About TrIP-Sen


Trip-Sen is a fabless semiconductor company focusing on telehealth and other healthcare applications, and developping modular, reconfigurable and cognitive Integrated Circuits (IC).

Value proposal


Implementation of a resonantly-driven CNOT gate with two different ESR resonances depending on the state of the adjacent control qubit. Frequency separation being controlled through J-coupling adjustment by the middle gate.

From Princeton Univ, Konstanz Univ, Maryland Univ and Tokyo Univ.


#qubit #SiliconQubit #SiliconOnInsulator #soitec #RabiFrequency #ESR #SpinResonance #Zeeman #QuantumDot #CNOT #microelectronics #semiconductor


Qubit spin resonance CNOT gate


Pierre Gandolfo, 2020



Interesting joint use of CNN and TCN (Temporal Convolutional Neural Network) with Pulsed Radar for robust hand-gesture recognition on edge AI devices with tight memory footprint and power constraints (<100kB and ~120mW).





Pierre Gandolfo, 2020



Nanoparticles patterning/microfluidics manipulation (e.g. bacteria, proteins biomarkers for blood sample analysis) though primarily acoustic radiating force in a confined evanescent Bessel beam. Opto-mechanical nano-resonator could then be used for single-particle mass-spectrometry.

From Univ of Grenoble and CEA-Leti

#microfluidics #massSpectrometry #proteomics #pointOfCare #peptide #bloodSampleAnalysis #ultrasouund #Bessel #BesselBeam #nanoresonator #piezoelectric #microelectronics #semiconductors #nanotechnologies #MEMS #CEA #Leti #biologicalAnalysis




Pierre Gandolfo, 2020



Fourier-transform spectrometer-on-chip in SOI (MZI interferometer) made of waveguides of varying lengths for OPL (Optical Path Length) modulation and combined with compressive-sensing spectrum reconstruction technique (for both narrow/sparse and broad/smooth spectral features retrieval); and as such not suffering from the typical trade-off between spectral resolution and SNR while being scalable through the use of a single photodetector.
Same spectrum reconstructions techniques could be further applied to other types of spectrometer-on-chip (e.g. Michelson and Fabry-Perrot interferometers with moveable MEMS micro-mirror as well as organic photodiodes array with small # of pixels)

From Massachusetts Institute of Technology (MIT)

#spectrometer #spectroscopy #NearInfrared #MachZehnder #Michelson #FabryPerrot #Interferometer #CompressiveSensing #RegularizedRegression #photodiode #RayleighCriterion #SiliconOnInsulator #Soitec #microelectronics #semiconductor




Pierre Gandolfo, 2020



Insightful that Terahertz imaging (although active) could be implemented in 65-nm bulk CMOS (so way beyond Ft/Fmax limits) through distributed resistive self-mixing operated in triode mode.
Each pixel incorporating small ring antenna behaving like an elementary magnetic dipole instead of Bessel function radiation pattern given its small size (80um pitch) vs. wavelength.

From ISEN/IEMN Lille Univ., STMicroelectronics Crolles and IHCT Bergische Univ:

#terahertz #terahertzImaging #bessel #selfMixing #RF #NoiseEquivallentPower #antenna #bolometer #NonDestructiveSensing #seeThru #xray #MachineVision #microelectronics #semiconductor





Pierre Gandolfo, 2020



Quantum radar/illumination (QI) but directly operated in the microwave domain (so without any Electro-Opto-Mechanical converter (EOM)), using a JPV (Josephson Parametric Converter) for signal-idler entanglement and a digital phase-conjugate receiver with M copies of chopped data (500 samples each) for target presence/absence discrimination. QI vs. Classical Illumination (CI) boosting the detection efficiency of low-reflectivity objects in environments with low signal power (e.g. non-invasive biomedical scanning, short-range radar type of apps).

From Institute of Science and Technology Austria, Massachusetts Institute of Technology, University of York and Camerino, INFN and CNR-INO

#quantumIllumination #classcicalIllumination #radar #entanglement




Pierre Gandolfo, 2020



Quantum radar/illumination (QI) with Optical Parametric Amplification (OPA) maintaining high returned signal-idler mode pairs cross-correlation but in low-brightness environments, through initial SPDC (Spontaneous Parametric Downconversion) signals entanglement, compounded with the processing gain of the Many-M mode pairs for sensitive target detection (Hypothesis H1).

From BBN Technologies and California Institute of Technology

#quantumIllumination #classcicalIllumination #radar #spdc #entanglement #microelectronics #semiconductor



Pierre Gandolfo, 2020



24/7 multianalyte “Lab-on-a-chip” sensor based on ion-sensitive FETs (in FD-SOI-UTB: Fully-depleted Silicon-On-Insulator with Ultra-Thin-Body) with stacked microfluidics interface (in SU-8 photoresist) to passively pump minute amounts of sweat for analysis through capillary action. Main benefits of this approach being non-invasive and multi-functional diagnosis although not exhaustive, sensing at rest, very small size (~9x9mm), very low-power consumption (20nW per FET), short fluid propagation latency (~1.7s), bio-compatibility and above all CMOS manufacturing compatibility.

From EPFL (École Polytechnique Fédérale de Lausanne)

#LabOnAChip #MEMS #sensor #ISFET #SiliconOnInsulator #FDSOI #microfluidic #teleheath #eheath #LiquidUnderTest #IonSensingMembrane #ContinuousHealthMonitoring #semiconductor #microelectronics #soitec




Pierre Gandolfo, 2020



Support of intrinsic neuroplasticity for recurrent spiking neural networks (SNN) by having a hybrid neuron (inc. NVM 1T1R circuitry and based on the leaky-integrate and fire (LIF) model) capable of measuring its own firing rate at regular intervals and adjusting it versus target rate.

From CEA-Leti Grenoble, ETH and Univ. Zurich, Politecnico di Torino and Tours Univ

#EdgeArtificialIntelligence #ArtificialIntelligence #AI #neuralnetwork #spikingneuralnetwork #recurrentneuralnetwork #neuromorphic #OxRAM #neuroplasticity #microelectronics #semiconductor #nanotechnology




Pierre Gandolfo, 2020



Corner quantum dots in spilt-gate devices with adjustable inter-dot coupling through SOI (Silicon on Insulator) backgate voltage, fast read-out through RF gate reflectometry and further being scalable to n-Qubit register for the joint implementation of CNOT gate and high-fidelity read-out channels.

From Univ. Grenoble, CEA-Inac and CEA-Leti:

#qubit #SiliconQubit #SiliconOnInsulator #RabiFrequency #Zeeman #QuantumDot #CNOT #CoulombBlockade #Pauli #SpinBlockade #Quantum #microelectronics #semiconductor #semiconductors




Pierre Gandolfo, 2020



Double Quantum Dots (DQD) with different Rashba/Dresselhaus contributions, thus resulting into a different Zeeman energy between the 2 QD’s that will engender a rotation along horizontal axis and so a change of the charge sensor (SET) current when moving back to (2,0)S ground state. This in addition to J energy exchange provide an all-electrical two-axis qubit control inside the Bloch sphere. Possibility also to disable the rotation by adjusting the angle of the applied magnetic field.


From Sandia National Laboratories and Sherbrooke Univ.



qubit SiliconQubit Hamiltonian SiliconOnInsulator Zeeman QuantumDot SpinOrbitCoupling Bloch BlochSphere CryoCMOS microelectronics semiconductor




Pierre Gandolfo, 2020



Logical qubit based on two-electron spin sates initialized/read in the (0.2)S charge ground state. Horizontal-axis rotation due to Zeeman energy difference between the two quantum dots (QD) can be compensated if needed through rotation around the vertical-axis by tuning the J exchange interaction. It should then be possible to enhance such logical qubit scheme (described in the below linked paper) as a scalable qubit array in FD-SOI (Fully-Depleted Silicon-On-Insulator): series of Face-to-Face spilt corner gates with inter-dot coupling each mediated by their respective backgate voltage (red arrow). A FD-SOI implementation would further allow to reduce heat dissipation of the surrounding control electronics (cryo-CMOS operation) by adjusting threshold voltage.


Original article from Harvard Univ, Weizmann Institute of Science and Santa Barbara Univ.



qubit SiliconQubit Hamiltonian SiliconOnInsulator Zeeman QuantumDot SpinOrbitCoupling Bloch BlochSphere CryoCMOS microelectronics semiconductor soitec




Pierre Gandolfo, 2020



Electrically-driven Yagi-Uda antenna operated in the optical instead of RF domain and embedded within an Al2O3 waveguide film for highly-directive inter-chiplet data communications with high bandwidth and small silicon footprint.


From Universität Würzburg



chiplet sip heterogeneousintegration soc semiconductors antenna datacommunication opticalnetwork microelectronics




Pierre Gandolfo, 2020



A secure wake-up radio against Denial of Sleep type of attacks, with no communication overhead and possibility of resynchronization (in case of missed ACK or brute-force attack), by being recursive while including past encrypted message from the main radio into the computation of each wake-up token. Possibility then to use a low-complexity, although not frequency-tunable, super-regenerative receiver with quench generator and cascode RF amplifier to speed up the oscillation envelope growth for bit decision making.


From Grenoble Alpes University and Ghent University



80211ba wakeupradio wireless radio denialofservice superregenerative bluetooth lorawan wifi halow zigbee 6lowpan #802154 wirelesshart MBAN WPAN MICS




Pierre Gandolfo, 2020



Unsupervised self-learning mode for fully-connected Spiking Neural Networks (SNN) based on stochastic Spike Timing Dependent Plasticity (STDP) for both Long-Term Potentiation/Depression (LTP/LTD) and built around NVM 1T1R circuitry matrix for the synapse memory elements.


From Univ. Grenoble Alpes and CEA-Leti (D. Garbin's thesis)



EdgeArtificialIntelligence ArtificialIntelligence AI neuralnetwork spikingneuralnetwork neuromorphic OxRAM neuroplasticity STDP LongTermPotentiation ShortTermPotentiation microelectronics semiconductor nanotechnology





Pierre Gandolfo, 2020



Neuromorphic cells for Spiking Neural Networks (SNN) based on the leaky-integrate and fire (LIF) model and leveraging a distributed architecture by combining NVM RRAM 1T1R circuit with differential pair integrator (co-integrated into BEOL/FEOL semiconductor manufacturing process) for both neurons and synapses. Such distributed architecture improving power efficiency and eliminating memory bottleneck for edge AI devices while reducing silicon footprint as well as better memory cycling endurance.


From CEA-Leti Grenoble, ETH and Univ. Zurich, Politecnico di Torino and Tours Univ.



EdgeArtificialIntelligence ArtificialIntelligence AI neuralnetwork spikingneuralnetwork neuromorphic RRAM FrontEndOfLine BackEndOfLline




Pierre Gandolfo, 2020



Good tutorial on the MUSIC (MUltiple SIgnals Classifier) algo for high-resolution DoA (Direction of Arrival) determination from multi-uncorrelated sources by leveraging the orthogonality between the noise subspace eigenvectors of the array correlation matrix with the array steering vectors while keeping a maximum field of view with an antenna spacing of λ/2.




radar adas autonomouscar beamforming localization RTLS





Pierre Gandolfo, 2020



VGA image sensor with end-of-column Compressive Sensing (CS) and inference done in the CS domain for embedded object recognition with dimensionality reduction of 1/480 and high classification accuracy, thus enabling drastic reduction of memory/processing requirements for edge AI devices with S-sparse signals.


From Univ of Grenoble-Alpes, CEA-Leti and UCLouvain



CompressiveSensing AnalogToInformationSensor EdgeArtificialIntelligence ArtificialIntelligence AI ObjectRecognition SparseSignal ImageSensor CMOSImageSensor CIS sensor CompressedSensing CompressiveSampling SparseSampling




Pierre Gandolfo, 2019



Closed loop control of an AC/AC matrix converter with a PR controller for motor VFD apps in order to mimize the effects of unbalanced input voltages and harmonics while benefiting from bi-directional switching (e.g. GaN-based switches) for 4-quadrant operation, regeneration capabilities, power factor control and no use of bulky DC-link reactive element.


From Louisiana State University



GaN ACAC powerconverter voltagesourceinverter inverter motor matrixconverter spacevector pwm galliumnitride




Pierre Gandolfo, 2019



Some technical insights for 3D ToF (Time-of-flight) on how SBI (Suppression of Background Illumination) in every pixel is typically implemented with PMD pixel architecture. Two reverse-biased diodes (charge wells for the collecting of generated charges from incident photons) adjacent to two PG-PMD (Photogate Photonic Mixer Devices) driven in phase opposition and tied to the transmit light modulated signal so that the effect of ambient light get removed by subtracting the outputs of both channels.


Section III of the following paper from Sinhgad College of Engineering for more details.



3dtof timeofflight authentication facerecognition handgesturerecognition depthsensing nir nearinfrared imagesensor 3dcamera





Pierre Gandolfo, 2019



Regarding Apple new U1 chip for UWB (ultra-wideband) positioning on the iPhone 11 Pro/Pro Max, will be interesting to see for which application types Apple intends to use it over time (e.g. car/home secure access, asset theft prevention…besides AirDrop file sharing), over which range they intend to operate, and which type of UWB pulse waveform they are using (e.g. IEEE 802.15.4a UWB-IR-like, .15.4f, .15.4z, proprietary) as an attacker could reduce the upper-bound distance if there is any time diversity/processing gain built into their UWB waveform by then exploiting the symbol structure predictability through ED/LC (Early Detect / Late Commit) type of attacks at the physical layer unless some form of random pulse reordering/blinding is applied or then limited to very short-range apps with one pulse per bit.


From ETH Zurich university



uwb ultrawideband apple iphone u1 positioning




Pierre Gandolfo, 2019



The use of Hadamard matrix as coding scheme for robust chip-to-chip communication links with wider eye opening (both horizontally and vertically) for a given data rate, reduced EMI/X-talk and lower common-mode noise at receiver end.




chiplet sip heterogeneousintegration soc semiconductors





Pierre Gandolfo, 2019



A rather recent and comprehensive survey of normally-off techniques for GaN-based HEMT (inc. pros/cons of each method) with the exception perhaps of the recessed-gate MIS-HEMT with floating reference connected to source.




powerapps gan matrixconverter induction motor pmsm ledlightingsystems acdc charger amplifier




Pierre Gandolfo, 2019



Reprogrammable and always-on sensing at reduced power budget with sensor data classification done in the analog domain.


From West Virginia Univ.



sensor iot wireless fpga ai artificialintelligence edgenetwork




Pierre Gandolfo, 2019



Silicon on Thin Buried Oxide (SOTB) from LEAP/Renesas: Failed to understand what's "new" here since SOITEC and STMicroelectronics with FD-SOI had long demonstrated (since 2002) the interface-coupling effects between front and back channels (fct. of thickness ratio between gate oxide and BOX) and how this could be used to alter not only the threshold voltage but also subthreshold slope and transconductance.






FDSOI soitec iot energyconsumption powermanagement SOI



Pierre Gandolfo, 2019



GaN HEMT benefits nicely illustrated here for this 2kVI vehicle inverter by moving away from the highly power-efficient/ruggetized but extremely complex dual buck inverter topology with MOSFET devices paralleling and additional circuitry for min. of body diodes reverse recovery losses at hard switching events.




gan powermanagement inverter powerapps





Pierre Gandolfo, 2019



A different type of resonant start-up-converter for energy harvesting (cold start) based on a piezoelectric transformer and Armstrong oscillator but not using any biasing inductance (from Univ. Normal Sup Cachan). While not described in their paper, that same silicon-integrable Rosen-type piezo transformer could be further reused for the step-up DC-DC main converter with high power efficiency given its high Q factor and resulting soft switching capabilities




energyharvesting powermanagement microelectronics internetofthings iot wireless


No alt text provided for this image

Pierre Gandolfo, 2019



A different radar approach achieving high range resolution without boosted bandwidth by progressively increasing the cross-correlation coherence window length and using regression method to a linear or oscillatory function in order to determine breakpoint (= fixed/moving range target) --> well-suited for densely-populated areas with less interference but at the cost of longer sweep time / duty cycle.


From Univ. Tel Aviv



radar adas autonomouscar drone radiofrequency




Pierre Gandolfo, 2019





Patents - Pierre Gandolfo


FR2.955.991 - Active-Inductor-Oscillator



US8755745B2 - Modular, reconfigurable and cogntive microsystem for the monitoring and remote control of smart communicating objects



US20030224787A1 - System and method of communication between multiple point-corrdinated wireless networks



Under construction


Version imprimable Version imprimable | Plan du site



Version imprimable Version imprimable | Plan du site
© TrIP-Sen - Pierre Gandolfo Microelectronics Naotechnologies