K
Kaushik Rajan
Researcher at Microsoft
Publications - 48
Citations - 657
Kaushik Rajan is an academic researcher from Microsoft. The author has contributed to research in topics: Cache & Iterative reconstruction. The author has an hindex of 13, co-authored 48 publications receiving 600 citations. Previous affiliations of Kaushik Rajan include Istituto Italiano di Tecnologia & Indian Institute of Science.
Papers
More filters
Proceedings ArticleDOI
Emulating Optimal Replacement with a Shepherd Cache
TL;DR: This work proposes a novel replacement strategy that mimics the replacement decisions of OPT and can cover 40% of the gap between OPT and LRU for a 2MB cache resulting in 7% overall speedup.
Journal ArticleDOI
Probabilistic shared cache management (PriSM)
TL;DR: The proposed Probabilistic Shared Cache Management (PriSM), a framework to manage the cache occupancy of different cores at cache block granularity by controlling their eviction probabilities, requires only simple hardware changes to implement, can scale to larger core count and is flexible enough to support a variety of performance goals.
Proceedings ArticleDOI
PerfOrator: eloquent performance models for Resource Optimization
TL;DR: PerfOrator is a novel approach to resource-to-performance modeling that employs nonlinear regression on profile runs to model arbitrary UDFs, calibration queries to generalize across hardware platforms, and analytical framework models to account for parallelism.
Proceedings ArticleDOI
NUcache: An efficient multicore cache organization based on Next-Use distance
TL;DR: A new PC-centric cache organization, NUcache, is proposed, which logically partitions the associative ways of a cache set into MainWays and DeliWays, and is shown to be more effective than other well-known cache-partitioning algorithms.
Journal ArticleDOI
Improving Image Quality in Electrical Impedance Tomography (EIT) Using Projection Error Propagation-Based Regularization (PEPR) Technique: A Simulation Study
TL;DR: A Projection Error Propagation-based Regularization (PEPR) method is proposed in this article to improve the reconstructed image quality in Electrical Impedance Tomography (EIT).