varelse

86 ms·

601.
by varelse 11y ago
And mess with your neighbor's life by placing your own landing placards in your fenced-off backyard.
602.
by varelse 11y ago
Grabbing quick and easy answers from the Google and SO leaves me more time for the really interesting problems in my experience so far. And it's the really interesting problems that make writing bespoke C/GPU kernels worth the ti
603.
by varelse 11y ago
I spun my wheels for the first half of my stint at a tech company by being the sort of full-stack developer who never really specialized in anything. During that time, I frequently noted that the engineers who were promoted to the top were
604.
by varelse 11y ago
For ~$30M per 1% share of the browser market, I'm surprised that: 1) There aren't more competitors 2) A subgroup of them hasn't agreed to a conformant subset of HTML/Javascript/CSS and stuck to it
605.
by varelse 11y ago
Don't anyone tell Ben Carson about this...
606.
by varelse 11y ago
Really, you guys believe technology sufficiently advanced to be indistinguishable from magic (h/t to Arthur C. Clarke) will be festooned with the alien tech equivalent of pop-up ads and banners easily detectable by puny earthlings? Al
607.
by varelse 11y ago
Project this a decade into the future and I envision all sorts of talk-radioish belly-aching from middle-aged GamerGaters who pretty much did it to themselves by sticking strictly to the easy bits and building MVPs without ever diving deep
608.
by varelse 11y ago
Why... So... Visible?
609.
by varelse 11y ago
"And bingo! I get to press the button again, woohoo!" Wonder what the other button does... On the bright side, they bucked the trend and cited Terry Winograd's research from the 1970s instead of pretending they made this i
610.
by varelse 11y ago
The tone of the article is way over the top, but I 100% agree with this quote: "[E]xperience has shown repeatedly that good designs arise only from evolutionary, exploratory interaction between one (or at most a small handful of) excep
611.
by varelse 11y ago
Call the rental service C2G: the sharing economy of your life, and it's unicorn time...
612.
by varelse 11y ago
They're not the only ones apparently: http://www.prnewswire.com/news-releases/origination-of-googl...
613.
by varelse 11y ago
I get the feeling that the original poster never actually played Star Raiders, a game in which all one does is blow up the enemy, occasionally pausing to chase the robot that delivers recharges from the starbases... That said, 36 years late
614.
by varelse 11y ago
More like a rough idea than full-blown design, but such a game could implicitly teach astronomy the way Balance of Power implicitly taught geography.
615.
by varelse 11y ago
Inplace is a nice late phase optimization, but the root problem here appears to be that Google's convolution kernels are crap compared to those in cuDNN3 and Neon (I asked Scott Gray about this directly and I trust his wisdom). No su
616.
by varelse 11y ago
Xen virtualization disables P2P copies ergo GPUs have what we call a "failure to communicate and some GPUs you just can't reach (without going through the CPU that is)."
617.
by varelse 11y ago
>do you use multi-GPU for your DL work? If so, how often? I'm using it right now, to great effect. I can't really say what for, but methinks I'll be using it every hour of every day for the foreseeable future.
618.
by varelse 11y ago
You might be seeing the final gasps of Google's longstanding and now-reversed anti-GPU stance in TF's initial GPU performance. NVIDIA support alone will make sure TF knocks it out of the park down the road. IMO it's crazy to
619.
by varelse 11y ago
TensorFlow is currently a single-server solution pthreads app. But assuming you're serious and not dabbling, you can build a ~27 TFLOP server for <$7K in parts or buy it assembled for ~$9K. http://exxactcorp.com/inde
620.
by varelse 11y ago
No but an SMX is the GPU equivalent of a CPU core... And there are 24 of them in a GM200...
621.
by varelse 11y ago
I suspect this is a just-so-tale because the capability is so new and the GPU programming skill required to implement it efficiently is so rare. TLDR: just say no to parameter server (most of the time). But IMO the real obstacle to horizon
622.
by varelse 11y ago
>Tensors are the future "Tensor said the tensor. Tension, apprehension, And dissension have begun." https://en.wikipedia.org/wiki/The_Demolished_Man Tensors were the future even before they were cool.
623.
by varelse 11y ago
Why yes... https://github.com/BIDData/BIDMach
624.
by varelse 11y ago
Interesting, it does appear to be gone... Basically, old NVIDIA marketing material used to state that GPUs double in performance every year or so whilst Moore's Law is slowing down w/r to CPU performance because clock speeds have
625.
by varelse 11y ago
I'm done ranting about Google and GPUs at last. From a technical standpoint, I believe TensorFlow vindicates my viewpoint and that's enough for me. This is closure I really needed, OK? And besides, now I can concentrate my ener
626.
by varelse 11y ago
Actually lasagne is pretty good, I wasn't targeting you (but then you're pretty good at winning kaggle competitions so perhaps there's a connection there, no?)... I'm thinking mostly of Theano, which, from a performance
627.
by varelse 11y ago
Might one conclude that this influential opinion might have created a big old stinky career mess for someone in the field that took many months to clean up and cost one a fair chunk of change to do so? So perhaps one could be forgiven for
628.
by varelse 11y ago
That... Was not... My experience there... But kudos to the deep learning guys for overcoming that potential energy barrier NVIDIA couldn't surmount on their own...
629.
by varelse 11y ago
Sure, we all start out that way. But by the 3rd or 4th year, you frequently just want to write that thesis and get on to the next thing because you've already advanced the field as much as you're going to as a grad student. I hav
630.
by varelse 11y ago
Sure, now. 4 years ago, not so much. One might even conclude that one person inside Google had written a brawny and influential paper that was right in a lot of ways, but wimpily dead-wrong to apply to GPUs.

More ›