Juan Valverde Alcala

Title: Juan Valverde Alcala: Patents in Processor Scheduling and Machine Learning

Introduction

Juan Valverde Alcala is an inventor associated with Munich, Germany. His patent work includes methods for predicting task contention in multicore architectures and scheduling tasks for processor systems. He is associated with Collins Aerospace Ireland Limited and has 3 patents.

Latest Patents

“Machine Learning Based Contention Delay Prediction in Multicore Architectures” describes a method for generating training data for a machine-learning-based Task Contention Model. The method runs microbenchmarks individually on a multiprocessor system, measures performance monitoring counters, analyzes correlations between the counters and benchmarks, and selects relevant counters to create a reduced data array. This data can support predictions of delays caused by contention between parallel tasks.

“Scheduling Tasks for Execution by a Processor System” presents a computer-implemented method for ordering tasks. It accesses an execution model, identifies tasks that are not direct-feedthrough tasks, and uses that information to determine an execution order.

Career Highlights

Alcala’s listed patents address processor performance and task execution. Together, the inventions focus on using measured system behavior and task characteristics to improve contention-delay prediction and scheduling.

Collaborations

His listed coworkers include Javier Mora De Sambricio and Michele Garraffa. His company affiliation is Collins Aerospace Ireland Limited.

Conclusion

Juan Valverde Alcala’s patent portfolio reflects work on machine-learning-based contention analysis and processor task scheduling, with applications in improving how computing systems predict delays and organize execution.