Zardavas D.,Data Center |
Phillips W.A.,The Surgical Center |
Loi S.,University of Melbourne
Breast Cancer Research
PIK3CA mutations represent one of the most common genetic aberrations in breast cancer. They have been reported to be present in over one-third of cases, with enrichment in the luminal and in human epidermal growth factor receptor 2-positive subtypes. Substantial preclinical data on the oncogenic properties of these mutations have been reported. However, whilst the preclinical data have clearly shown an association with robust activation of the pathway and resistance to common therapies used in breast cancer, the clinical data reported up to now do not support that the PIK3CA mutated genotype is associated with high levels of pathway activation or with a poor prognosis. We speculate that this may be due to the minimal use of transgenic mice models thus far. In this review, we discuss both the preclinical and clinical data associated with PIK3CA mutations and their potential implications. Prospective clinical trials stratifying by PIK3CA genotype will be necessary to determine if the mutation also predicts for increased sensitivity to agents targeting the phosphoinositide 3-kinase pathway. © 2014 Zardavas et al.; licensee BioMed Central Ltd. Source
News Article | March 24, 2016
Tucked behind a rotating display of the 2017 Honda Civic Hatchback prototype at the New York International Auto Show on Wednesday was the automaker's Clarity Fuel Cell sedan. It was just sitting there without any media members flocking to it. In a way, the moment symbolizes everything Honda doesn't want to see transpire once the Fuel Cell sedan hits the production line in Japan, before being made available in California later this year. Honda has done its part, researching, developing and improving hydrogen-powered vehicles for nearly 15 years, dating back to the introduction of its first-generation fuel cell vehicle, the FCX, in 2002. But now, if the greener technology is ever going to be adopted on a wider scale, Honda will need states, regulators and companies to do their part, too. The reason Honda will be strategically retailing the vehicle to customers in select California markets this year is because only certain locations in the Golden State are equipped to sustain the fuel cell stack once its driving range exceeds the Clarity's touted 300 miles-plus. As of mid-June 2015, the United States Department of Energy's Alternative Fuels Data Center counted only 23 hydrogen fueling stations in the U.S., with a majority of them being in California. As Honda spokesperson Chris Martin explained the fuel cell stack technology to us at the NYIAS, he couldn't help but stress the importance of more public hydrogen fueling stations popping up across the country to establish the sustainability that would enable the Clarity to truly thrive in the future. Until then, we would hate to see the Clarity Fuel Cell sedan sit on the production room floor with nobody paying it any mind. It's too good a technology to be wasted. Since the FCX in 2002, Honda has really ramped up its hydrogen-powered technology, launching its second-generation fuel cell vehicle, the FCX Clarity, in July 2008. That vehicle had a V-flow fuel cell stack positioned in its center tunnel, providing 240 miles of driving range. The automaker improved that with the new Clarity, whose fuel cell stack is 33 percent more compact with a 60 percent increase in power density in comparison with that of the FCX Clarity. That and Honda figured out a way to fit the entire fuel cell stack under the hood of the car, as we saw up close at NYIAS, freeing up room for a more spacious cabin and capacity for a fifth passenger to be seated comfortably. In addition, the new Clarity's range of 300-plus miles gives it at least 60 more miles than its predecessor. It would be a shame if the Clarity becomes a hit in California, only for Honda to be forced to sit tight with the technology because there aren't enough sustainable hydrogen-fueling stations out there.
News Article | August 22, 2016
A moment of inspiration during a wiring diagram review has saved more than $2 million in material and labor costs for the Trinity supercomputer at Los Alamos National Laboratory. The Laboratory’s High Performance Computing (HPC) facilities team, led by Data Center Manager Ron Velarde of HPC-Design (HPC-DES). discovered the potential to re-engineer Cray’s initial wiring diagram for the power feed to Trinity’s computing racks. As a result, the facilities team was able to perform a redesign, approved by Cray and the Laboratory, that ultimately saved the Lab an estimated $2.6 M in material and labor costs. Sandia National Laboratories adopted this redesign for their Trinity-like systems and Lawrence Berkeley National Laboratory is considering it as well. “My team and I are always looking to improve engineering means and methods for supercomputer installations in our data centers,” said Velarde. Trinity, Los Alamos’s latest major high-performance computer, is designed to provide increased computational capability for the NNSA Nuclear Security Enterprise in order to improve geometric and physics accuracy in calculations that can be completed in weeks — not years. The facilities team also used prefabricated copper tray cable rather than fabricating the 22,000 feet of cables on site, which was the plan originally proposed for the project. For high-performance computing systems in the Laboratory Data and Communications Center (LDCC), where power cables are being routed over the computers for the first time, the facilities team worked with the Lab’s electrical standards personnel to approve new aluminum power cables, saving 20 percent in materials cost and a factor of three in weight. This pioneering use of aluminum is also expected to generate cost and weight savings in other areas of the Laboratory in addition to the high-performance computing data centers. The Trinity Phase 1 system, based on the Haswell processor, is currently performing simulations. The Trinity Phase 2 system, based on the Knight’s Landing processor, is being delivered this summer. As an example of the scales involved, Phase 2 of the Trinity system, which began to arrive in June, weighs just under 100 tons. Trinity in its completed form will weigh about 175 tons and contain twelve miles of copper cable and 44 miles of optical cable. More 16,000 gallons of water will circulate in the inner process loop to cool Trinity at a flow rate of around 10,000 gallons per minute.
News Article | April 11, 2016
The landscape of programming languages is dynamic and bewilderingly complex, with literally hundreds of languages in use right now across the world. Each has its own pros and cons, its own quirks, its own trajectory of waning or waxing popularity. And each is buoyed by its own community of programmers, bound into a subculture of sorts. In this way, we can say a language, and the community supporting it, has a “character.” Motherboard talked to David Stewart, an Intel® Software Evangelist, open-source software expert and manager in the Intel Data Center Software Technology group, about a widely accessible programming language, Python. It’s notable for its accessibility to neophyte programmers, and the civil, inclusive tenor of discourse within its community, particularly towards women. The popularity of Python has grown steadily in recent years. It’s now among the most common “first languages” taught to novice programmers. Recently, several women-lead Python groups have emerged, hosting workshops to teach women how to code. Guided by Stewart’s deep understanding of Python, Motherboard had the opportunity to discuss what’s in store for the future. Motherboard: Let’s start from the beginning. Can you tell me what Python is? David Stewart: Python is a programming language. It’s become pretty much ubiquitous in cloud computing. In fact, the biggest projects in open-source cloud computing infrastructure, like OpenStack, are all written in Python. “Pythonic” programming has also become extremely popular with people who want to write in this new area called "DevOps”—a combination of people who are doing software development and system operations. And folks are trying to write code doing system-administration kinds of tasks, too, or tasks that don’t require a ton of programming, that you want to do kind of quickly. If you look at the top computer science schools in the country today, the majority of them are teaching this language Python as a first language. What's great about it from a teaching perspective is it’s extremely friendly, even for non-programmers. You don’t have to spend a lot of time fully describing the data types and stuff like that. You more or less just start programming and the language itself figures out what the type information is and handles that all correctly. MB: When did Python become so popular? DS: I actually have a chart here that goes back to 2006. We subscribe to a service by a company called Evans Data Corporation that tracks the popularity of languages. They show Python today with 7 million programmers. It’s like the fifth most popular programming language in the world. It has gradually grown as the use of languages like C has declined over time. My first exposure with Python was back in 2010, when I took over a project that would build the whole Linux operating system on your desktop in about an hour, which is sort of stunning. This thing is downloading software source code from about a thousand projects all over the internet, and then configures those sources and patches it, compiles it, and turns it into a complete Linux operating system. It does this in about an hour. And the whole thing is written in Python. I'm like, "What? You’ve got to be kidding me." It has some incredibly advanced features, even though it’s "Pythonic" and easy and accessible. It’s very quick to program in. When it comes down to it, people are very interested in saving time. MB: I understand Python has a reputation for being friendly to women. What’s going on here? DS: Of course, it’s kind of ironic to have a man talk about a topic like this. I've been doing work with open-source for a long, long time. A lot of the communities, unfortunately, are kind of unfriendly to women. It’s my observation that the interaction between members of the Python community are much more civil and constructive than in the Linux community, in particular. The latest Python conference I went to was in Europe last July. It’s called Euro Python. One of the things that was really interesting about it was that the first keynote was actually given by a couple of women who had in the previous year created a nonprofit organization called Django Girls. These are a pair of women who work at a company in the UK, who are both from Poland and both have the first name Ola. The way they gave the keynote was fantastic. It was told as a fable about a little squirrel; she was very interested in technology but felt excluded by the other animals in the forest. This fable was told with a bunch of watercolor drawings that one of the Olas had actually created. In their first year of operation, Django Girls held 100 workshops all over the world teaching women how to develop websites in Python. Actually, my daughter is 24-years-old, and she’s been trying to figure out what she wants to do with her life. I suggested she go to one of these Django Girls conferences, because it’s a workshop all about hands-on tutorials helping women learn how to program. MB: What is it about Python that makes it friendly to women? Is it something about the actual language itself? Or is it more of a subcultural thing within the community? DS: One thing that I think causes this is the founder of the Python project, a guy named Guido van Rossum. He’s referred to as the “BDFL”—the Benevolent Dictator for Life. The way to think of him is like Linus Torvalds of Linux. Most of his keynote at that conference was answering questions from the people who had attended. And he actually said, “Let's alternate between men and women asking questions.”On the second day of the conference, he was wearing a shirt from PyLadies, another nonprofit like Django Girls that helps women learn how to program on Python. I’ve talked to Guido about this. I commented on how accessible it seems to be to women, and he admitted this is a high priority for himself. And Guido, by the way, is highly respected—I would say probably loved by the members of the community. He sets the tone. It may also be due to the fact that the style of the Python language itself is much more accessible. People talk about coding that is "Pythonic." As they develop the language they’re really looking to preserve this "Pythonic" aspect, which tends to make it more accessible and easier to program. Again, my daughter, she didn’t have any programming background going into this Django Girls workshop. She was pretty anxious, worried that maybe she wasn’t smart enough to program. When she came out of it, she knew she could do it. She was able to succeed, and I think it’s a huge confidence boost. MB: So what kind of Python-related work are you involved in? DS: My interest here at Intel is, that since so much software is now being written in Python, we want to make sure Intel has the best chip to run it on. So we're doing a lot of work to really dig deep into the language, and find ways that we can increase performance. I'm very interested in PyPy, a Python project that can accelerate the performance a lot. Usually when you're working on servers, and you're looking to improve performance, you have some massive server system you’re trying to make faster. A lot of times we're improving a language a few percent here and there. There’s a project called Swift, it's part of OpenStack. It’s an open-source project for doing something like object storage, and it’s written in Python. So we started looking at the performance of this thing, and started chipping away a few percentages at a time. And eventually we decided to try this PyPy implementation, because it uses a technique we found extremely useful in Java and other languages. So we used this version of Python called PyPy and suddenly we got a 2x performance improvement in Swift. It’s like a 111% throughput improvement, and 86% response time improvement. We’re going “holy cow, this is fantastic.” Now we're starting to dig into this and see if there's some way we can make this implementation more mainstream. We're also finding that a lot of people are using Python for this whole area of machine learning, and analytics and technical computing. This is a huge deal as well. So if someone writes a new algorithm for, let’s say, facial recognition, and they're writing it in Python, we'd like to give them a big acceleration in performance. So Intel is actually creating a Python distribution. We're creating a Python product ourselves, too. This is a different part of Intel than I’m involved in, but they're actually creating an Intel-accelerated Python product. But we want to make it easy no matter if you want to stick with the open-source version or the product that we're providing. We're very happy whatever a customer might choose. MB: Alright, so Python is popular, fast and easy to use. It sounds pretty sweet. Is this the programming language of the future? DS: “Always the future is changing,” Yoda said. As I look at this chart of trends from 2006 to 2015, it’s definitely been trending up consistently, at least among open-source projects. This whole language thing is very dynamic. Back when I was a young software engineer in the ‘80s, there was a very famous computer scientist named Brian Reid who I remember giving a talk that basically claimed that the world does not need another general purpose language. Maybe we need some special purpose languages, but not another general purpose one. He was kind of closing the patent office. And since that time, my gosh, all the popular languages I’m looking at have been invented. I think it’s an incredibly surprising area of innovation and development, and it continues to be something people are experimenting with. I’d say Python is the language of today. Is it also the language of the future? It will be if all these students come out of college knowing Python and using it. But Apple's developing languages. Google’s developing languages. Microsoft’s developing languages. They all have their strengths and weaknesses. We're trying to see where the trends are going, where people are spending their time, and making sure Intel is the best chip to run any of this stuff on. And we're trying to innovate new ideas, and suggest some language features, without trying to rip away control from anyone else. I think we're more in the role of trying to come beside them and help out, and be an asset to those communities. To learn more about Intel® Software Evangelists, please visit evangelists.intel.com
News Article | January 25, 2016
In a warehouse basement in Brooklyn’s Red Hook neighborhood late last year, a handful of self-taught network engineers gathered to casually discuss how they might make Time Warner Cable irrelevant in their lives. Toppling—or at least subverting—a telecom monopoly is the dream of many an American, who are fed up with bait-and-switch advertising campaigns, arbitrary data caps, attacks on net neutrality, overzealous political lobbying, lackluster customer service, and passive-aggressive service cancellation experiences that are a common experience of simply being a broadband internet customer these days. The folks at NYC Mesh are actually doing something about it. Motherboard made a podcast about NYC Mesh earlier this month. You can subscribe to Radio Motherboard on iTunes. On any given weekend, Brian Hall and his fellow organizers can be found around the city, installing directional wifi routers on rooftops. Anyone in the city who lives near another person on the network is welcome to join, and NYC Mesh volunteers will help you install a rooftop router. “Two weeks ago, Jonathan decided he wanted to put a router in his backyard,” Hall told me, referring to one of his organizers. “We were initially going to put it on the top of a telegraph pole. But we couldn’t figure out how to get up there, so he started climbing a tree, it was so far up. There’s now a router at the top of a tree in Park Slope.” The DIY network relies on “mesh” routing. The concept is quite simple. Your home wifi router provides internet to anyone with a wifi-capable device in your home. But routers also have the ability to connect to and talk to each other. By “meshing” them, or connecting them together, you are creating a larger wifi zone. As long as one of the routers is connected to the internet in some way, it’s possible for anyone within range of any of the routers to get onto the internet. You could, theoretically, connect many routers together to create a giant wifi hotspot that covers huge parts of New York City or any other geographical area. Of course, it’s not quite that simple. Every time a connection “hops,” it slows down. If your router is connected to the internet but your neighbor’s is not, your internet speed will be faster than his or hers. The more hops, the more latency. This means that, for a mesh connection to be a viable alternative to a standard one, many of the routers must be connected to the internet. Currently, NYC Mesh has about 40 “nodes,” or routers on the network (they are mostly smattered around Manhattan’s East Village and Williamsburg in Brooklyn), with more than 100 people waiting for an install. The network is also capable of “meshing” through the internet, meaning that if two routers are independently connected to the internet, they can also talk to each other to be part of the same network, which is why there are various pockets of meshes in the NYC network. NYC Mesh explains its mission in this presentation. What NYC Mesh has accomplished so far is interesting. It’s a fun hobby project, a cool proof-of-concept, and an interesting experiment in what’s possible with the cooperation of a handful of strangers. But all of the nodes are eventually routed through a Time Warner Cable internet connection, which doesn’t do much good if you’re trying to create what could eventually be an alternative to Time Warner. It has to become more serious. And so Hall has arranged agreements with two massive internet exchanges in New York City that will allow him to install two “super nodes” that have ranges of several miles in March of this year. Super nodes are essentially high-powered wireless transmitters that are connected to an internet exchange, which can be thought of as the “backbone of the internet.” In many cases, normal internet service providers buy or lease their bandwidth from these internet exchanges, which have names the layperson probably hasn’t heard of. The first super node will be at the Sabey Data Center, which is on the Manhattan side of the Brooklyn Bridge. The second will be at 325 Hudson, which bills itself as a “carrier-neutral interconnection facility” that “offers direct access to transatlantic cables, major metro and regional fiber providers, and peering exchanges.” “They would be connecting us directly into the backbone of the internet,” Hall said. “We’ve met a lot of helpful people in high places willing to donate bandwidth and to give us cheap space on a couple internet exchange places in downtown Manhattan. It’s quite exciting because we’ll get cheap internet and can broadcast it to Brooklyn and downtown Manhattan.” The importance of these super nodes to the future of the network can’t really be understated: They have a range of several miles and are connected directly into the backbone of the internet, which allows NYC Mesh to become its own internet service provider. Routers can also connect directly to the super node, meaning that people who want to join it will no longer need to live within a few hundred feet of the nearest node. Instead, everyone who lives in lower Manhattan or along the Brooklyn waterfront should be able to connect directly to the network. Hall says that routers connected to the super node will be able to get download speeds of more than 100 mbps, which is faster than many Time Warner Cable connection options. A $95 router on your rooftop would be able to connect to the super node, and Hall says he’s planning on charging people a modest maintenance fee (“we’re thinking like $10 a month or something”) to join. The super node will initially be funded by businesses in lower Manhattan who are hoping to create redundancies should their internet service go down during a storm, for instance. There are still a lot of moving parts, and neither of the nodes has been installed yet, but experts tell me there’s no reason why NYC Mesh can’t succeed. Guifi.net is believed to be the largest mesh network in the world—it’s an autonomous, free, community-owned network that has roughly 30,000 nodes located all over Spain. In Cuba, an illegal mesh network called SNet has flourished, allowing people there to play games and exchange files on a local network (Cuba’s access to the “real” internet is severely limited by the government). Isaac Wilder and the Free Network Foundation, meanwhile, have open sourced tools to start small networks around the US. Jinyang Li, a New York University professor who helped develop early mesh networking technology, told me there simply hasn’t been much interest in mesh networking in the United States so far. “Mesh networks have been the only choice for people in a lot of developing economies,” Li said. “I think in the United States the infrastructure has been so good, so there hasn’t been incentive to build these networks, but there’s no reason it wouldn’t work.” As people become more and more fed up with monopolistic ISPs, there’s growing sentiment that there must be another way. “Everyone seems to hate Time Warner, that’s the thing that unifies the city,” Hall said. “It’s going to be a while before we replace Time Warner, but there’s some hope of it happening.”