GraphStream's mission is to help organizations gain competitive advantage by deploying advanced data infrastructure that is radically more scalable, simpler to operate, and less expensive than what their competitors are still using.
GraphStream began operations in early 2003. Our business, engineering, and support leadership includes alumni from Silicon Graphics, HP, IBM, and other pioneering suppliers of scalable data infrastructure.
GraphStream is privately held by a group of investors with a focus on organic, sustained long-term growth. From the very beginning, our growth has been 100% funded by operations, and the company has been consistently profitable.
The company got its start building advanced scalable software+hardware infrastructure for scientific and engineering workloads. In GraphStream's first year, we worked together with key building block suppliers Intel, NVIDIA, and Mellanox to produce the world's first commercially integrated scalable platforms with GPU co-processors in every node, distributed storage, and InfiniBand networking. We built many systems based on this architecture, including a number of installations that were the largest of their type in the world at time of deployment.
The image below shows part of an early GraphStream installation at Lawrence Livermore National Laboratory. GraphStream built and tested the complete 257-server, $USD 2million system in our production facility, then delivered it as a set of 18 fully pre-integrated, pre-cabled cabinets, for rapid on-site installation. LLNL used this system to generate visualizations of simulation results transferred via a 256 Gbit/sec multiple-link Ethernet connection from the adjacent BlueGene/L system, at that time #1 on the Top500 list of the world's most powerful supercomputers.
GraphStream designed, built, installed, and supported numerous additional systems of this type for LLNL and for many other clients, including:
After GraphStream's initial success providing scalable data infrastructure for scientific and engineering workloads, we expanded our focus to include infrastructure for the larger and more diverse set of workloads that enterprises typically run.