Appro
| Company type | Private |
|---|---|
| Industry | Supercomputing, High Performance Computing |
| Founded | 1991 |
| Founder | Daniel Kim |
| Defunct | 2012 |
| Fate | Acquired by Cray Inc. |
| Successor | Cray Inc. |
| Headquarters | Milpitas, California , United States |
Area served | Worldwide |
Key people | Daniel Kim (President and CEO) |
| Products | Supercomputers, HPC cluster solutions, servers |
| Parent | Cray Inc. (from 2012) |
Appro specialized in creating supercomputing solutions tailored for the High Performance Computing (HPC) sector, focusing on medium to large-scale system implementations. Its corporate headquarters were located in Milpitas, California, with a computing facility based in The Woodlands, Texas. Additionally, the company operated manufacturing and support branches in South Korea and Japan.
In November 2012, Cray Inc. completed the acquisition of Appro.
Company history
Established in 1991 by Daniel Kim, who served as President and CEO, Appro initially grew as an OEM producer of high-performance rackmount computer systems. The company quickly became recognized as one of the fastest-growing suppliers within the rackmount server market.
Transitioning in 2002 from OEM manufacturing, Appro began marketing its own branded high-density servers aimed at HPC and Internet computing sectors.
By 2005, Appro had broadened its manufacturing, engineering, and product development operations in South Korea. That year marked the introduction of its first blade cluster, the Appro HyperBlade Server Cluster Solution, alongside the Appro 1142H—a pioneering 1U server featuring four processors designed specifically for HPC applications.
In 2006, Appro released the Xtreme-X Supercomputer powered by x86 processors. This flagship model combined robust performance with high availability tailored for scalable data center environments and incorporated the Appro Cluster Engine (ACE) Management Software to facilitate remote administration.
Between 2008 and 2009, the company launched the HyperPower Cluster, a hybrid CPU/GPU system utilizing NVIDIA Tesla GPUs and CUDA technology. This period also saw the introduction of professional services and cluster management software, as well as the energy-conscious Appro GreenBlade System.
In 2011, Appro debuted an advanced version of its Xtreme-X Supercomputer that integrated cutting-edge processing units and accelerator technologies. Enhancements included improved performance efficiency per watt, multiple high-speed networking choices, and specialized professional services. This system was designed to meet diverse HPC requirements—such as capacity computing, hybrid configurations, data-intensive tasks, and capability computing—and secured significant contracts with institutions like San Diego Supercomputer Center, Lockheed Martin (Department of Defense), NNSA (Department of Energy), University of Tsukuba, plus national laboratories including Lawrence Livermore, Sandia, and Los Alamos.
On November 9, 2012, Cray Inc. announced its intention to acquire Appro for $21.8 million to strengthen its HPC cluster portfolio. The transaction was finalized on November 21, 2012.
Customers and Supercomputing Projects
- National Oceanic and Atmospheric Administration (NOAA) – Delivered a 113TF supercomputer supporting the Hurricane Forecast Improvement Project (HFIP) in 2012
- Virginia Tech, Advanced Research Computing (ARC) – Developed the Blue-Ridge Supercomputer (2011) featuring Intel Xeon Phi coprocessors
- Kyoto University – Implemented a 210TF supercomputer for the Academic Center for Computing and Media Studies (ACCMS) in 2011
- San Diego Supercomputer Center (SDSC) – Supplied the Gordon Supercomputer with 35M IOPS (2011), Trestles at 100TF capacity (2010), and DASH cluster (2009)
- Tri-lab Linux Capacity Cluster (TLCC) Projects for NNSA – Supported TLCC2 (2011) and TLCC07 (2007) deployments at Lawrence Livermore, Sandia, and Los Alamos National Laboratories
- Los Alamos National Laboratory – Provided the Mustang supercomputer delivering 353TF performance in 2011
- University of Tsukuba – Installed the HA-PACS supercomputer rated at 802TF (2011) and a 95TF system at the Center for Computational Sciences (2009)
- Lockheed Martin – Supplied a 147.5TF supercomputer backing the DoD High Performance Computing Modernization Program (HPCMP)
- Sandia National Laboratory – Established an Intel MIC Testbed Experimental Cluster in 2011
- Idaho National Laboratory – Delivered the Fission supercomputer in 2011
- Lawrence Livermore National Laboratory – Supported multiple projects including Hyperion (2010), a 110TF Graph supercomputer (2009), and several Peloton Project clusters such as Rhea, Zeus, Minos, and Atlas totaling 222TF capacity (2007)
- Renault Formula 1 Team – Provided a 38TF supercomputer for its Computational Fluid Dynamics Center located in the UK (2009)
Awards and honors
- In November 2012, the Appro Xtreme-X supercomputer, known as Beacon, achieved the top position on the Green500 List for energy efficiency. This system, implemented by the National Institute for Computational Sciences at the University of Tennessee, incorporates Appro's GreenBlade GB824M platform equipped with Intel Xeon E5-2670 CPUs and Intel Xeon Phi 5110P accelerators.
- The HPCwire Editors' Choice Award in 2012 honored Appro Xtreme-X with accolades for "Best HPC Cluster Solution" and "Best HPC Storage Product," recognizing its partnership with SDSC on the Data Oasis Storage System utilized by the Gordon Supercomputer.
- Appro secured the fourth spot among vendors on the November 2012 Top500 list, featuring 24 of its supercomputers; the highest-ranking machine was Zin at Lawrence Livermore National Laboratory, positioned at number 29.
- During the 2012 Intel Solution Summit, Appro received the Datacenter Solution Innovation Award in acknowledgment of its GreenBlade and Xtreme-X innovations.
- In 2011, HPCwire's Reader and Editor Choice acknowledged Appro as one of the "Top 5 Vendors to Watch" and awarded the company "Best HPC Server Product" for its Xtreme-X system.