Qualcomm Incorporated

Qualcomm Mems Technologies, Inc.

Sameer Wadhwa

Ankit Srivastava

Lennart Karl-Axel Mathe

Marzio Pedrali-Noy

Seyed Arash Mirhaj

Nitz Saputra

Yi Wang

Dongwon Seo

Lennart Karl Mathe

Ren Li

Mustafa Keskin

Guoqing Miao

Subbarao Surendra Chakkirala

Ashwin Raghunathan

Mahmoud Mohamed Rashad Ibrahim Elhebeary

James Hendrie McIntyre

Hrishikesh Vijaykumar Panchawagh

Zhongze Wang

Kostadin Dimitrov Djordjev

Yipeng Lu

Dinesh Jagannath Alladi

Xiaonan Chen

Nicholas Ian Buchan

Sandeep Louis D'Souza

Andrew Weil

Nathan Felix Altman

Kentaro Yamamoto

Jae Hyeong Seo

Suren Mohan

Masoud Ensafdaran

Bo-Ren Wang

Jaswinder Singh

Lingli Xia

Mowen Yang

Weihua Chen

Shunxi Wang

Xiaoke Wen

Ashwin Ragunathan

Raghunatha Reddy Sripuram

Kezhen Liu

A compute-in-memory array is provided in which each neuron includes a capacitor and an output transistor. During an evaluation phase, a filter weight voltage and the binary state of an input bit controls whether the output transistor conducts or is switched off to affect a voltage of a read bit line connected to the output transistor.

Charge-pump-based current-mode neuron for machine learning

A duty cycle correction circuit includes four pairs of serially coupled transistors. A first two of the serial pairs of transistors couple between an internal node for complement output clock signal and ground. A second two of the serial pairs of transistors couple between the internal node and a power supply node for a power supply voltage. Each serial pair is controlled by a corresponding pair of quadrature clock signals in which one of the quadrature clock signal is delayed with respect to the other quadrature clock signal be one quarter of a clock period. The first two serial pairs of transistors thus combine to discharge the internal node for one-half clock period whereas the second two serial pairs of transistors combine to charge the internal node for one-half clock period so that the complement output clock signal has a 50% duty cycle.

Quadrature duty cycle correction circuit

A first inverter in a clock generation circuit is coupled to an input clock signal and has multiple driver slices. Each driver slice includes first transistors that have gates coupled to the input clock signal, second transistors that have sources coupled to rails of a power supply. Each of the second transistors has a drain coupled to a source of one of the first transistors. The second transistors are turned on or turned off based on signaling state of a differential enable signal. A tuning resistor is coupled to the drains of the first transistors and further coupled to an output of the first inverter. A second inverter outputs a quadrature version of the input clock signal and has an input coupled to the output of the first inverter. A first tunable capacitor is coupled to the output of the first inverter.

Hybrid clocking scheme for SERDES physical layer circuits

A clock conditioning circuit includes a duty cycle correction circuit. The duty cycle correction circuit has a first input capacitor, a first self-biasing inverter and a variable capacitor. The first self-biasing inverter has an input coupled to the first input capacitor. The variable capacitor may be coupled to the first input capacitor. The variable capacitor may be configured to receive a first clock signal. A capacitance of the variable capacitor may be programmable by a capacitance control signal.

Programmable self-biasing inverter for correcting duty cycle distortion

Methods and apparatus for supplying power to a dynamic load, such as a neural network circuit. One example power supply circuit generally includes a voltage regulator circuit and a distribution circuit coupled to one or more outputs of the voltage regulator circuit. The distribution circuit is configured to output different amounts of current based on changes in the dynamic load. For certain aspects, the dynamic load includes a neural network circuit having a plurality of segments. In this case, the distribution circuit may be configured to output the different amounts of current based on which segment in the plurality of segments of the neural network circuit is active.

Neural-network-based power management for neural network loads

Certain aspects of the present disclosure provide techniques for performing machine learning computations in a compute in memory (CIM) array comprising a plurality of bit cells, including: determining that a sparsity of input data to a machine learning model exceeds an input data sparsity threshold; disabling one or more bit cells in the CIM array based on the sparsity of the input data prior to processing the input data; processing the input data with bit cells not disabled in the CIM array to generate an output value; applying a compensation to the output value based on the sparsity to generate a compensated output value; and outputting the compensated output value.

Sparsity-aware compute-in-memory

A transmitter includes driver slices coupled to its output. Each driver slice includes a first differential predriver that is selectively enabled and disabled by a first switch based on a control code configuration. A second differential predriver provides a first differential buffered data signal to a first group of driver slices when the second differential predriver is enabled. A second switch enables the second differential predriver when the control code is configured to enable the first differential predriver in at least one driver slice in the first group of driver slices. A third differential predriver provides a second differential buffered data signal to a second group of driver slices when the third differential predriver is enabled. A third switch enables the third differential predriver when the control code is configured to enable the first differential predriver in at least one driver slice in the second group of driver slices.

Low power auto-scalable differential predriver

A low-power phase interpolator circuit has a phase generator that receives an input clock signal and uses the input clock signal to generate multiple intermediate clock signals with different phase shifts; a phase rotator circuit that outputs phase-adjusted clock signals, each phase-adjusted clock signal having a phase that lies within a range bounded by phases of two of the intermediate clock signals; a frequency doubler circuit that receives a plurality of the phase-adjusted clock signals and outputs two frequency-doubled clock signals having a 180° phase difference; and a quadrature clock generation circuit that receives the two frequency-doubled clock signals and provides four output signals that include in-phase and quadrature versions of the two frequency-doubled clock signals.

Low-power inter-die communication using delay lines

Various embodiments include devices and methods for a multi-bit multiplier-accumulator (MAC). Some embodiments may include an analog adder having a first adder capacitor. The first adder capacitor may add a plurality of single-bit MAC outputs by receiving the plurality of single-bit MAC outputs from a plurality of single-bit MACs, and storing the plurality of single-bit MAC outputs. In some embodiments, the analog adder may output a multi-bit MAC output based on addition of the stored plurality of single-bit MAC outputs.

Analog adders for multi-bit MAC arrays in reconfigurable analog based neural networks

A delay cell for a delay locked loop, DLL, based serial link is disclosed. The delay cell has a first stage and a second stage, wherein an output of the first stage is an input to the second stage, the first stage comprising a resistive digital to analog converter, R-DAC and the second stage comprising a current starved delay cell.

Low jitter delay cell

A low-power phase interpolator circuit has a phase generator that receives an input clock signal and uses the input clock signal to generate multiple intermediate clock signals with different phase shifts; a phase rotator circuit that outputs phase-adjusted clock signals, each phase-adjusted clock signal having a phase that lies within a range bounded by phases of two of the intermediate clock signals; a frequency doubler circuit that receives a plurality of the phase-adjusted clock signals and outputs two frequency-doubled clock signals having a 180&#xb0; phase difference; and a quadrature clock generation circuit that receives the two frequency-doubled clock signals and provides four output signals that include in-phase and quadrature versions of the two frequency-doubled clock signals.

A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate for multiplying the stored bit with an input vector bit. An output node for the logic gate connects to a second plate of a capacitor. A first plate of the capacitor connects to a read bit line. A write driver controls a power supply voltage to the cross-coupled inverters, the first switch, and the second switch to capacitively write the stored bit to the pair of cross-coupled inverters.

Compute-in-memory bitcell with capacitively-coupled write operation

A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate formed by a pair of switches for multiplying the stored bit with an input vector bit. A controller controls the pair of switches responsive to a sign bit during a computation phase of operation and controls the pair of switches responsive to a magnitude bit during an execution phase of operation.

Compute-in-memory with ternary activation

Certain aspects of the present disclosure provide apparatus and techniques for capacitance modulation to mitigate pixel leakage in ultrasonic sensors. For example, certain aspects are directed to an ultrasonic sensor including a column line, a pixel having a transistor coupled to the column line, a pixel control circuit coupled to a drain the transistor of the pixel. The ultrasonic sensor may also include a column control circuit coupled to a source of the transistor, wherein at least one of the pixel control circuit or the column control circuit is configured to couple at least one of the drain or the source of the transistor, respectively, to an electric ground during a hold phase of the ultrasonic sensor, and a receiver circuit coupled to the column line.

Capacitance modulation to mitigate pixel leakage in ultrasonic sensors

Certain aspects of the present disclosure provide techniques for look-ahead column sensing for fast voltage-mode read on ultrasonic sensors. For example, certain aspects are directed to an ultrasonic sensor that generally includes a column line, a pixel having a transistor coupled between a voltage rail and the column line, a receiver circuit, and a first column control circuit coupled between the receiver circuit and the pixel, the first column control circuit being configured to electrically isolate the column line from the receiver circuit during a look-ahead settling phase of the ultrasonic sensor, and electrically couple the column line to the receiver circuit during a sensing phase of the ultrasonic sensor.

Growing community of inventors

San Diego, CA, United States of America

Sameer Wadhwa