May 28, 2013
While Grid Engine may have deep roots in HPC’s academic open source community, it’s firmly entrenched in an ever-increasing number of large-scale enterprise settings, at least according to the backers of one of its newest supported incarnations.
This business reach isn’t surprising in itself, but what is rather remarkable is that Univa, which forked its own distro in 2011 with hired hands from the original Sun development crew, claims 80 percent of its still-budding Grid Engine revenue is coming from big business instead of Grid Engine’s mainstay home in big labs and universities.
Univa CEO, Gary Tyreman, told us that although they’re not disassociating themselves with HPC and its perceived thick association with academia, he's “not ashamed to say” that they’re out to make a solid business out of the trusty scheduler. One could read between the lines and guess that there’s just enough money in HPC for Grid Engine, and as Tyreman says, the people that truly need the reliability and full capabilities are willing to pay for it.
What they really have with their Grid Engine product—or what italicizes that 80 percent of their bottom line—is more appropriately called “enterprise technical computing,” explains Tyreman.
While there are plenty of universities and labs running Grid Engine, he claims that there are many opportunities for supported Grid Engine for their post-HPC future. What’s really driving their far reach into enterprise, he said, is the use of Grid Engine in “big data” and general business computing (at BP, major airlines, etc.) where the applications are not HPC, but are high-stakes in that if a server goes down, it could cost a great deal of money.
“The considerations are completely different,” he noted, pointing to how scientific computing users and Fortune 500s consider the capabilities and cost/benefits of scheduler choices.
During a conversation last week in the wake of their news about a new partnership to enhance Grid Engine’s role in the European auto manufacturing sector, Tyreman brought up Platform Computing (and LSF in particular) repeatedly—both as the symbol of monetizing workload management and as the opportunity for his own company to step in to do the same.
Since IBM’s acquisition of Platform in late 2011, Tyreman says some of the longstanding Platform LSF customers are looking for a new direction. While he noted that “some will like the IBM methodology and process, others do not," it was clear that they're finding success with shops that are running both Grid Engine and a proprietary piece like LSF, sometimes with equal weight, other times in distributed 70/30% ratios depending on how it makes sense for the applications. He sees the Platform to IBM changeover, along with added opportunities from new classes of big data and business applications as two main streams feeding the growth of their supported Grid Engine.
For instance, in terms of the science + computing partnership mentioned previously, Univa has been able to exploit some of the gaps left following the Platform Computing transition. Their partner is the largest provider of HPC and big data technology services for the European auto industry and has recently sought to move beyond their 15-year relationship with LSF. This helps Univa get a foothold at BMW and other auto manufacturers—adding to their emphasis on enterprise reach.
When asked if this is a tough sell all around (given the transition difficulties, the fact that they’re already using an open source version already—likely for some time) Tyreman says that his engineers are where the value point is. Further, he claims that the features go far beyond native Grid Engine, including cloud-bursting capabilities, advanced reporting and other features that are more aligned with a Platform-like professional offering.
With over 10,000 sites using Grid Engine, Tyreman said Univa has a chance to enjoy more than the double-x growth they’ve seen since 2011. Still, at the recent IBM analyst event, Big Blue finally spelled out some of its plans for the Platform buy, at least according to Addison Snell of Intersect360 Research during our own Weekly Soundbite recording last week.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?