SiGe epitaxial memory for neuromorphic computing with reproducible high performance based on engineered dislocations.

Although several types of architecture combining memory cells and transistors have been used to demonstrate artificial synaptic arrays, they usually present limited scalability and high power consumption. Transistor-free analog switching devices may overcome these limitations, yet the typical switching process they rely on-formation of filaments in an amorphous medium-is not easily controlled and hence hampers the spatial and temporal reproducibility of the performance. Here, we demonstrate analog resistive switching devices that possess desired characteristics for neuromorphic computing networks with minimal performance variations using a single-crystalline SiGe layer epitaxially grown on Si as a switching medium. Such epitaxial random access memories utilize threading dislocations in SiGe to confine metal filaments in a defined, one-dimensional channel. This confinement results in drastically enhanced switching uniformity and long retention/high endurance with a high analog on/off ratio. Simulations using the MNIST handwritten recognition data set prove that epitaxial random access memories can operate with an online learning accuracy of 95.1%.

Unforeseen decreases in dissolved oxygen levels affect tube formation kinetics in collagen gels.

The availability of oxygen (O(2)) is a critical parameter affecting vascular tube formation. In this study, we hypothesize that dissolved oxygen (DO) levels in collagen gels change during the three-dimensional (3D) culture of human umbilical vein endothelial cells (HUVECs) in atmospheric conditions and that such changes affect the kinetics of tube formation through the production of reactive oxygen species (ROS). We demonstrate a decrease in O(2) tension during 3D cultures of HUVECs. Noninvasive measurements of DO levels during culture under atmospheric conditions revealed a profound decrease that reached as low as 2% O(2) at the end of 24 h. After media replacement, DO levels rose rapidly and equilibrated at ∼15% O(2), creating a reoxygenated environment. To accurately estimate DO gradients in 3D collagen gels, we developed a 3D mathematical model and determined the Michaelis-Menten parameters, V(max) and K(m), of HUVECs in collagen gels. We detected an increase in ROS levels throughout the culture period. Using diphenyliodonium to inhibit ROS production resulted in the complete inhibition of tube formation. Interference RNA studies further showed that hypoxia-inducible factors (HIFs)-1α and -2α are not involved in the formation of 3D tubes in collagen gels. We conclude that ROS affect the tubulogenesis process through HIFα-independent pathways, where the levels of ROS are influenced by the uncontrolled variations in DO levels. This study is the first demonstration of the critical and unexpected role of O(2) during 3D in vitro culture models of tubulogenesis in atmospheric conditions.

Alloying conducting channels for reliable neuromorphic computing.

A memristor1 has been proposed as an artificial synapse for emerging neuromorphic computing applications2,3. To train a neural network in memristor arrays, changes in weight values in the form of device conductance should be distinct and uniform3. An electrochemical metallization (ECM) memory4,5, typically based on silicon (Si), has demonstrated a good analogue switching capability6,7 owing to the high mobility of metal ions in the Si switching medium8. However, the large stochasticity of the ion movement results in switching variability. Here we demonstrate a Si memristor with alloyed conduction channels that shows a stable and controllable device operation, which enables the large-scale implementation of crossbar arrays. The conduction channel is formed by conventional silver (Ag) as a primary mobile metal alloyed with silicidable copper (Cu) that stabilizes switching. In an optimal alloying ratio, Cu effectively regulates the Ag movement, which contributes to a substantial improvement in the spatial/temporal switching uniformity, a stable data retention over a large conductance range and a substantially enhanced programmed symmetry in analogue conductance states. This alloyed memristor allows the fabrication of large-scale crossbar arrays that feature a high device yield and accurate analogue programming capability. Thus, our discovery of an alloyed memristor is a key step paving the way beyond von Neumann computing.

Publisher Correction: Alloying conducting channels for reliable neuromorphic computing.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Recapitulating physiological and pathological shear stress and oxygen to model vasculature in health and disease.

Studying human vascular disease in conventional cell cultures and in animal models does not effectively mimic the complex vascular microenvironment and may not accurately predict vascular responses in humans. We utilized a microfluidic device to recapitulate both shear stress and O2 levels in health and disease, establishing a microfluidic vascular model (µVM). Maintaining human endothelial cells (ECs) in healthy-mimicking conditions resulted in conversion to a physiological phenotype namely cell elongation, reduced proliferation, lowered angiogenic gene expression and formation of actin cortical rim and continuous barrier. We next examined the responses of the healthy µVM to a vasotoxic cancer drug, 5-Fluorouracil (5-FU), in comparison with an in vivo mouse model. We found that 5-FU does not induce apoptosis rather vascular hyperpermeability, which can be alleviated by Resveratrol treatment. This effect was confirmed by in vivo findings identifying a vasoprotecting strategy by the adjunct therapy of 5-FU with Resveratrol. The µVM of ischemic disease demonstrated the transition of ECs from a quiescent to an activated state, with higher proliferation rate, upregulation of angiogenic genes, and impaired barrier integrity. The µVM offers opportunities to study and predict human ECs with physiologically relevant phenotypes in healthy, pathological and drug-treated environments.