Industry analysis examines FPGA deployment trends in datacenters as alternative and complementary acceleration to GPU computing.

FPGAs from Intel and Xilinx have been steadily carving out niches in datacenter applications where low power, high performance, and configurability may trump programming challenges. AWS and Microsoft have made significant strides in bringing FPGA solutions to market. Xilinx and Amazon Web Services have been working with solution providers to create shrink-wrapped applications and tools which use AWS F1 FPGA instances, and Microsoft has recently announced some pretty stellar results in its Project BrainWave AI program using Intel (Altera) FPGAs. Nearly a year prior, the author noted that AWS needed to go from 3 solutions to 30 to convince the market of real demand, and then to 100 to have a material impact on the market. At that time of writing, AWS is now at 20 Amazon Marketplace Instances (AMIs).

Xilinx provides FPGAs to Amazon, which can be configured in Amazon F1 elastic cloud instances to accelerate a wide variety of workloads. However, using this technology traditionally takes significant work by the few engineers in the industry who are skilled in both hardware and software design. To address this barrier to adoption, Amazon and Xilinx in 2017 initially worked with solution providers Edico Genome, NGCodec, and RYFT to deliver genomics, image processing, and complex analytic solutions, respectively, as a shrink-wrapped solution using F1 AMIs.

Looking at the 20 AMIs now available on AWS F1 hardware, roughly half of these AMIs are FPGA development tools, and half are quasi-applications, mostly APIs which can be called from applications. The growth from 4 AMIs to 20 in the last 9 months is a milestone and indicates interest at least from the application supply-side of the equation. There has been good progress in support for Machine Learning inference tools, with five AMIs providing tools ranging from DeePhi's speech recognition app to optimized tools for Apache Spark. Notably, Xilinx announced that it has acquired the Chinese company DeePhi for an undisclosed amount, signaling its investment in deep learning capabilities.

GPUs like NVIDIA's Volta TensorCore V100 have quickly become the gold standard for training deep neural networks in every cloud datacenter, including Microsoft, but the job of using the resulting DNNs is up for grabs for GPUs, FPGAs, CPUs, and custom chips like Google's TPU. Microsoft has demonstrated its intent to use FPGAs here, and Amazon has continued to ramp up its portfolio of Xilinx-powered FPGA solutions for a variety of apps. What remains unclear is whether and when FPGAs might cross the chasm and reach meaningful volume in the data center, with more cloud providers jumping on board for internal apps and making them externally available for accelerated solutions. However, given Microsoft's impressive results with Project BrainWave, FPGAs' day appears to be on the horizon, even as it can take years to produce an industrial-grade solution built on FPGAs.