April 03, 2013
DALLAS, Texas, April 3 — Dynastack Inc., a Dallas based cloud ecosystems and IT services provider, today announced their internal benchmark testing results and plans for the rollout of their OpenStack public cloud.
Based on their initial benchmark results it was found that OpenStack Compute instances can spawn in as little as 10 seconds for smaller instances and consistently under 20 seconds for larger instances. Though the numbers are internal, Dynastack’s results are still 40 seconds faster than Rackspace Cloud, AWS and other competing products.
“We’re still looking at different vendors for our final InfiniBand implementation at the Dallas datacenter. But after our initial results, we’re really excited with the bonded 100 Gbps InfiniBand FDR interconnect results. Once we’ve refined it I expect the benchmarking to be even faster than what we’re looking at today”, said Dynastack CTO, Stephen Schuresko. “We expect to have InfiniBand fully implemented by the end of Q3 or early Q4 of this year.”
Over the last 4 years InfiniBand, with its lower latency and higher density, has outpaced traditional Gigabit Ethernet interconnects and become the de facto interconnect standard for datacenters and high-performance computing IaaS providers. The technology is proven and the adoption rate is bearing that out; so, it makes sense that Dynastack is building around InfiniBand technology.
“This is just the first stage for us in rolling out our HPC and GPU Cluster solutions. And truly, this is our vision for the super-computing inspired cloud ecosystems that we will offer to all of our customers,” Dynastack CTO Stephen continued. “We set out from the very beginning to make Dynastack the fastest, highest performing OpenStack offering available. Our numbers already show that and I really can’t wait for the independent testing to start in order to prove it.”
As applications continue to scale and distribute it is becoming increasingly more important that the underlying network keeps up with the processing power available. This network bottleneck is gradually shifting but latency problems still exist, though now only in the microseconds, it still lags well behind CPU speeds. But it’s not only about speed it’s also about power and carbon footprint.
“100% of our datacenters run on SuperMicro MicroCloud servers. So our platform isn’t just the highest performing cloud it’s also green. We can fit 20 times more servers in a 42U rack than we can with traditional hardware providers like Dell or HP and we still use less electricity,” said Dynastack CEO, Jared Rice. “This is very important for how we envision the data center utility model shifting in the future, which is simply another challenge that we will meet head on. We will continue to invest in innovation and we will also give back to the communities around us, not just Dallas, but also the open source community with powerful OpenStack contributions. Our goal is to change the technology and communications playing field forever. We will be disruptive and a force to be reckoned with.”
Dynastack offers a wide array of hosting products encompassing everything from traditional shared hosting to high-performance cloud computing (HPC) services. Their scheduled formal launch is April 8th though they are offering advance accounts for select users.
About Dynastack, Inc.
Dynastack is a Dallas based cloud ecosystems and IT services provider that focuses on helping their customers coevolve and support their business ecology with the most innovative technologies available.
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?