Barcelona, Catalonia, Spain
Over thirty years experience in High Performance Computing in academia and industry. Fourteen years experience in ocean modelling/geophysical fluid dynamics, with emphasis on mathematical and computational aspects of large-scale climate modelling. Developer and maintainer of the STREAM benchmark, the de facto industry standard measure of sustained memory bandwidth. Twelve years experience in the computer industry at SGI, IBM, AMD. Focus on Performance Architecture, Application Performance Analysis, Microbenchmarking, and Strategic Market Analysis. Specialties: Performance Analysis/Modelling/Tuning for applications in High Performance & Technical Computing (HPC). System Architecture for tightly-coupled parallel computer systems (homogeneous and heterogeneous), including low-level HW and SW coherence, consistency, and communication/synchronization protocols. HPC Market Analysis, including Segmentation, Sizing, Requirements, Competitive Analysis, and Strategic Planning.
HPC performance analysis and architecture
Performance analysis for applications and systems in High Performance Computing. Research in Computer Architectures for High Performance Computing.
. Technology lead for AMD's "Torrenza" initiative -- making AMD64 the preferred platform for hosting innovative application acceleration solutions. . Microbenchmarking in support of detailed performance analysis of Opteron memory subsystem and HyperTransport efficiency. . System Architecture protocol design for fine-grain communication and synchronization in homogeneous and heterogeneous systems. . Strategic planning for products related to the High Performance Computing Market.
HPC Performance Team Lead for IBM eServer System Performance organization: . Performance Analysis for HPC codes . Computer System Architecture . Microprocessor Design Engineer (emphasis on memory hierarchy) . HPC Market Analysis & Strategic Planning . HPC Competitive Analysis HPC Performance Analyst in the POWER4/POWER4+ and POWER5/POWER5+ design groups. Large-Scale Architecture Team Lead for IBM's DARPA-funded High Productivity Computing Systems (HPCS) project.
Lead Performance Analyst for the NUMA server architecture group at SGI (March 1998-July 1999). Led performance analysis for the design of the SGI Altix3000 series, and consulted on the design of the SGI Origin3000 and Cray X1. Member of the Technical Staff in Performance Engineering (July 1996-March 1997). Worked with Microprocessor and Server design teams to optimize performance. Tuned industry standard and customer-specific benchmarks.