Search for: All records

Abstract We review two magnetic tunnel junction (MTJ) approaches for compact, low-power, CMOS-integrated true random number generation (TRNG). The first employs passive-read, easy-plane superparamagnetic MTJs (sMTJs) that generate thermal-fluctuation-driven bitstreams at 0.5–1 Gb s⁻¹per device. The second uses MTJs with magnetically stable free layers, operated with stochastic write pulses to achieve switching probabilities of about 0.5 (i.e. write error rates of $\simeq 0.5$), achieving $\gtrsim 0.1$ Gb s⁻¹per device; we refer to these as stochastic-write MTJs (SW-MTJs). Randomness from both approaches has been validated using the NIST SP 800-22r1a test suites. sMTJ approach uses a read-only cell with low power and can be compatible with most advanced CMOS nodes, while SW-MTJs leverage standard CMOS MTJ process flows, enabling co-integration with embedded spin-transfer torque magnetic random access memory. Both approaches can achieve deep sub-0.01 µm²MTJ footprints and offer orders-of-magnitude better energy efficiency than CPU/GPU-based generators, enabling placement near logic for high-throughput random bitstreams for probabilistic computing, statistical modeling, and cryptography. In terms of performance, sMTJs generally suit applications requiring very high data-rate random bits near logic processors, such as probabilistic computing or large-scale statistical modeling. Whereas SW-MTJs are attractive option for edge-oriented microcontrollers, providing entropy sources for computing or cryptographic enhancement. We highlight the strengths, limitations, and integration challenges of each approach, emphasizing the need to reduce device-to-device variability in sMTJs—particularly by mitigating magnetostriction-induced in-plane anisotropy—and to improve temporal stability in SW-MTJs for robust, large-scale deployment. 

Abstract Increasing the number of organ donations after circulatory death (DCD) has been identified as one of the most important ways of addressing the ongoing organ shortage. While recent technological advances in organ transplantation have increased their success rate, a substantial challenge in increasing the number of DCD donations resides in the uncertainty regarding the timing of cardiac death after terminal extubation, impacting the risk of prolonged ischemic organ injury, and negatively affecting post-transplant outcomes. In this study, we trained and externally validated an ODE-RNN model, which combines recurrent neural network with neural ordinary equations and excels in processing irregularly-sampled time series data. The model is designed to predict time-to-death following terminal extubation in the intensive care unit (ICU) using the history of clinical observations. Our model was trained on a cohort of 3,238 patients from Yale New Haven Hospital, and validated on an external cohort of 1,908 patients from six hospitals across Connecticut. The model achieved accuracies of$$95.3~\pm ~1.0\%$$and$$95.4~\pm ~0.7\%$$for predicting whether death would occur in the first 30 and 60 minutes, respectively, with a calibration error of$$0.024~\pm ~0.009$$. Heart rate, respiratory rate, mean arterial blood pressure (MAP), oxygen saturation (SpO2), and Glasgow Coma Scale (GCS) scores were identified as the most important predictors. Surpassing existing clinical scores, our model sets the stage for reduced organ acquisition costs and improved post-transplant outcomes. 

Abstract B-trees are widely recognized as one of the most important index structures in database systems, providing efficient query processing capabilities. Over the past few decades, many techniques have been developed to enhance the efficiency of B-trees from various perspectives. Among them,B-tree compressionis an important technique introduced as early as the 1970s to improve both space efficiency and query performance. Since then, several B-tree compression techniques have been developed. However, to our surprise, we have found that these B-tree compression techniques werenevercompared against each other in prior works. Consequently, many important questions remain unanswered, such as whether B-tree compression is truly effective or not. If it is effective, under what scenarios and which B-tree compression methods should be employed? In this paper, we conduct an experimental evaluation of seven widely used B-tree compression techniques using both synthetic and real datasets. Based on our evaluation, we present lessons and insights regarding the use of B-tree compression that can be leveraged to guide system design decisions in modern databases.