University of California

Carnegie Mellon University

Bina Technologies, Inc.

Accumente, LLC

Jike Chong

Ian R Lane

Senaka Buthpitiya

Jungsuk Kim

Youngmin Yi

Marghoob Mohiyuddin

Narges Bani Asadi

Henry Chen

Austin Doupnik

Ekaterina I Gonina

The present disclosure provides systems and methods for nucleic acid sequence analysis. A system for processing raw nucleic acid sequence data from a genomic sequencer comprises a data processing server having a housing contained therein one or more processing modules. The one or more processing modules can each comprise an electronic control unit programmed to align nucleic acid sequence data from a genomic sequencing device and perform one or more of variant analysis and structural variant analysis on the nucleic acid sequence data. The system can further comprise a computer server in communication with the processing server. The computer server can be programmed or otherwise configured to process and/or analyze the aligned nucleic acid sequence data.

Systems and methods for processing nucleic acid sequence data

The present invention describes methods for performing large-scale graph traversal calculations on parallel processor platforms. The invention describes methods for on-the-fly hypothesis rescoring that utilizes graphic processing units (GPUs) in combination with utilizing central processing units (CPUs) of computing devices. The invention is described in one embodiment as applied to the task of large vocabulary continuous speech recognition.

Methods for hybrid GPU/CPU data processing

A method of optimizing the calculation of matching scores between phone states and acoustic frames across a matrix of an expected progression of phone states aligned with an observed progression of acoustic frames within an utterance is provided. The matrix has a plurality of cells associated with a characteristic acoustic frame and a characteristic phone state. A first set and second set of cells that meet a threshold probability of matching a first phone state or a second phone state, respectively, are determined. The phone states are stored on a local cache of a first core and a second core, respectively. The first and second sets of cells are also provided to the first core and second core, respectively. Further, matching scores of each characteristic state and characteristic observation of each cell of the first set of cells and of the second set of cells are calculated.

Utilizing multiple processing units for rapid training of hidden markov models

Methods for faster statistical inference in computation based recognition problems on highly parallel processors with multiple cores on-a-chip are disclosed, which include: selectively flattening levels of the recognition network to improve inference speed (improving the recognition model); selectively duplicating parts of the recognition network to minimize a critical section in atomic accesses to as few as one atomic instruction (improving the recognition procedure); and combining weight and source port into one 32-bit word to minimize the number of atomic operations. These methods have been implemented on an NVIDIA GTX 280 processor in a Large Vocabulary Continuous Speech Recognition (LVCSR) embodiment, and achieve more than a 10&#xd7; speed up compared to a highly optimized sequential implementation on an Intel Core i7 processor.

Method and system for parallel statistical inference on highly parallel platforms

Growing community of inventors

Sunnyvale, CA, United States of America

Jike Chong