

                Comment

                Working...
                X