A
Ayose Falcón
Researcher at Intel
Publications - 37
Citations - 781
Ayose Falcón is an academic researcher from Intel. The author has contributed to research in topics: Branch predictor & Thread (computing). The author has an hindex of 13, co-authored 37 publications receiving 772 citations. Previous affiliations of Ayose Falcón include Hewlett-Packard & Polytechnic University of Catalonia.
Papers
More filters
Patent
Reconfigurable processing unit
Marc Lupon,Enric Herrero Abellanas,Ayose Falcón,Fernando Latorre,Pedro López,Frederico Pratas +5 more
TL;DR: In this paper, a processor core and a number of calculation modules that each is configurable to perform any one of operations for a convolutional neuron network system are configured to perform convolution operations, averaging operations and dot product operations.
Proceedings ArticleDOI
Combining Simulation and Virtualization through Dynamic Sampling
TL;DR: A novel dynamic sampling mechanism that overcomes this problem and enables the use of VMs for timing simulation, and allows unmodified OS and applications over emulated hardware at near-native speed, yet providing a way to insert timing measurements that yield a final accuracy similar to state-of-the-art sampling methods.
Patent
Processing device for performing convolution operations
Enric Herrero Abellanas,Marc Lupon,Ayose Falcón,Frederico Pratas,Fernando Latorre,Pedro Lopez +5 more
TL;DR: In this paper, a convolutional filter is applied to a plurality of input data elements represented by a two-dimensional array, with the convolver unit comprising a multipliers coupled to two or more sets of latches.
Proceedings ArticleDOI
Author retrospective for software trace cache
TL;DR: This work evaluates and analyzes in detail the impact of the STC, and code layout optimizations in general, on the three main aspects of fetch performance; the instruction cache hit rate, the effective fetch width, and the branch prediction accuracy.
Journal ArticleDOI
Prophet/Critic Hybrid Branch Prediction
TL;DR: The prophet/critic hybrid conditional branch predictor, which has two component predictors that play the role of either prophet or critic, shows an 8K + 8K byte prophet/Critic hybrid has 39% fewer mispredicts than a 16K byte gskew predictor-a predictor similar to that of the proposed Compaq* Alpha* EV8 processor - across a wide range of applications.