LegUp Computing Inc., a UTEST company from 2015 announced today that it closed a seed funding round led by Intel Capital.
LegUp offers a cloud platform that enables software developers to program, deploy, scale, and manage FPGA devices for accelerating high performance applications without requiring hardware expertise. The technology enables the next generation of low-latency and high-throughput computing on the vast amount of real-time data processed in the cloud. LegUp Computing, Inc., was spawned from years of research in the Dept. of Electrical and Computer Engineering at the University of Toronto to commercialize the award-winning open-source LegUp high-level synthesis tool.
Today’s datacenters rely on general purpose CPUs augmented with accelerators like GPUs and TPUs for specific tasks. FPGAs have traditionally been used to implement specialized functions, such as network switches, but now FPGAs have been deployed in several cloud service providers. These FPGA compute nodes can offer orders of magnitude higher performance at lower power than traditional servers along with flexibility to handle evolving workloads, an example being Bing search by Microsoft.
In critical enterprise applications, real-time data must be processed by machine intelligence to make decisions in millisecond timeframes. These applications include real-time bidding for advertisements, fraud detection, video analytics, and elastic search. In-memory databases that require high-throughput; speech recognition, which is latency sensitive; and other workloads such as genetic sequencing also can be run on these FPGA compute platforms. Traditional servers are overwhelmed by millisecond real-time requirements that can easily be met by FPGAs. LegUp unlocks and democratizes the FPGAs that are already deployed in public clouds, making them accessible to software engineers.
Dr. Andrew Canis, the CEO of LegUp Computing, Inc., said: “We envision LegUp enabling the next generation of real-time computing in the datacenter, as FPGAs are capable of directly connecting to the network layer offering throughput of 40-400 Gbps. LegUp abstracts away the low-level hardware details and allows software engineers to leverage FPGA capabilities. Because of our relationship with Intel, LegUp is poised to take advantage of Intel’s Stratix 10 FPGAs to deliver optimized solutions for database, networking and machine learning applications.”
LegUp offers a novel C/C++ based programming model using threading libraries familiar to software engineers, and maps software threads into parallel hardware on the FPGA. The company offers software libraries to support a dataflow streaming model, well suited to accelerating network-connected applications. LegUp supports multiple FPGA vendors and platforms, including embedded SoC boards and cloud FPGAs at leading cloud service providers. With LegUp-generated hardware, NoSQL databases with sub-millisecond latencies handling millions of requests per second become possible. LegUp will also offer compression, encryption and machine learning modules for processing the growing amount of data in the cloud. Value-added customizations and tooling for enterprise customers will include real-time decision making systems, big-data analytics, artificial intelligence, genomics, video transcoding, cyber security and financial trading.
According to Jason Anderson, Professor of Electrical and Computer Engineering at the University of Toronto: “The deployment of FPGAs in the cloud is a game-changer for domain-specific compute acceleration. LegUp’s technology overcomes a key obstacle to using FPGAs, namely, the need for hardware-design skills. The Intel funding round of LegUp Computing, Inc. will help transition the important early work done at the university to successful commercial viability.”
Mike Mayberry, chief technology officer and senior vice president of Intel Corporation and managing director of Intel Labs, stated: “The investment in LegUp Computing, Inc. is part of a worldwide effort from Intel Labs and Intel Capital to identify promising new startups from university collaborations. FPGAs provide unique technical advantages in a variety of applications, including compute acceleration, and are an integral part of Intel’s comprehensive product offering.”