Did you miss the session at the Data Summit? See on-demand here.
The market for Edge AI chips is designed to accelerate offline AI workloads, often in self-contained hardware, maturing into fast clips. Based on converging trends such as digital transformation, cloud-native technology and the Internet of Things, the value of the same AI hardware segment could reach $ 38.87 billion by 2030, according to Values Reports estimates.
Values point to lower latency and increased demand for real-time processing as well as reduction in storage and operation costs as factors driving the use of AI chips. Indeed, these types of chips can enable better performance and lower power consumption by reducing the need for cloud-dependent devices for data processing. But the same AI chips are built in another way: for example, because it lacks the computing power of a cloud datacenter, only selective tasks can be performed on the same device.
Quadrick is one of the many startups diving enthusiastically into the AI Edge market, promising to overcome the historical hurdles of Edge hardware. Today, Quadrick announced a $ 21 million Series B funding round, co-led by Denson’s NSITEXE and MegaChips, with the participation of Livewood VC, Pear VC, Uncork Capital, and Cota Capital, to accelerate production of the same AI chips the company claims. doing. “Speed up the entire application pipeline” on the device without the need for a powerful general purpose processor.
Secret chip sauce
Quadrick, based in Burlingham, California, was founded in 2016 by Virbhan Khetarpal, Nigel Drago and Daniel Firu. All three are from MIT and Carnegie Mellon and the formerly co-founded cryptocurrency computing company 21 Inc.
“The founding team was building smart robots while facing the inadequacy of Nvidia and Intel’s existing compute platforms,” a Quadric spokesperson told VentureBeat via email. “Unless it is rebuilt from the ground up, the processors used for edge computing are not measurable. The Quadrick was founded to create new processor architectures; One that normalizes the dataflow paradigm and delivers high levels of power efficiency for a wide range of algorithms in machine learning, computer vision, DSP, graph processing and linear algebra. “
Quadric claims that its 1.1-billion-transistor, 16-nanometer chip uses only 4.5W of power and packs 4GB of memory with 256 “vertex cores”, designed to speed up some of the algorithmic workloads involved in common AI applications. Workloads do not include training, or the step of developing an AI system that requires the system to feed large amounts of data in order to learn to predict. Instead, they make guesses, which is the point at which the system can make predictions based on new data coming.
Quadrick’s unique ability to handle both neural backbones and classical dynamic data-parallel algorithms in a unified architecture is helping to create AI for everyone, everywhere. Most other solutions combine high-power processor clusters with application-specific neural processing units, “a Quadrick spokesperson said. The company further explains on its website:” The architecture is instruction-based … [it] Is a software programming model designed for ease of use by the developer. The software programming model allows the developer to express graph-based and non-graph-based algorithms together.
Quadric offers plug-and-play AI models for applications in the warehousing, construction, transportation and agricultural industries. The company previously claimed that Denso plans to integrate its Edge chip technology, which works with any machine with an M.2 motherboard expansion slot, into future self-driving vehicle solutions.
Expansion of Edge Market
Deloitte estimates that more than 750 million Edge AI chips that work on the device have been sold to date, representing $ 2.6 billion in revenue.
The spokesperson added, “The intensity of data generated in industries is increasing rapidly, so to handle these data volumes, the next generation of innovation in computing will be outside the datacenter and near the network edge,” the spokesperson added. “Quadric helps enterprises create data solutions that are sensitive to privacy and optimize latency and bandwidth costs.”
The Quadric AI competes against companies including Storm, Axelera, Deep Vision, Flex Logix, Sima.ai, Blaize and Hailo, with the latter raising more than 20 320 million at a valuation of over $ 1 billion. As ZDNet’s Tiernan Ray highlights in a recent issue, Venture Financing has supercharged the AI Edge chip market, with dozens of vendors (one report counting over 60) scrambling to get a piece of the growing cash heap.
But Quadrick believes its AI chip architecture sets it apart in a growing space. The company maintains that its investment will enable Quadrick, which reportedly has five customers, to release the next version of its chip architecture; Improve the performance of the software development kit it sends along with its chips; And roll out new products for integration into system-on-chips.
“Most other companies in the same computing space are building specific workload accelerators. In contrast, Quadrick’s software is future-proof against the dynamic background of centric architecture algorithms and AI models, “the spokesperson explained.
To date, the 35-employee Quadrick has raised $ 34 million in venture capital and $ 2 million in debt, including a $ 15 million round in May 2019.
Venturebeat’s mission Transformative Enterprise is about to become a digital town square for technology decision makers to gain knowledge about technology and transactions. Learn more