Sign in / up

back to article Machine learning library TensorFlow can count to potato... I mean, 1.7

Open-source machine-learning boffins rejoice! Numerical computation library TensorFlow 1.7.0 made a discreet appearance this morning, just a month after 1.6 dropped. TensorFlow was originally developed by Google Brain for internal use before going open source in November 2015. The processor-intensive library makes full use of …

COMMENTS

Post your comment
House rules Send corrections
Add to 'My topics' unstar *
  1. Tom 7

    Performance is variable,

    I've got an I7 8 core laptop. Tensor flow runs one the 'standard' handwritten number recognition rests in around 20 minutes on all 8 cores. This is a bit faster than on my ancient gx720. Caffe runs the same test in about 25 minutes on one core out of the box and a noticeably faster when compiled for the chipset and runs 5c cooler too. I didnt try recompiling Tensor because it didnt tell me it might be a good idea.

    Now the GPU on the RaspberryPis is rated at 24Glops and my I7 is rated at 42 or so I'm looking forward to seeing that working ( or rather a box of Pi zeros working - in theory the PiZeros work out about twice as expensive as the cheapest Titan so they are value for money for up to a box load).

    I must play with Snap and see if it is that much faster.

    I cant think of any reason why Google would release slow software on the world tho...

    1 1 Reply
    1. doublelayer Silver badge
      Reply Icon

      Re: Performance is variable,

      As I understand it, you can't play with snap, because IBM will only let it run on high-end (I.E. more expensive than you or anyone else can afford) servers made by their hardware partner, the totally-open-and-not-unreasonable computer corporation. Oh sorry, that's actually spelled IBM. So while it might actually run faster than anything else, you can't verify said measurements until you have already committed to using IBM stuff, at which point I'm sure they get the benefits of having designed for their hardware. If it's really important to them, I'm sure they will find a way to show you how bad TensorFlow could be if you deliberately brea...is in practice when compared to the code cranked out by the aforementioned interested-in-machine-learning-not-all-of-your-research-budget computer corporation. In other words, keep playing with TensorFlow and the other machine learning systems that we have now; those will be the important ones later.

      0 1 Reply
  2. Alister

    you say potato, I say tomato

    Let's call the whole thing Pi...

    3 0 Reply
  3. Anonymous Coward
    Anonymous Coward

    Jolly Good show, one for the AI team

    Counting potatoes will be really necessary once AI has eliminated Humans from the planet.

    They will need to determine where the next threat will come from.

    Potatoes being about the next in line in intelligence to AI deploying humans.

    3 0 Reply
    1. allthecoolshortnamesweretaken
      Reply Icon

      Re: Jolly Good show, one for the AI team

      You are obviously forgetting squirrels.

      0 0 Reply

POST COMMENT House rules

Not a member of The Register? Create a new account here.

Forgotten password ?

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like