A Look Inside China’s Chart-Topping New Supercomputer

Much to the surprise of the supercomputing community, which is gathered in Germany for the International Supercomputing Conference this morning, news arrived that a new system has dramatically topped the Top 500 list of the world’s fastest and largest machines. And like the last one that took this group by surprise a few years ago, the new system is also in China.

Recall that the reigning supercomputer in China, the Tianhe-2 machine, has stood firmly at the top of that list for three years, outpacing the U.S. “Titan” system at Oak Ridge National Laboratory. We have a more detailed analysis …

A Look Inside China’s Chart-Topping New Supercomputer was written by Nicole Hemsoth at The Next Platform.

Nvidia Rounds Out Pascal Tesla Accelerator Lineup

Nvidia wants for its latest “Pascal” GP100 generation of GPUs to be broadly adopted in the market, not just used in capability-class supercomputers that push the limits of performance for traditional HPC workloads as well as for emerging machine learning systems. And to accomplish this, Nvidia needs to put Pascal GPUs into a number of distinct devices that fit into different system form factors and offer various capabilities at multiple price points.

At the International Supercomputing Conference in Frankfurt, Germany, Nvidia is therefore taking the wraps off two new Tesla accelerators based on the Pascal GPUs that plug into systems …

Nvidia Rounds Out Pascal Tesla Accelerator Lineup was written by Timothy Prickett Morgan at The Next Platform.

What Will GPU Accelerated AI Lend to Traditional Supercomputing?

This week at the International Supercomputing Conference (ISC ’16) we are expecting a wave of vendors and high performance computing pros to blur the borders between traditional supercomputing and what is around the corner on the application front—artificial intelligence and machine learning.

For some, merging those two areas is a stretch, but for others, particularly GPU maker, Nvidia, which just extended its supercomputing/deep learning roadmap this morning, the story is far more direct since much of the recent deep learning work has hinged on GPUs for training of neural networks and machine learning algorithms.

We have written extensively over …

What Will GPU Accelerated AI Lend to Traditional Supercomputing? was written by Nicole Hemsoth at The Next Platform.

DockerCon 2016 Day 2 Keynote

This is a liveblog for the day 2 keynote of DockerCon 2016, which wraps up today in Seattle, WA. While today’s pre-keynote warm-up doesn’t include laser-equipped kittens, the music is much more upbeat and energetic (as opposed to yesterday’s more somber, dramatic music). If the number of laptops on the podium is any indicator (yesterday it was a cue to the number of demos planned), then today’s keynote will include a few demos as well.

Ben Golub kicks off the day 2 keynote—with the requisite coffee shot that is a sacrifice to the “demo gods”—and offers up some thanks to the supporters of last night’s party at the Space Needle. Golub quick reviews the key announcements and demos from the day 1 keynote (see my liveblog here). Today, though, will be focused on democratizing Docker in the enterprise. In referring to Docker’s adoption in the enterprise, Golub shares some numbers that vary widely, and admits that it’s really difficult to know what the real adoption rate is. He points to multiple “critical transformations” occurring within the enterprise: application modernization, cloud adoption, and DevOps (process/procedure/culture changes).

This leads Golub into a discussion of anti-patterns, or fallacies. The first fallacy he Continue reading

DockerCon 2016 Day 1 Keynote

This is a liveblog for the day 1 keynote of DockerCon 2016, taking place over the next couple of days in Seattle, WA. Before the keynote starts in earnest, Gordon the Turtle entertains attendees with some “special” Docker containers that affect the display on the main stage: showing butterflies, playing sounds, launching a Docker-customized version of Pac-Man, or initiating a full-out battle of laser-shooting kittens.

The keynote starts with Ben Golub taking the stage to kick things off. Golub begins his portion with a quick “look back” at milestones from previous Docker events and the history of Docker (the open source project). Golub calls out a few particular sessions—protein folding, data analysis in sports, and extending a video game—and then unveils that these sessions are being presented by kids under the age of 13.

This leads Golub into a review of the efforts of Docker (the company) to democratize containers:

  • Increasing usability
  • Enhancing portability
  • Extending community

Golub gives a “shout out” to the technologies underpinning modern Linux containers (namespaces, cgroups, etc., and their predecessors) and calls out the 2,900+ contributors to the open source Docker project. He then spends the next several minutes talking about various metrics—pull requests, containers Continue reading

Tesla review: What you need to know about charging

Before you buy an electric car, you need to understand charging. It’s a huge deal. You think it works almost like filling the gas tank. It doesn’t. Before going on long trips, you first need to do math and a bit of planning.


The Math

Like BMW model numbers indicate engine size, Tesla model numbers indicate the size of the battery, so my "Tesla S P90D" has a 90kwh (killowatt-hour) battery, with a 286mile range. Their lowest end model is the “Tesla S 60”, which has a 60kwh hour battery, or a 208mile advertised range.

In the United States, a typical plug is a 120volt circuit with a maximum of 15amps. Doing the math, this is how long it’ll take for me to recharge the battery:


That’s right, 1.4 days (or 2.1 days for a 90kwh car). This is the absolute worse case scenario, mind you, but it demonstrates that you have to pay attention to charging. You can't simply drive up to a station, fill up the tank in a couple minutes, and drive away.

Let’s say you live in Austin, Texas, and you have a meeting in Dallas. You think that you can drive up to Continue reading

Intel Management Engine’s security through obscurity should scare the **** out of you

For a long time there was a commonly held belief that open source products were inherently more secure because there was nothing hidden. The thought was that with the code for popular applications out in the open, there’d be scores of good guys looking at every line and bugs and flaws would be few and far between. Alas, this turned out to be a pipe dream because even the most examined code can still contain flaws so obscure and arcane, even highly skilled and incredibly talented coders can’t find them. Why? It’s usually because the good guys don’t have the time to play hacker as intensely as the real hackers do. For the bad guys, the rewards for finding exploitable flaws are tangible while for the good guys, the cost of not finding flaws far exceeds, by orders of magnitude, the value of the few flaws they do find because those flaws are the most easily found.To read this article in full or to leave a comment, please click here

Intel Management Engine’s security through obscurity should scare the **** out of you

For a long time there was a commonly held belief that open source products were inherently more secure because there was nothing hidden. The thought was that with the code for popular applications out in the open, there’d be scores of good guys looking at every line and bugs and flaws would be few and far between. Alas, this turned out to be a pipe dream because even the most examined code can still contain flaws so obscure and arcane, even highly skilled and incredibly talented coders can’t find them. Why? It’s usually because the good guys don’t have the time to play hacker as intensely as the real hackers do. For the bad guys, the rewards for finding exploitable flaws are tangible while for the good guys, the cost of not finding flaws far exceeds, by orders of magnitude, the value of the few flaws they do find because those flaws are the most easily found.To read this article in full or to leave a comment, please click here

DockerCon Cool Hack Challenge: Tyrion Cannister Neural Style GUI

As students at Holberton School, a software engineering school based in San Francisco, Siphan and I are exposed to lots of exciting technology! The main goal of the school is to produce full-stack engineers in two years. Although we are only four months into the program, we are already learning how to use the Docker platform.

A few weeks ago, we held our very first hackathon at school – and of course it was focused on Docker! The school’s founders (one of whom was an original member of the Docker marketing #boomteam) thought now was the perfect time for us to participate in a hackathon, so they teamed up with Docker to make it happen.

In the spirit of DockerCon, our guidelines were basically the same as those of DockerCon Cool Hack Challenge: make the most awesome things we could think of using Docker, and do it in ten hours. Cue the intense music.

Continue reading

Ethereum/TheDAO hack simplified

The news in the Bitcoin world is the Ethereum/DAO hack. I thought I'd write up a simplified explanation.

What is Bitcoin?

I'm sure you know, or have an idea what Bitcoin is, but I'll explain it again in terms that are important here.

Bitcoin is just a public ledger (the "blockchain"), of all transaction there ever was. This ledger is huge (80-gigabytes) and growing, but Moore's Law says computers grow even faster, so that shouldn't be a problem.

Each entry in the ledger says to move the coins received in these previous entries, and give them to this recipient. In other words:
move these coins I received there, to this guy here
In the future when that guy spends the coin in a new transaction, he'll refer back to this transaction here. That's why it's called a "chain" -- every transaction refers to a previous one, back to the original creation of the coins.

Actually, these transactions aren't fixed as simply moving money around. Instead, a script is used, written in  a bitcoin-specific programming language that can do things like add, multiply, and compare numbers. However, it's not a full programming language. It can't call functions or execute loops. It's Continue reading

The fight over a new Wi-Fi channel is coming to a head

Globalstar’s plan to open a new Wi-Fi channel under its control is nearing the moment of truth after years of regulatory wrangling.The U.S. Federal Communications Commission is considering Globalstar’s plan, and a vote could come at any time. News reports suggest it might be close.If Globalstar gets its way, Wi-Fi users in the U.S. will have one more channel, which could reduce congestion and improve performance. But both their mobile devices and the hotspots they use would need firmware upgrades to take advantage of the new frequency, and the channel wouldn’t necessarily be open to everyone.To read this article in full or to leave a comment, please click here

The fight over a new Wi-Fi channel is coming to a head

Globalstar’s plan to open a new Wi-Fi channel under its control is nearing the moment of truth after years of regulatory wrangling.The U.S. Federal Communications Commission is considering Globalstar’s plan, and a vote could come at any time. News reports suggest it might be close.If Globalstar gets its way, Wi-Fi users in the U.S. will have one more channel, which could reduce congestion and improve performance. But both their mobile devices and the hotspots they use would need firmware upgrades to take advantage of the new frequency, and the channel wouldn’t necessarily be open to everyone.To read this article in full or to leave a comment, please click here

How to build a network of Linux routers using quagga

This post lists the commands required on each node to build a network of three Ubuntu Linux routers. Each router is connected to the other two routers and is running quagga. Each router is also connected to a PC running Ubuntu Linux.

three-nodes-kr

I use this network configuration to evaluate network emulators and open-source networking software in a simple scenario. Readers may find these commands useful in building their own configuration scripts.

I provide “copy and paste” commands so the network can be configured quickly.

Creating a basic topology

The physical — or virtual — network installation and the management network setup is outside the scope of this post. The method used to build the lab topology depends on the equipment, and/or the network emulator and hypervisor technology you are using.

I assume you already have six machines running and connected in a network as shown above, and I assume you have a management network set up so that each machine can communicate with the host computer and with the Internet.

Router configuration

Each router needs to install the quagga router package, configure quagga, and then configure the network using the quagga VTY shell. Optionally, quagga daemon configuration files may be Continue reading

5 things you should know about password managers

New data breaches are coming to light almost weekly and they reveal a simple but troubling fact: many people still choose weak passwords and reuse them across multiple sites. The reality is, remembering dozens of complex passwords is almost impossible, and carrying them around on a scrap of paper that you have to keep updating is a huge hassle. That’s why password managers exist. Here’s why they’re important, and how to get the most out of them.To read this article in full or to leave a comment, please click here

5 things you should know about password managers

New data breaches are coming to light almost weekly and they reveal a simple but troubling fact: many people still choose weak passwords and reuse them across multiple sites. The reality is, remembering dozens of complex passwords is almost impossible, and carrying them around on a scrap of paper that you have to keep updating is a huge hassle. That’s why password managers exist. Here’s why they’re important, and how to get the most out of them.To read this article in full or to leave a comment, please click here

58% off RAVPower FileHub Wireless Hard Drive, WiFi Bridge & Sharing Media Streamer – Deal Alert

You may find many uses for this versatile and powerful file hub from RAVPower, currently discounted 58% on Amazon from $75.99 down to just $31.99. Let's run down the features: Let up to 5 devices wirelessly connect to the FileHub wireless network, to access its SD card and/or USB storage to share files Media server capabilities, allowing wirelessly connected devices to stream audio and video from any attached storage device Move files freely between all connected methods of storage and devices, including tablets and phones Bridge an existing wireless network, create a new secure network and share with family or friends, or bridge to connect to devices such as Chromecast in hotels Built-in power bank charges your device when you can't access a power source RAVPower's FileHub currently averages 4 out of 5 stars from over 1,700 customers (read reviews). For more information or to explore buying options, see the discounted FileHub now on Amazon.To read this article in full or to leave a comment, please click here

1 … 2,897 2,898 2,899 2,900 2,901 … 3,856