In this slidecast, the Radio Free HPC team interviews Fritz Ferstl, CTO of Univa. Topics include Big Data, HPC, and the continuing convergence of both.
While what we think of as traditional HPC may differ greatly from Big Data analytics, that seems to be changing. With a long history in high performance computing and customers in both worlds, Ferstl shares his unique perspective on where the two worlds overlap and where the potential is greatest for synergy in the future.
This has to be our best show yet, so be sure to check it out.
In this follow-up podcast to the GPU Technology Conference, the Radio Free HPC team mulls over a talk by GE’s Dustin Franklin, GPU app specialist. Dustin’s topic was GPU-direct RDMA; was this a first look at real-world RDMA with GPU-to-GPU communications?
Follow along as the guys describe flow charts on technical slides that are not yet approved by viewing for the “great unwashed masses” – but make no mistake, they’re impressed by what they saw. Dan “knows a guy” who can divulge more, and offers to arrange an inquisition with Henry. Henry promised to “be nice,” whatever he means by that. Rich missed this GTC session and several others while “conducting interviews,” whatever he means by that. Dan offers another characterization. And this just in: there’s a great deal of information available on the Internet.
In this podcast, the Radio Free HPC team discusses the recent buzz surrounding FGPAs. After being sidelined by accelerators, they’re increasingly being used in appliances.
Big vendors are talking about FPGAs not only for appliances but for general-purpose systems as performance assists. Are we headed back to the future? The guys discuss the ins and outs of FPGAs and why, in some cases, they could be a huge win for the organizations that implement them. But is the architecture flexible enough? For enterprise and Big Data, perhaps it is. If you need to perform the same algorithms over and over again, FPGAs could be a perfect fit. As with all things tech, there are a few cautionary notes to be sounded. Amassing more and more appliances can lead down a tricky road. Will their use in workload-optimized systems lead to vendor lock-in? Can you really teach an old FPGA new tricks? And can they be weaponized?
Most importantly: how are servers like cattle? Tune in to find out…
Rule #1: You do not talk about Exascale. (Kind of like rule #1 of Fight Club, except the guys keep breaking it.) Why not? Because too many of the people talking about Exascale are having the wrong conversation about it.
What should the conversation be? Should it be about the systems themselves, or about the work that can be done only with those systems — the science that we can’t yet do? Spoiler alert: Dan and Henry disagree on this. But a peaceful vibe reigns once again as they discuss what The Exascale Report calls “The Three Noble Truths” of Exascale, which sounds kind of Zen and cool — as if it was coined by Exascale Samurai.
And finally… is it time to talk about Zetta-scale?
In this podcast, the Radio Free HPC team looks at success factors for technology Startups. Prompted by a recent interview with Sun Microsystems co-founder Andy Bechtolshiem, the discussion centers around lessons learned from Sun’s decline and eventual acquisition by Oracle.
In this podcast, the Radio Free HPC team is still talking about the recently concluded SC12 conference in Salt Lake City. The conversation starts with a short review of Thanksgiving dinner (including disgusting eating noises added in at no additional charge) before moving on to more weighty topics such as Intel’s formal introduction of their Xeon Phi coprocessor, including some performance and price information.
Rich and Henry think that Intel has a strong hand with Phi, but Dan isn’t so sure…
In this wrap-up review of SC12, the Radio Free HPC team discusses the Student Cluster Competition, covering the teams, results, and a discussion of how the competition has evolved over the years and where it should go in the future.
In this podcast, the Radio Free HPC team quits griping about the horrible WiFi at SC12 and moves on to a truly big issue: Are LINPACK and HPCC benchmarks useful? Should they be constantly re-evaluated? And shouldn’t you really test machines on the kinds of workloads they’re designed to run?
The catalyst for this discussion is the Blue Waters system, for which no LINPACK numbers have been submitted. Yes, it’s behind schedule, and sure, they’re busy doing the science… but is it also a shot across the bow? Are they rebelling against industry philosophy? If they are, that’s a good thing, according to Henry – because a system is about what you plan to do with it, not how many flops you can get out of it. Rich agrees: if you get a giant LINPACK number on a system that has reliability issues, and you can’t output real science because all your time and money is invested in brute computation, what good is it? And the industry sectors doing meaningful work – where are their systems on the Top500? They’re not playing this game.
Spoiler alert: Henry agrees with Dan on something. Really. It’s at the 10:00 mark, if you’ve got to see it to believe it. We hardly believed it ourselves.
In this podcast, the Radio Free HPC team regroups after SC12 to discuss an industry trend that was in evidence at the show: vendor consolidation.
Cray just acquired Appro
Intel acquired Qlogic Truescale InfiniBand, Whamcloud, and Cray interconnect IP
IBM bought Platform
Xyratex bought ClusterStor
Hitachi acquired BlueArc
NetApp bought Engenio storage
And so on…
The guys discuss how acquisitions need to be integrated the ‘right’ way and how it’s more than just slapping up new logos on websites and combining slide decks. They also talk about some cautionary tales in the world of tech acquisitions along with some success stories. Dan offers a success story: IBM and Platform. Rich predicts which major vendor will next be swallowed whole; Henry predicts a challenge to Intel’s x86 dominance.
In “SC After Hours” chatter, Dan describes “My Dinner With Henry” in Salt Lake City, and Rich is accused of being the biggest Apple fanboi in all of HPC and, perhaps, the world. Watch for a special cameo appearance by someone who knows all about buyouts: Larry Ellison.
In this podcast, the Radio Free HPC team discusses what they’re expecting to see at this week’s SC12 conference in Salt Lake City. Get the scoop on what’s new, what’s old, and what’s just plain played out.
In this podcast, the Radio Free HPC team takes a look at Glacier, Amazon’s cloud archive and backup offering.
Amazon is pitching Glacier as a solution for customers who don’t need frequent access to their data and can handle retrieval times of several hours. The big enticements are low, low cost — as little as a penny per gigabyte per month — and durability. Dan and Henry weed through each facet of Amazon’s marketing claims and — well — rip each one to shreds. Henry thinks this is aimed at the unsuspecting/unfortunate home or small business consumer, as anyone with technology expertise will run far, far away from Glacier. Dan compares it to the “Roach Motel” of storage: once you’re in, you can never get out. And don’t even get them started on the definition of “durable.”
Viewer tip: keep an eye on the “consecutive hours awake” timer at the bottom of the screen.
In this podcast, the Radio Free HPC team looks at tape storage. Is that parrot completely dead, or is it just resting? Is tape now legacy technology, or is it alive and well? And can’t our data just all go in the cloud anyway? In more ‘legacy’ talk, Dan posits that Henry whittled the first punch cards by hand, and Rich claims that Henry invented the chad. Henry retaliates by claiming to have more hair… a must-see.
In this podcast, the Radio Free HPC team takes a look at the OSCON Open Source Convention, where Tim Tim O’Reilly’s presentation on the “Clothesline Paradox” aptly illustrated the way developers create value. Since many large companies such as Comcast make a living on open-source software, Dan digresses into a string of complaints about his Comcast bill, but Henry and Rich reel in the discussion. Was the Internet created out of generosity, or enlightened self-interest? And we hear again from one of our sponsors: Glade ‘Data Center Edition’ air fresheners.
Power is a major challenge standing in the way of the Exascale computing. While the target is to consume 20 MW or less for an exascale machine, current technology trends will not take us there by 2018. In this podcast, the Radio Free HPC team discusses why this is such a tough challenge, where such a system might need to be hosted, and types of infrastructure that will need to be considered. Along the way, you’ll hear scary “power” music and figure out how this all relates to Mad Max, lasers, unicorns, and Planet of the Apes.