Re: ehhh
It's not just "is it down". It's "what happened". Site owners will take to Twitter to explain downtime.
532 publicly visible posts • joined 4 Dec 2012
Knock it if you want, but real time tweets are a pretty good or at least a possible way to gauge public opinion, get instant reactions to live events, see if/when websites have stopped working etc.
What kills me is that there is no search box on Twitter's front page (not logged in). I had to type search.twitter.com to get a redirect to https://twitter.com/search-home.
https://twitter.com/hashtag/BrianWilliamsMisremembers?src=tren
Google's move is just a continuation of hoovering up data and adding more semantic wolfram-like search. Like snippets from Wikipedia, Google News results, or now a box of 2-3 tweets. As long as it doesn't crowd search results/vertical space it is probably good. For someone who still uses Google anyway.
They're in secretive venture capital and trade secrets mode.
The fact that Google, NASA, Lockheed, etc. have given D-Wave millions made them look a lot better, but their refusal to be academic about how it works tarnishes that. I don't count Google's quantum computing side effort against D-Wave since D-Wave admits they aren't making a general-purpose quantum computer.
I've been called a Google Glass apologist, but this is what the device should have been. Full lens display enabling augmented reality, not a corner display that causes eye strain. Go ahead, wear it while driving... the people who can afford the best of these products in 5 years will be among the first to adopt driverless cars as well.
If Intel makes smaller and more powerful "Quark" type chips, and more fashion designers get brought on board, it could be possible for one of these companies (Google, Microsoft, Sony, etc.) to make a smartglasses model that are indistinguishable from normal glasses (at least the ones with thicker frames) or sunglasses. Tiny cameras are Cold War technology... make it small enough, and every spectacle bearer could be a hidden glasshole. The self-contained nature of HoloLens (and whatever this marketing speak secondary GPU does) might enable less input lag than Oculus or a smartphone-tethered pair of glasses. So possible less AR/VR barfing.
In any case, professionals and niches are a good first target. Surgeons for Glass, NASA for HoloLens.
They have to be pretty small and pretty far away from Neptune.
The many TNOs found since the 90s and 2000s are very small, Pluto-like, with very eccentric orbits and satellites in some cases. More recent finds are more like Ceres.
X marks the... They SAID there was a mystery planet there – NASA
However, despite finding thousands of new stars in its survey of the sky, WISE was unable to spot any object the size of Saturn or larger to a distance of 10,000 astronomical units (1.49597871 × 1015m, or 9.29558073 × 1011 miles) and nothing bigger than Jupiter out to 26,000 AU.
It's the "glocal" scale of it that impresses me. A dispute, scandal, or ban in so many countries and major cities, and still they manage to operate and achieve a $40 billion valuation. I'm already anticipating autonomous Uber with the drivers kicked to the curb like human trash. Taxi companies would have done that same thing in 10 years in a non-Uber world, but Uber could put itself into a monopoly position.
In the off-chance that Cameron could somehow convince our Silicon Valley friendly Obama to back his stupid plan to weaken encryption, and somehow get the result passed by a Republican-controlled Congress, the tech community will go to war.
Make Cameron retire plz Brits so we don't have to laugh, cry and spit at this nonsense.
(2011) Wireline Costs And Caps: A Few Facts
"2 cents to 5 cents per gigabyte. The actual bandwidth cost to a large carrier like Time Warner or AT&T, depending on how you do the accounting. $1/month/customer. Going Down: Bandwidth usage growth per customer. The rate has been about 30% per year, with the rate slightly falling the last few years. The growth in average usage is actually going down slightly, per Cisco VNI and the MINTS data of Professor Andrew Odlyzko. Going Down: Capital investment required. Going Up: Profit Margins. Prices for broadband have generally been going up in the U.S. since 2007 while costs drop."
CISCO: The Zettabyte Era - Trends and Analysis
"Global IP traffic has increased fivefold over the past 5 years, and will increase threefold over thenext 5years. Overall, IP traffic will grow at a compound annual growth rate (CAGR) of 21 percent from 2013 to 2018."
Wikipedia: "On September 12, 2014, Google announced that development on VP10 had begun and that after the release of VP10 they plan to have an 18 month gap between releases of video standards." and "HEVC is said to double the data compression ratio compared to H.264/MPEG-4 AVC at the same level of video quality."
If the average broadband was faster and uncapped, with more competitors in each market, there wouldn't be a need for net neutrality rules. Multiple players have sold 1 Gbps for $60-70 a month (not just GORGLE), and that should be capable of handling multiple 8K streams. Or even more video once successors to H.265 and VP9 have appeared. There's no excuse for peddling 5-50 Mbps at $30 a month and caps and throttling when fixed bandwidth costs have fallen exponentially. Prioritization my arse. Get 1-10 Gbps if you need more oomph. Prioritization is practically a security risk anyway. Everything should be encrypted, in which case you can't or won't prioritize.
Who knows if Wheeler's 25/3 broadband proposal would do any good. Greedy rural ISPs can just drop the word broadband from their advertising. "Git yer 1 MEGAbit MEGA Internet Access!"
This is one half of a teraflop as we know it. NVIDIA's slick marketing got them a slick headline. It's also called "Erista" not "Arista".
NVIDIA CES 2015 Press Conference Liveblog
"11:12PM EST - Confirmed that it's Erista
11:12PM EST - NVIDIA is proclaiming it a 1 TFLOPS GPU, though this is at FP16 as opposed to the more normal FP32 metric for TFLOPS"
100 GFLOPS/W? Isn't that leaps and bounds ahead of all CPUs and GPUs out there?
Apparently Quadro K6000 is capable of "5.2TFLOPs/sec for single precision (FP32)" consuming 225 W. So about 23 GFLOPS/W. Double precision for the K6000 card would be 1/3 of that (7.7 GFLOPS/W). November 2014's Green500 #1 most efficient supercomputer achieved 5.27 GFLOPS/W.
So the real story here is the power efficiency of Arista/Erista/X1, even if it's mainly for graphics performance, mobiles, and tasks like vehicular vision rather than supercomputing. If they could combine X1's power efficiency with 1/3 FP64, you could build a 33 MW exaflop supercomputer (not really though). Some cards out there also do 1/2 FP64.
Edit: According to Anandtech it is 1024 GFLOPS of FP16 operations, 512 GFLOPS of FP32. I don't see 10W anywhere, but I see what looks like 64 "single precision GFLOPS/W normalized" on page 1. Feel free to crack the code.