vegabook

39 ms·

391.
by vegabook 10y ago
I can't help thinking that we need a new scientific programming language that is properly vectorised and directly targets GPUs, as opposed to all these hacks that unroll loops. Now that a large proportion of data science and scientfic
392.
by vegabook 10y ago
+1. fwiw from me. All the best. I'll keep an eye on you guys.
393.
by vegabook 10y ago
I am aware of the fact that I've pushed the boat out on criticism, but this, unfortunately is a hard-assed market. Personally I think you have been too moonshot and not specific enough. I may be wrong but hopefully my views give you a
394.
by vegabook 10y ago
maybe you should fire the guy who wrote that terrible post.
395.
by vegabook 10y ago
What you have to keep in mind, is that AMD is valued at a quarter of Nvidia, and at 1/25th of Intel. Therefore the market has a huge scepticism discount built in. AMD doesn't have to hit the ball out of the park with Zen. It just
396.
by vegabook 10y ago
yet we should keep in mind that Intel makes its own chips , while Nvidia clear has an availability problem[1] which it cannot fix on its own. Therefore to the extent that Intel controls its own destiny better, we may find that to be a majo
397.
by vegabook 10y ago
@tunesmith I should add to my comment that I agree with you that many non tech people often treat technical debt as free. It's a big problem and is a major source of failure.
398.
by vegabook 10y ago
This is analogous to saying no business needs financial debt, and can bootstrap entirely out of cashflow. It's baloney. There are many situations where you can do quick-and-dirty versus (slow) elegant-and-enduring, and there exist bus
399.
by vegabook 10y ago
I disagree that the interest costs of technical debt feel invisible. I personally feel a very heavy burden when cutting technical corners / coding to specifics rather than the elegant generic cases. Indeed I spend a large amount of tim
400.
by vegabook 10y ago
> Technical debt always reflects an operations problem This is a clueless statement. Technical debt is just like any other debt. It's about getting product out fast during a phase where market-share build is crucial, building a do
401.
by vegabook 10y ago
Well I have the previous generation "Precision" M3800 which is supposed to be the top of the range "serious" line from Dell, and I can vouch for that generation's coil whine too (It's fully loaded including the
402.
by vegabook 10y ago
I use R multiple hours every day, I love it, I love the ecosystem, but I can't help thinking that it's showing its age. It does a great job on static analysis and visualization of smaller (< 1 gigabyte) data sets, but is seriou
403.
by vegabook 10y ago
...at enormously different price/flop, basically because it restricts RAM size and disables ECC in the Core chips. It's why we need AMD's Zen to be competitive again, so that this price gouging ends. Same for Tesla/Gefor
404.
by vegabook 10y ago
k but a) back pressure as per my original post with Elixir (or Apache Storm, or Flink, amongst many others), and b) if you need latency-aware cross-timezone you should probably be letting another piece of distributed infrastructure, explici
405.
by vegabook 10y ago
There is a high negative correlation between ecological and monetary wealth. Time to choose. btw southern hemisphere looks good on this measure. As a bonus, zero nukes, and low likelihood of being an attractive target.
406.
by vegabook 10y ago
just lean on the shoulders of giants who have thought all these points out before, and use Erlang (or Elixir). That includes back pressure. http://elixir-lang.org/blog/2016/07/14/announcing-genstage/
407.
by vegabook 10y ago
it's not just the hardware. El Capitan, the latest and greatest, still ships with a stone age Vim 7.3 from August 2010 . And we're told not to touch it because /usr/bin is sacrosanct. Bunch of add ins doesn't even
408.
by vegabook 10y ago
Homejoy, Datajoy, killjoy? Maybe the clue is in the name!
409.
by vegabook 10y ago
who said they were an alternative to relational databases? The problem with the post was not that the 4 mentioned were bad, only that they were the "only four worth using" which is an eye-rollingly ridiculous statement. Anybody wh
410.
by vegabook 10y ago
Clearly the relational databases are very good and very useful in more cases than time series. However timeseries is huge, and growing - it's not a niche case anymore. Anyhow my beef was with the assertion that the 4 relational databas
411.
by vegabook 10y ago
of course I've heard of column store relational databases, but they're not nearly as efficient nor scalable as Cassandra which has been built from the ground up for this. Fact is, if you don't need referential integrity, ther
412.
by vegabook 10y ago
> Given that, there are only four dbs worth using: Postgre, Oracle, DB2, and MS SQL This is total bullshit. Even a modest size database (3-5 terabyte) for a small company like mine, chokes a relational database bigtime on time series.
413.
by vegabook 10y ago
well put, if your axiomatic paradigm is laissez-faire economics, whose credibility is fast being eroded as its negative consequences (exclusive returns to capital only) become all too awfully apparent. Terms meant as derogatory such as &q
414.
by vegabook 10y ago
> The key, and I think this applies to any language, is not NOT BACK DOWN. Some things will be hugely painful, if you back down you just defer the pain [...] don't let them This is very stressful. Forcing an interlocutor to speak
415.
by vegabook 10y ago
Domain experts who have no clue about tech are doomed. They are still very numerous in business, treating the tech people as support staff, rather than strategic staff. Generally sprinkling tech in a haphazard way on top of an existing orga
416.
by vegabook 10y ago
Excellent article, well thought out, a parsimonious distillation of the essence of day-to-day management. However it ignores a crucial issue in all management jobs, and that is credibility erosion. "Boss doesn't have a clue about
417.
by vegabook 10y ago
well to raise 77 million something in their pitch is working. I'm guessing it's implicitness, ie, unlike actual local directories, as in dropbox et al, you'll just see the cloud, but there will still be a fat, local, cache. S
418.
by vegabook 10y ago
I buy like 5 books a month just to try to keep up. I've got math-overload not to mention latency-depression and algo-stress. I have cuda-mares at night. Trying to keep up is so personally costly that I sometimes wonder if I shouldn&#x2
419.
by vegabook 10y ago
but that's Dropbox. So contrary to what the article asserts, this is just another dropbox.
420.
by vegabook 10y ago
In my opinion it is a fundamental axiom of computer science that some local storage will always be necessary. We've just seen AMD put SSDs directly onto a video card [1] for precisely the reason that local bandwidth is supremely impo

More ›