Announcement

Collapse
No announcement yet.

Benchmarks Of The NVIDIA Tegra K1 & Its Hotness

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

  • Benchmarks Of The NVIDIA Tegra K1 & Its Hotness

    Phoronix: Benchmarks Of The NVIDIA Tegra K1 & Its Hotness

    Here's our first public benchmarks of the NVIDIA Jetson TK1 ARM development board powered by the Tegra K1 SoC with quad-core+1 Cortex-A15 and NVIDIA Kepler GPU. There's also some thermal metrics for those concerned about the active-cooling on this development board...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Can we get an Ubuntu powered ultrabook with that SoC please?

    Comment


    • #3
      do you mean netbook?

      Comment


      • #4
        Originally posted by Siuoq View Post
        do you mean netbook?
        Not necessarily - the Kepler unit on this thing should make just about any netbook IGP look useless*, and a quad-A15 without the usual power/heat constraints should stand up fairly well against lower-end x86 dualcores. Still slow by Ultrabook standards, but I expect it'll be into the low-end real laptop range.

        I can't wait to see the benchmarks of this thing!

        *Well, a turnip could probably make a netbook's IGP look useless...

        Comment


        • #5
          I think most of us are much more interested in their Denver Tegra

          Comment


          • #6
            comparisons on compute

            The compute looks quite impressive! I compared the http://openbenchmarking.org/result/1...KH-NVIDIATEG94 with http://openbenchmarking.org/result/1...PL-1303138FO84 and it generally beats all other SOCs, slotting somewhere between a core2 duo and the i3 on most marks, with cray perf being a positive outlier. Now I can't wait to see grapics perfs!

            Let's see how this stacks up against Mullins. I have a hunch...

            Comment


            • #7
              Originally posted by SyXbiT View Post
              I think most of us are much more interested in their Denver Tegra
              The 64bit version , right ?

              That will be very interesting...in special if someone makes a mITX MoBo with it.
              There is a company that make(s) a mITX with a Tegra3 (and that uses 4GB RAM) and a bunch of connectors....very low power.

              But then again, with the Kabini, Mullins and Beema (not to mention the lower and lower power drains from Intel offers), i wonder for how long ARM will continue be competitive in a compute_power/power_drain ratio.
              Not to mention the compatibility of x86 and it's advantages for many years to come....

              Because even this K1 will drain a good chunk more than a Tegra3, wonder Denver....but looking forward with great interest to Michael's comparison vs AM1,etc....should be interesting.

              As for custom heatsink (with or without standard size fan) for this MoBo, easy to do if you have a minimum of skill in DIY stuff....very easy to achieve in special because only two mounting points...just like it is to do for AM1.

              Comment


              • #8
                Originally posted by godlike_panos View Post
                Can we get an Ubuntu powered ultrabook with that SoC please?
                You could try using Picuntu on a RK3188 tablet

                Here is my comparison agains celeron J1900 http://openbenchmarking.org/result/1...PL-1405010KH96 - quite comparable results.

                Comment


                • #9
                  How about es2_info(mesa-utils-extra)? Doubt there's nothing for supporting wayland but one can always dream.

                  Comment


                  • #10
                    "With the active cooling, the Tegra K1 peaked at 41 Celsius, bottomed out at 30 Celsius when idling, and averaged to 36 Celsisus during the many benchmarks run within the aforelinked result file."
                    Shouldn't this chip use mW when idling? If you use 700mW when idle shouldn't the chip go to almost room temp? Like 24C or something?

                    Comment

                    Working...
                    X