Table of Contents
- Market Categories and Deployment Types
- Key Criteria Comparison
- GigaOm Radar
- Vendor Insights
- Analyst’s Take
- About Anand Joshi
Artificial intelligence has become increasingly important in enterprise applications, where huge data sets need to be ingested and coerced into meaning. Only recently, however, is the performance getting a real hardware boost. In the past few years, many chipsets have appeared that accelerate enterprise AI applications. This report looks at the vendors and chipsets for the enterprise market that were shipping as of 2020. There are a number of vendors that have developed chipsets for AI acceleration that are still in the sampling phase, and therefore not included in this report. Also excluded are chipsets for AI acceleration for the edge market. They will be covered in a separate report.
Intel’s CPU and NVIDIA’s GPU products are the leaders in this new market, and both companies have generated several billion dollars in revenue from AI acceleration chipsets. Both continue to invest heavily and build the software infrastructure necessary to support deployment of AI applications. AMD is a challenger but lags behind in software, and its GPUs do not support the data path required by AI algorithms. However, AMD is investing heavily, and we expect to see rapid progress. Xilinx’s FPGA is another challenger, and its recent acquisition by AMD suggests the possibility of new and exciting innovations.
Heterogeneous computing—in which systems use more than one kind of processor— is being promoted by Intel as a means to accelerate AI, and AMD is working on a similar approach. This might combine GPU, GPU, FPGA, and ASICs on a single chip, rather than each as a discrete chip as they are now. This innovation would provide a variety of paths to choose for AI acceleration in the future.
We included ASICs from Google and Amazon in this report because they are both available, though the user base is smaller than for Intel and NVIDIA chips. We also look at three ASIC startups: Cerebras, Habana, and Graphcore, whose products are also being used by customers. These challengers are leading innovation in AI acceleration, although their software still is somewhat in its infancy compared to the leaders.
This AI Chipset Radar report evaluates the products from these companies in conjunction with the key criteria and evaluation metrics defined in the Key Criteria for AI chipsets report.
How to Read this Report
This GigaOm report is one of a series of documents that helps IT organizations assess competing solutions in the context of well-defined features and criteria. For a fuller understanding consider reviewing the following reports:
Key Criteria report: A detailed market sector analysis that assesses the impact that key product features and criteria have on top-line solution characteristics—such as scalability, performance, and TCO—that drive purchase decisions.
GigaOm Radar report: A forward-looking analysis that plots the relative value and progression of vendor solutions along multiple axes based on strategy and execution. The Radar report includes a breakdown of each vendor’s offering in the sector.
Solution Profile: An in-depth vendor analysis that builds on the framework developed in the Key Criteria and Radar reports to assess a company’s engagement within a technology sector. This analysis includes forward-looking guidance around both strategy and product.