Advanced Computing in the Age of AI | Saturday, July 13, 2024

Intel, Alibaba Demo FPGAs in Cloud 


Chinese online retailer Alibaba is collaborating with Intel Corp. on a pilot program to leverage the chipmaker's Xeon-FPGA platform to accelerate the delivery of cloud-based enterprise applications.

The partners said Alibaba's Aliyun cloud service would incorporate Xeon-based servers and software development tools built around Intel's Arria 10 GX FPGAs. The pilot program is designed to accelerate cloud-based application performance and provide Aliyun customers with an alternative cloud platform for running business applications along with demanding data and scientific workloads.

The cloud-based workload acceleration testing is intended as "alternative to investing in on-premises FPGA infrastructure," the partners said Thursday (March 9). "Adding an FPGA-based acceleration offering means [cloud customers] can access that powerful computing without the cost or requirement of building out their own infrastructure," Jin Li, Alibaba Cloud's senior director, noted in a statement.

Intel (NASDAQ: INTC) and cloud providers such as Amazon Web Services (NASDAQ: AMZN) have been promoting FPGAs as a way to tuning cloud platforms to accelerate diverse workloads running in datacenter, including machine learning, video and data encryption.

Intel further asserts that the advent of new data analytics services along with traditional IT are creating a "compute gap" in datacenters that FPGAs can help plug. The chipmaker maintains that traditional CPU approaches for boosting performance such as frequency scaling have hit a wall. Meanwhile, the shift to multicore processing in the datacenter has not kept pace with workloads demands.

"Multicore processing has diminishing returns in terms of increasing true application performance due to limits in I/O and memory bandwidth," an Intel white paper argues.

Hence, the chipmaker promotes its FPGA technology as a way of accelerating server and storage performance in datacenters via the ability to tweak high-bandwidth, low-latency connections to networks and storage. For storage applications, Intel promotes its FPGAs as alleviating I/O bottlenecks while improving storage network performance.

Alibaba (NYSE: BABA), meanwhile, joins a growing list of cloud providers and hyper-scalers incorporating FPGA technology into their operations. Along with AWS, which introduced a new cloud instance in December that pairs Intel Broadwell processors with up to eight FPGAs, Intel also has been working with Facebook (NASDAQ: FB) to integrate the programmable chip technology.

Intel announced plans last year to deliver libraries based on the combination of its next-generation Xeon processors and Arria 10 GX FPGAs. The FPGA technology comes via Intel's 2015 acquisition of programmable logic vendor Altera Corp.

Intel and Facebook have collaborated on an Open Compute Project framework for datacenter processors that incorporate computer logic in the form of FPGAs and GPUs. As part of its effort to accelerate cloud computing, Amazon also has rolled out an "elastic GPU service" as it and others look for new approaches to handling demanding workloads that are taxing general-purpose processors.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).