All the sessions from Transform 2021 are available on-demand now. Watch now.
Hewlett Packard Enterprise (HPE) today announced it has acquired Determined AI, a San Francisco, California-based startup developing an open source platform for building machine learning models. The two companies say the deal will combine Determined AI’s software with HPE’s high-performance computing (HPC) offerings. Terms of the deal weren’t publicly disclosed.
Building and training machine learning models are among the most demanding stages in AI development because researchers must face challenges in HPC. These include setting up and managing parallel workloads and configuring infrastructure that spans compute, storage, fabric, and accelerators. Researchers also need to know how to program, schedule, and train their models to maximize utilization of the infrastructure they’ve set up.
Determined AI, which was cofounded in 2017 by Ameet Talwalkar, Evan Sparks, and Neil Conway, counts among its key contributors Ph.D. students and faculty from the University of California, Berkeley and Carnegie Mellon. The company’s platform helps set up, fine-tune, manage, and share workstations and clusters that run on-premises or in the cloud, as well as speeding up model training via capabilities like accelerator scheduling, advanced hyperparameter optimization, and neural architecture search. (Hyperparameters are variables whose values are used to control the learning process, while neural architecture search is a technique for automating the design of particular AI models.)
Determined AI claims it accelerated AI-guided drug discovery for one customer from three days to three hours. “The Determined AI team is excited to join HPE, who shares our vision to realize the potential of AI,” Sparks said. “Over the last several years, building AI applications has become extremely compute-, data-, and communication-intensive. By combining with HPE’s industry-leading HPC and AI solutions, we can accelerate our mission to build cutting-edge AI applications and significantly expand our customer reach.”
Growth in HPC
With the Determined AI purchase, HPE is chasing after an HPC market that’s becoming red hot. According to IDC, the accelerated AI server segment is expected to grow by 38% each year to reach $18 billion by 2024. And Intersect360 Research notes that the demand for HPC will increase by more than 40% to reach almost $55 billion in revenue by 2024.
In May, HPE expanded its GreenLake hybrid cloud services platform with a new data services cloud console, a suite of infrastructure management services, and HPE Alletra, cloud-native infrastructure for edge-to-cloud data workflows. In December 2020, HPE took the wraps off HPE GreenLake cloud services, a pay-per-use service providing access to fully managed, prebundled services based on HPC systems.
“As we enter the Age of Insight, our customers recognize the need to add machine learning to deliver better and faster answers from their data,” HPE SVP Justin Hotard said in a press release. “AI-powered technologies will play an increasingly critical role in turning data into readily available, actionable information to fuel this new era. Determined AI’s unique open source platform allows machine learning engineers to build models faster and deliver business value sooner without having to worry about the underlying infrastructure. I am pleased to welcome the world-class Determined AI team, who share our vision to make AI more accessible for our customers and users, into the HPE family.”
VentureBeatVentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact. Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:
- up-to-date information on the subjects of interest to you
- our newsletters
- gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
- networking features, and more