But don’t just take our word for it. Read what our customers have to say.
Assoc Prof / Mechanical Eng
Professor-Math & CS
Colorado School of Mines
Brett S. Phinney, PhD
UC-Davis Genome Center
My cluster has been working perfectly since I took delivery of it in April. Since this time, I have been running continuously 24 hours a day/ 7 days a week and I have not had a single problem yet. The performance is excellent — a bit better than I had expected.
Asst Professor/Physics & Astronomy
University of Iowa
Our US CMS Tier-2 site purchased our first compute installment, a 64-node dual CPU, dual-core Opteron cluster from TeamHPC. They were dedicated to installing this system both correctly and quickly, and interfaced effectively with our local technical personnel to make this process an enjoyable success.
Dr. David R. Swanson
Director – Research Computing Facility
University of Nebraska-Lincoln
When I was looking to purchase an Opteron system, the TeamHPC sales staff provided rapid and accurate answers to all of my questions, a very competitive quote and then great communication during the construction of my particular system. Their performance at all stages of this purchase have been top-notch and I am extremely happy with the computers we purchased from them. A very professional organization!
Semmes Distinguished Professor of Chemistry
I was very pleased with my recent cluster purchase. TeamHPC's system experts were responsive to our technical needs while the sales staff was attentive to our budgetary constraints. With TeamHPC's help, our turn-key system came in under budget and on time.
Assoc Professor/Computer Science
We purchased 64 dual processor nodes and TeamHPC was great to work with. They gave us an excellent price, worked with us to design the right system, set it all up efficiently, and put in an enormous effort, at no extra cost, assessing timing issues with some of our codes. They also provided some very helpful education for our students working with the machine.
Izabela Szlufarska, PhD
Dane Morgan, PhD
University of Wisconsin
Howard Hughes Medical Institute/
University of Texas Southwestern Medical Center
Director of Technical Support
ICOP Digital, Inc.
Lone Star Internet, Inc
Even now that the machine is running here, they have continued to answer my questions and provide technical support above my expectations. I would definitely purchase future computing machines from TeamHPC.
Senior Research Scientist
Calspan-UB Research Center, Inc.
Systems and Technologies Solutions
GEMS Lab – SAIC
In searching for a vendor, we found TeamHPC outstanding in their technical understanding of our needs. For example, we asked various vendors for diskless compute nodes. Most ignored that request and quoted us compute nodes with disks.
By contrast, TeamHPC clearly understood why we wanted diskless compute nodes: 1) so we would only have to maintain one operating system image for all compute nodes, 2) because if our application swapped to disk, its performance would be gone anyway and 3) because we did not want to invest money or admin. time in buying disks and replacing failed disks in the compute nodes.
TeamHPC was also very cooperative in providing us the most computing power with the fewest non-essential frills—no expensive commercial job queuing software (we are successfully using the open source clustering software “torque” version 1.1.0p5, saving us $16,000 over a commercial solution), no expensive Myrinet interconnect (Myrinet is nice, but not needed by our application on a 32-node cluster), etc. In other words, TeamHPC bent over backwards to get us the most computational bang/buck.
And sure enough, TeamHPC gave us the lowest price, but without any fear of going with the lowest bidder: Throughout the whole process, TeamHPC displayed outstanding technical knowledge of providing Linux superclusters.
Having won the contract, TeamHPC assembled the computer on their site. We connected remotely and performed our own burn-in testing, ensuring that any early hardware failures would take place before arrival at our site.
When delivery came, it was simply a matter of situating the rack, inserting the 32 pre-assembled nodes, the disk array, and the Ethernet switch, connecting all the Ethernet cables and the power cables, turning it on, and doing some local software setup. We were computing the next day.
Since then, we've found the computer fast and reliable. The software setup done by TeamHPC is so smooth that maintaining the whole cluster is about as hard as maintaining 2 or 3 Linux workstations would be.
Air Force Research Laboratory