 do in a long weekend with enough free food and clouds really really amazing what they're doing there I think that research overall is a really interesting area there's so many applications for science and other forms of research using open stack and other cloud technologies and we're gonna hear in a little bit from the Texas Association of Computing Center in Austin but first we're gonna see a quick video and learn a little bit more about what they are about how can we use the cloud produce new scientific results how can we take apart the cloud and understand how to do it better our mission at TAC and it's a great mission to have we seek out problems that need large-scale computing or data approaches to solve them and enable them to do their innovations faster I'm Dan Stanzian and I'm the executive director here at the Texas Advanced Computing Center at UT Austin so we build large-scale computing platforms to promote any field that's impacted by data which is almost all of them so we do large-scale supercomputers we do cloud systems visualization systems large-scale data storage systems machine learning systems all kinds of platforms it's always gratifying to see somebody who's running a hundred or a thousand or a million times faster than they used to because of the platforms we put out there there are examples that people in Arizona basically they are solving some of the interesting stuff in the space you know with the open stack all the way to you know scientists that they are trying to build the big data systems around the brain health to all the machine learning algorithm that humanity now start looking to do prediction of things that in the past we were not capable through our funding from the National Science Foundation TAC is a national resource and we're open to any US researcher and foreign collaborators of US researchers we have about 350 institutions spans about 2,500 different funded projects and then also partners from several dozen countries around the world our scientists are smart people they come tap into the basically this massive scale computation that Dan created here so they could solve the next humanity is basically problem hopefully they could solve cancer here so we have two large-scale projects that started using open stack both in about the past year we have one silo in jet stream where we sort of fix the configuration and let the users run their science and a separate system in chameleon where users can come in and reconfigure the system in the way that they want to do basic computing research so with jet stream and chameleon both coming online there's a whole new range of users we can support that we couldn't support in our old model the amount of data that we are seeing is driving our economy it's the beginning of humanity moving from the industrial revolution to informatic revolution which is data is the base of it computation is base of it you know and hopefully we see you know the cloud getting bigger and bigger and bigger so they start in pushing the science to the next step it's pretty incredible to think about what people are doing when it comes to open stack and science and to tell us more about it the star of this video we have the executive director from the Texas Advanced Computing Center I get it right this time from the UT Austin right here in Austin Texas Dan Stanzione welcome Dan I appreciate you making the travel in from the other side of town you've been through Austin traffic that's not as trivial as you might think that's a fair point we need to get those sensors out here in our city make it a little smarter okay so jet stream and chameleon I heard that in the video so what are those exactly so there are two big systems and funded projects we have through the National Science Foundation jet stream is a partnership between Indiana University and TAC and a number of other partners to build a scientific cloud sort of is a peer to our HPC resources that we've always provided to the open science research community so as we move from just pure simulation to sort of other modes of computing that people want to do in science we built jet stream to sort of serve that need chameleon on the other hand is to deal with all the other users that we normally leave out of those projects so chameleon is led by the University of Chicago and Argonne National Labs along with TAC in Ohio State and Northwestern and UT San Antonio and other partners who will be mad at me later for leaving their name out so but that system is more focused on how do we build the cloud in our traditional HPC systems we run big simulations for people designing aircraft or doing genomics work or you know large scale physics we don't really like people messing with the kernel or the disk scheduling algorithm or any of those things so chameleon is a platform to do that it's to serve the computer science research community for people who want to do things from bare metal up and really understand how to construct systems and OpenStack is instrumental in how we run both of those platforms so a lot of the hackers here would probably like to get on chameleon and like really really do some crazy stuff so in terms of some of the science work you know what are people running on on these clouds like Jetstream yeah so we have a lot of users on both systems at the moment chameleon came first it's been up for six or eight months now I'm in production we have about 700 users we have around 150 projects on that machine today there's all sorts of things from people in Arkansas trying to do different classifications of kinds of cybersecurity attacks to researchers in Pennsylvania who are looking at sort of the what's the sort of theoretical minimum for overhead of container and virtualization technologies so and there's a number of other orchestration for things like how the Large Hadron Collider can run large numbers of jobs there's just really a wide range of sort of development applications on chameleon Jetstream is more focused on the end users in science it's just going into production in the next couple weeks there's about 300 users on it in the early user period now some of the early science work there's a lot of there's psychology there's computer science there's biology a lot of sort of eco bio lately from management of endangered fish species to workflows to understand the evolution of birds to I think the latest and sort of most oddball one was a little app to determine what kind of snake just bit you okay we are in Texas so that could be a big hit here for some of the people probably a rattlesnake that's true spoiler alert it was a rattlesnake okay so obviously there's a lot of research going on where's the funding come from to do all this awesome stuff so for both those projects and for a lot of the projects we have attack it's largely from the National Science Foundation so the federal government through NSF and a number of other agencies funds a lot of research at hundreds of universities and for tens of thousands of projects around the country and so they have a little bit of a mandate to provide infrastructure to do that as well so just as in the past we've bought telescopes and genome sequencers and particle accelerators and traditional supercomputers now we have funding through group called the division of advanced cyber infrastructure to build the national cyber set of instruments that we need and so all the funding is coming primarily from there wow well I think that's a good use of taxpayer dollars in my opinion is actually do some some open source work and and use that for a lot of different science well thank you so much Dan it's very fascinating what you're doing here in town thanks a lot mark appreciate it