Gartner Identifies Think Silicon a ‘Cool Vendor’ in Novel Semiconductors for Neural Networks
Toronto/CA – Patras/GR – MAY 26th, 2016: Think Silicon S.A. a leading provider of ultra-low power Graphic Processing Unit (GPU) and imaging solution-IP for wearable, mobile devices and IoT platforms, today announced that is has been recognized in the April, 27th 2016 ‘Cool Vendor in Novel Semiconductors for Neural Networks, 2016 report by Gartner1, a leading information technology analyst organization.
Each year Gartner identifies new Cool Vendors selected for the “Cool Vendors” report which are innovative, impactful and intriguing in key technology and publishes a series of research reports highlighting these innovative vendors and their products and services. According to the report, “neural networks are one of the hottest areas in technology. We highlight custom silicon solutions that will shorten the timeline for mainstream deployment of DNNs (Deep Neural Networks) in existing and new applications.
“We believe it is an honor to see our efforts have been further validated by a designation in the Cool Vendor report by Gartner. The Think Silicon team designed from bottom-up a ultra-lowpower GPU platform which scales widely for different applications and markets and enabling our customers to improve their business performance”, said co-founder and CEO George Sidiropoulos.
At CES 2016 in Las Vegas, Think Silicon launched NEMA|p (PICO) and NEMA|t (tiny) the world smallest and most power efficient 2D and 3D Graphics Processor Units (GPU). Based on ultralow- power (ULP) saving methods, the NEMA-GPU increases the battery life of small-display devices up to 100% (from average 3 to 6 days). Think Silicon leverages the ULP technology with the GPGPU component by providing the compute engine for customized FPGA and SoC.
Think Silicon is working on a prototype with thirty-two (32) GPGPU cores (<0.2mm2 per core @ 28nm), containing a fixed-point arithmetic and has extensive multithreading capabilities (128 threads per core). The GPU defines a new class of products with a competitive architecture to execute highly parallel DNN applications and offer substantial power savings over existing solutions.
Gartner does not endorse any vendor, product or service depicted in our research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.
1 Gartner “Cool Vendors in ‘Novel Semiconductors for Neural Networks, 2016” by Michele Reitz, Martin Reynolds, James F. Hines, Nathan Nuttal, Gerald van Hoy, 27 April 2017.
Gartner subscribers may access the full report here