While many eyes, including mine, were on CES this week, today, I wanted to talk about a tech announcement that may have flown under the radar in the run-up to the holidays. I’ve been following Synaptics for some time now and watched it progress from Synaptics 1.0 (PC trackpads, etc.) to 2.0 (mobile, fingerprint readers) and now, what it calls Synaptics 3.0—a much more focused Synaptics, expanding aggressively into new, rapidly growing markets such as consumer IoT and automotive. Under new CEO Michael Hurston (who took the reins in 2019), Synaptics has been able to break out of the stagnation it fell into in the mid-2010s. While Synaptics has long had an impressive portfolio of IP—1,800+ patents and counting—it struggled to develop effective roadmaps and engage constructively with customers. A big part of the company’s comeback is due to its renewed focus and diversification into IoT, the success of which was evident in Synaptics’ recent Q1 earnings results. And that is the category where today’s news falls: the introduction of a new offering, the Katana Platform, an SoC for low power edge AI. Let’s take a look at what Katana is all about.
What is it?
The Katana Edge AI platform is the result of a new partnership with edge AI developer, Eta Compute, geared to, in Synaptics’ language, “address a growing industry gap for solutions that enable battery-powered devices for consumer and industrial IoT markets.” In the development of Katana, Synaptics contributed its ultra-low-power SoC architecture while Eta Compute brought the AI brawn.
Inside of the platform is Synaptics’ multi-core processor architecture, comprised of power and energy-optimized neural network and domain-specific processing cores and a healthy amount of on-chip memory. Synaptics says the platform also leverages specific architectural techniques designed to save energy in each of its unique modes of operation.
The silicon is supported by Eta Compute’s TENSAI Flow AI software and its power and performance-optimized libraries. This combination is crucial—to get the levels of efficiency industries desire, the underlying silicon must have the kind of software optimization techniques TENSAI Flow brings to the table. Additionally, Eta Compute supports Katana with its neural network compilers, models and various AI applications.
MORE FOR YOU
Specific AI applications Katana is well-suited for include object and people recognition and counting, asset or inventory tracking, environmental sensing and voice and sound detection. Many environments stand to benefit from the deployment of these ultra-low-power edge AI applications. Katana should find a receptive audience in factories, farms, office buildings, retail and smart homes, just to name a few. The platform looks like an excellent option for businesses looking to digitally transform their operations in a sustainable, cost-efficient fashion.
Katana in action
I recently had a call with Synaptics, where we went a little more in-depth on Katana and its potential use cases. One example given was that of a proposed smart office solution. A battery-powered camera, equipped with Katana’s edge intelligence, could monitor how many people are going in and out of, say, a specific conference room to better understand space utilization and manage energy usage more efficiently. This device would be different from a typical surveillance camera. It would only have a high enough resolution and frame rate to perform its function, thereby consuming less power (and perhaps alleviating privacy concerns). Additionally, traditional, non-battery operated cameras can be complicated and time-consuming, expensive and generally just a pain to install. You may have to hire contractors to open up ceiling tiles and run cables, and it may require specific permits. All of this, for a relatively basic, limited function camera. One can see the allure of a battery-operated device in this and other analogous situations. The ultimate goal in this scenario is reducing the associated costs and headaches of running a building, which can be significant.
Procurement and deployment
The two companies plan to work together to deliver specific application-focused kits in the interest of accelerating development and deployment. According to Synaptics, these kits will contain pre-trained ML models and reference designs. For those with the training know-how, customers will also be able to train models with their datasets, leveraging popular frameworks like TensorFlow, Caffe and ONNX. Those wanting to level up can pair Katana with Synaptics’ wireless connectivity offerings to deliver complete system-level modules and solutions.
A broad partnership
Additionally, as part of the partnership, Synaptics led Eta Compute’s recent Series C funding round of $12.5 million, bringing Eta Compute’s total funds raised to $31.5 million. The companies also announced that Satish Ganesan, Synaptics’ Chief Strategy Officer, will join Eta Compute’s Board of Directors. All of this to say, their collaboration on Katana does not appear to be a one-off—they’re laying the groundwork for a broader engagement, which I’m sure will bear additional fruit in the coming years.
Synaptics’ expansion into edge IoT is an intuitive fit, and I’m glad to see it continue to grow its business. I love that Synaptics is no longer treading water—it is leveraging its impressive patent portfolio much more effectively than in previous years, and we see that borne out in the company’s earnings.
Synaptics has an uncanny ability to quickly invest in and build technology for one business that can quickly transfer and adapt to others. By pairing its own bread-and-butter with software solutions from partners like Eta Compute, Synaptics can push even further and more rapidly into these new, high-growth markets. Furthermore, its low-power prowess helps grow the IoT sector as a whole, where there is a considerable growing demand for battery-operated, low-power devices. Synaptics has found its niche and earned a seat at the IoT table. I’ll continue to watch with interest.
Note: Moor Insights & Strategy writers and editors may have contributed to this article.
Moor Insights & Strategy, like all research and analyst firms, provides or has provided paid research, analysis, advising, or consulting to many high-tech companies in the industry, including 8×8, Advanced Micro Devices, Amazon, Applied Micro, ARM, Aruba Networks, AT&T, AWS, A-10 Strategies, Bitfusion, Blaize, Box, Broadcom, Calix, Cisco Systems, Clear Software, Cloudera, Clumio, Cognitive Systems, CompuCom, Dell, Dell EMC, Dell Technologies, Diablo Technologies, Digital Optics, Dreamchain, Echelon, Ericsson, Extreme Networks, Flex, Foxconn, Frame (now VMware), Fujitsu, Gen Z Consortium, Glue Networks, GlobalFoundries, Google (Nest-Revolve), Google Cloud, HP Inc., Hewlett Packard Enterprise, Honeywell, Huawei Technologies, IBM, Ion VR, Inseego, Infosys, Intel, Interdigital, Jabil Circuit, Konica Minolta, Lattice Semiconductor, Lenovo, Linux Foundation, MapBox, Marvell, Mavenir, Marseille Inc, Mayfair Equity, Meraki (Cisco), Mesophere, Microsoft, Mojo Networks, National Instruments, NetApp, Nightwatch, NOKIA (Alcatel-Lucent), Nortek, Novumind, NVIDIA, Nuvia, ON Semiconductor, ONUG, OpenStack Foundation, Oracle, Poly, Panasas, Peraso, Pexip, Pixelworks, Plume Design, Poly, Portworx, Pure Storage, Qualcomm, Rackspace, Rambus, Rayvolt E-Bikes, Red Hat, Residio, Samsung Electronics, SAP, SAS, Scale Computing, Schneider Electric, Silver Peak, SONY, Springpath, Spirent, Splunk, Sprint, Stratus Technologies, Symantec, Synaptics, Syniverse, Synopsys, Tanium, TE Connectivity, TensTorrent, Tobii Technology, T-Mobile, Twitter, Unity Technologies, UiPath, Verizon Communications, Vidyo, VMware, Wave Computing, Wellsmith, Xilinx, Zebra, Zededa, and Zoho which may be cited in blogs and research.