Wow, nervana, and now movidius. Intel are eating a lot of non nvidia machine learning hardware. Theyre also really into opencl, half the talks at this years international workshop on opencl were from intel http://www.iwocl.org/attend/sessions/ Interesting times :)
its possible to be experienced and not say no you know? the opposite is also true... the trick really is not saying yes or no, but understanding what your boss really wants, and offering a realistic way forward to achieve that.
this is really old, and intel sponsored. so intel research states that gpus are 2.5 times faster, nvidia says two orders of magnitude faster. presumably the truth is somewhere in between. in any case, it depends on the problem you are trying to solve, but in machine learning space, using convolutional neural networks, gpus are clearly the dominant mechanism recently.
have you actually tried writing highly optimizer cuda? its reallyhard. nothing is "for free". a lot of the caching (not all) has to be hand coded. you can see how much attention to detail scott gray puts into his kernels eg at https://github.com/NervanaSystems/maxas
1. arrange a 1:1 skype call with anyone and everyone on the team who might be either interested in what you are creating or to whom you might be able to contribute something useful. 2. try to find someone who clearly understands the needs of the organization and who has some painpoint you can help with. this is unlikely to be the ceo, because their painpoints are very high level and abstract. it will be some technical guy who has a bunch of tasks to do, some of which theyd be happy to offload. do those tasks for them, earn karma, youre in.
its actually harder to read, in my opinion, since the % character is often used for formatting,eg '%s' % somevar.
personally, i dislike such spammy prs. it adds nothing to my code, and its not like accepting the pr is going to attract you as a developer on my code.
on a related note, i went into a brick and mortar shop to buy earphones the other day. turns out they wont let you try them, for various reasons. in the end, i looked on the internet for reviews, and bought some headphones on amazon instead, and they are perfect, awesome. so, the brick and mortar electronic equipment store adds what value?
i guess if we are hacking wifi, we probably know approximately which continent they are in. but even if they are typing say pinyin, theres a spacebar between each word, and three characters per word, seems like an obvious heuristic?
yeah, calling it less may have been fun at the time, but its poor ui design. if you call a method 'add_one', having it multiply by three might be what you intend, but its not very intuitive.
Im not sure i understand why this is a big deal. As the guy said, it was designed to send a message. As far as drugs, youd have to be insane to do drugs in china. The laws on drugs are really harsh. As for the reason, presumably the organizer hadnt paid the right bribes.
They are. Not sure that changes my point much? The user database/registration are entirely separate. There is no lockin being circumvented by virtue of both being tencent.
Well, there's an opportunity. Five years ago, everyone in China used to use qq, which is kind of like Chinese Facebook: everything public. Then wechat came out, and now I never use qq any more, I mean, noone does.
It's not as good as Chinese wechat though, because it's really spammy: you click on like, everyone can see. On wechat you only see photos, likes, comments, from your direct friends. It is much more private. Once you've tried it, it's kind of hard to put up with the very public nature of Facebook, I find.
To be fair, when I was a kid, we didn't have internet. We only had library, and computer journals, the paper kind. Nowadays, for research papers, read arxiv. For news read hackernews. For learning stuff, Google, stack overflow. Fiction? Kindle app.
I dont know about "substantial" but ultra-violet does technically cause damage. You can see this in posters left in particularly sunny shop windows for extended periods of time.
Not really. No more than Counter Strike let people decide for themselves how many points they deserve... In our freetime, we play games in systems that decide our worth for us. I see no reason why work should be any different. In a management role, I get far happier, more productive people, when I reach out and find out what they're doing. People like it when people take an interest in them, and what they are doing.
The problem is that, unlike exams and stuff, where lots of effort went into studying how long each question should take a 'typical' person, in actual work, the problem hasn't actually been solved before, at least not by the current team, and no-one really knows how long it would take. There's no obvious theoretical objective way of measuring. So, you can waste two thirds of your time on estimating and negotiating times and pays for little micro-projects (which is doable, though at a huge overhead), or simply work normal hours, and check that everyone is approximately working to the best of their ability.
This latter method is by far the easier method, for everyone. I have worked remotely quite a lot, using both methods, and the second method is far more sustainable, in my opinion. (I mean, you just sync your hours with your workmates, and just make sure you're always online in the online chat application; if the company doesnt chat in your absence on online chat, well, it might not be the right company for you; but many do. A lot).
Not really. Computers let us create any virtual environment we want, fulfill all of our fantasies. so what do we use it for? Play cs: simulate approximately tribal warfare, and hunting for food. Genetically this is what millions of years of evolution programmed us to enjoy.