DeePhi Technology Co., Ltd. was founded in 2016 by Han Song, Yao Song and Yu Wanglutz and is doing business in providing deep learning platforms that are convenient, efficient, and cost-effective. Deep compression, compilation toolchain, deep learning processing unit (DPU) design, FPGA development, and system-level optimization are some of the technologies used by them. Xilinx acquired DeePhi Technology for an undisclosed amount.
DeePhi Technology has a total of 16 patents globally, out of which 3 has been granted. Of these 16 patents more than 93% patents are active. China is where DeePhi Technology has filed maximum number of patents and it also seems reasonable as the biggest market for DeePhi Technology is China, it has generated an annual revenue of $15 million in the year 2020. Parallelly, China seems to be the main focused R&D center and is also the origin country of DeePhi Technology.
Do read about some of the most popular patents of DeePhi Technology which have been covered by us in this article and also you can find DeePhi Technology’s patents information, the worldwide patent filing activity and its patent filing trend over the years, and many other stats over DeePhi Technology’s patent portfolio.
How many patents does DeePhi Technology have?
DeePhi Technology has a total of 16 patents globally. These patents belong to 11 unique patent families. Out of 16 patents, 15 patents are active.
How many DeePhi Technology patents are Alive/Dead?
How Many Patents did DeePhi Technology File Every Year?
Are you wondering why there is a drop in patent filing for the last two years? It is because a patent application can take up to 18 months to get published. Certainly, it doesn’t suggest a decrease in the patent filing.
|Years of Patent Filing or Grant||DeePhi Technology Application Filed||DeePhi Technology Patents Granted|
How Many Patents did DeePhi Technology File in Different Countries?
Countries in which DeePhi Technology Filed Patents
|United States Of America||1|
Where are Research Centers of DeePhi Technology Patents Located?
10 Best DeePhi Technology Patents
CN107239825A is the most popular patent in the DeePhi Technology portfolio. It has received 87 citations so far from companies like Qualcomm Incorporated, Microsoft and Amazon Technologies.
10 Best DeePhi Technology Patents
|Publication Number||Citation Count|
What Percentage of DeePhi Technology US Patent Applications were Granted?
DeePhi Technology (Excluding its subsidiaries) has filed 24 patent applications at USPTO so far (Excluding Design and PCT applications). Out of these 18 have been granted leading to a grant rate of 81.82%
Which Law Firms Filed Most US Patents for DeePhi Technology?
|Law Firm||Total Application||Success Rate|
|Ipro Pllc Xilinx||6||80.00%|
|Patterson & Sheridan Llp Xilinx||3||50.00%|
Xilinx acquired Chinese machine learning startup Beijing-based DeePhi Technology Co Ltd.
“Xilinx is accompanying DeePhi Tech along its journey to explore the potential of machine learning and is supporting our innovation as one of our early investors. We look forward to continuing our joint efforts with Xilinx to bring our solutions to the next level in performance,” said Yi Shan, CTO of DeePhi Tech.
Their 3 most popular patent topics include: Artificial neural networks, Artificial intelligence and computational neuroscience.
List of DeePhi Tech Patents
|Publication Number||Title (English)|
|CN107689224B||Deep neural network compression method with reasonable mask code|
|CN107657316B||General processor synergy system design of the neural network processor|
|CN108171244A||Object recognition method and system|
|CN108154522A||The target tracking system|
|CN107688850A||A deep neural network compression method|
|CN107657263A||An advanced treatment unit for realizing the ANN|
|CN107239825A||Considering load balancing depth neural network compression method|
|CN107239823A||A device and method for for realizing sparse neural network|
|WO2019085378A1||Hardware implementation device and method for high-speed full-connection calculation|
|WO2018223822A1||Pruning- and distillation-based convolutional neural network compression method|
|WO2019085377A1||Target tracking hardware implementation system and method|
|WO2019080484A1||Method of pruning convolutional neural network based on feature map variation|
|WO2019085379A1||Hardware realization circuit of deep learning softmax classifier and method for controlling same|
|WO2018188463A1||Programming model oriented to neural network heterogeneous computing platform|
|WO2019080483A1||Neural network computation acceleration method and system based on non-uniform quantization and look-up table|
|US10762426B2||Multi-iteration compression for deep neural networks|