John Abbott covers systems, storage and software infrastructure topics for 451 Research, a part of S&P Global Market Intelligence. Over a career that spans more than 30 years, he has pioneered specialist technology coverage in such areas as Unix, supercomputing, system architecture, software development and storage.
As one of the co-founders of The 451 Group in October 1999, John ran analyst operations from the company's San Francisco office. He has been a principal author on many 451 Research Special Reports, including those on storage virtualization and blade servers – the first comprehensive surveys of either subject to be published. More recently John has focused on topics such as converged infrastructure, new systems architectures, AI and deep learning accelerators. He helped establish 4SIGHT, the 451 Research framework for the forward-looking, long-term coverage of emerging technologies.
John began covering the technology sector in 1984, building on his previous experience as a technical author and direct involvement using mainframes, early PCs and Unix workstations. As a freelance journalist, he contributed to publications including Computing, Computer Weekly, The Financial Times and The Times. In 1987, he was appointed editor of ComputerWire's weekly Unix newsletter, Unigram.X, and later became editor of the company's daily Computergram International service, first in London and subsequently in San Francisco. He established the 451 Research office in San Francisco and lived there for over a decade.
John studied music at the University of Keele and has an MA in modern English literature from the University of London.
The reductions apply to AWS's P2 and P3 GPU-based instances for all SageMaker components, including hosting, training, processing, notebooks and batch transform. The goal is to reduce time to value when hosting the entire machine learning pipeline on AWS.
Backed by automotive giants such as Daimler and Denso, the chip startup is focusing on plug-and-go ease of deployment with small form-factor cards and modules. Blaize has launched its first software development tools, including the Picasso SDK and the code-free AI Studio, and is specifically targeting AI at the edge with its dataflow processor.
The US Department of Defense has embraced the concept of disaggregated composable infrastructure to dynamically mix the proportions of CPU, GPU, memory and storage for AI workloads. It recently awarded over $52m in contracts to composable infrastructure startup Liqid, which has just 60 employees.
A look back at the vendor's acquisition targets over the past three years reveals its traditional preoccupations. Most of its deals involve process-specific simulation tools than can be slotted into its existing platform to extend functionality and market reach. In the future, Altair could explore a larger simulation purchase.
The compiler specialist's goal is to help AI accelerator vendors gain design wins by establishing an open standards-based cross-architecture software layer and ecosystem. Current software options for AI and neural network workloads are typically tied to a single architecture, and often support only a limited number of graph compilers.
The GPU maker plans to maintain Arm's vendor-independent business licensing model while investing in its embedded 'edge of AI' strategy. Our data shows that the deal, which also includes a substantial earnout, stands as the third-largest tech transaction since 2002.
New Alert Set
"My Alert"
Failed to Set Alert
"My Alert"
Here are a few tips for navigating our Research.
Quickly access any of our research and data products.
Browse the research & data that aligns with your specific interest.
Discover new and more flexible ways of finding relevant content.
Set notifications for the topics & companies you care about most.
Access any page in the dashboard from here.
Hand picked content for you and your team.
