publications

2026

1. CGO

   PolyUFC: Polyhedral Compilation Meets Roofline Analysis for Uncore Frequency Capping

   Nilesh Rajendra Shah , M V V S Manoj Kumar , Dhairya Baxi , and 1 more author

   In 2026 IEEE/ACM International Symposium on Code Generation and Optimization (CGO) , 2026

   Abs

   

   We present PolyUFC, an MLIR based compilation flow for uncore frequency capping that combines (performance and power) roofline analyses and polyhedral compilation-based static analysis for characterization of affine programs. We introduce a parametric mathematical model that links operational intensity and uncore frequency to derive frequency caps, validated through empirical evaluation on real hardware. By embedding these caps into Pluto optimized code generated by Polygeist, we achieve improvements in Energy Delay Product (EDP) up to 42% on compute-bound, and up to 54% on bandwidth-bound programs—carefully selected from ML-models from vision/NLP domains and PolyBench—over Intel UFS driver. Our framework is retargetable across multiple micro-architectures; and can handle multiple optimization goals like performance, energy and EDP, and is applicable across inter/intra dialects.

2. LATHC

   PolyUFC: Polyhedral Compilation Meets Roofline Analysis for Uncore Frequency Capping

   Nilesh Rajendra Shah , and Ramakrishna Upadrasta

   In LATHC 2026 workshop (associated with CGO 2026) , 2026

   Informal proceedings.

   Abs

   We present PolyUFC, an MLIR based compilation flow for uncore frequency capping that combines (performance and power) roofline analyses and polyhedral compilation-based static analysis for characterization of affine programs. We introduce a parametric mathematical model that links operational intensity and uncore frequency to derive frequency caps, validated through empirical evaluation on real hardware. By embedding these caps into Pluto optimized code generated by Polygeist, we achieve improvements in Energy Delay Product (EDP) up to 42% on compute-bound, and up to 54% on bandwidth-bound programs—carefully selected from ML-models from vision/NLP domains and PolyBench—over Intel UFS driver. Our framework is retargetable across multiple micro-architectures; and can handle multiple optimization goals like performance, energy and EDP, and is applicable across inter/intra dialects. https://jnamaral.github.io/LATHC26/program/#PolyUFC

2023

1. IMPACT

   GeMS: Towards Generating Millions of SCoPs

   In IMPACT 2023 workshop (associated with HIPEAC 2023) , 2023

   Informal proceedings.

   Abs

   We propose an automatic polyhedral loop generator for Generating Millions of SCoPs/loops (GeMS). In this presentation, we first give a motivation and an outline of our generator. Using GeMS, we generate a labeled set of compute and memory bounded kernels by imposing constraints on cache locality. Our initial loop-generator is able to generate 24K kernels for a given configuration. We then illustrate the characteristics of the generated kernels by studying their IPC statistics. Our experimental evaluation involves using various input parameter sizes on two machines that have different cache configurations. We conclude with a listing of possible applications where GeMS could be useful. We also suggest ways to extend GeMS using various core polyhedral compilation libraries and data structures.

2022

1. TACO

   BullsEye : Scalable and Accurate Approximation Framework for Cache Miss Calculation

   Nilesh Rajendra Shah , Ashitabh Misra , Antoine Miné , and 2 more authors

   ACM Trans. Archit. Code Optim., Nov 2022

   Abs

   

   For Affine Control Programs or Static Control Programs (SCoP), symbolic counting of reuse distances could induce polynomials for each reuse pair. These polynomials along with cache capacity constraints lead to non-affine (semi-algebraic) sets; and counting these sets is considered to be a hard problem. The state-of-the-art methods use various exact enumeration techniques relying on existing cardinality algorithms that can efficiently count affine sets.We propose BullsEye , a novel, scalable, accurate, and problem-size independent approximation framework. It is an analytical cache model for fully associative caches with LRU replacement policy focusing on sampling and linearization of non-affine stack distance polynomials. First, we propose a simple domain sampling method that can improve the scalability of exact enumeration. Second, we propose linearization techniques relying on Handelman’s theorem and Bernstein’s representation. To improve the scalability of the Handelman’s theorem linearization technique, we propose template (Interval or Octagon) sub-polyhedral approximations.Our methods obtain significant compile-time improvements with high-accuracy when compared to HayStack on important polyhedral compilation kernels such as nussinov, cholesky, and adi from PolyBench, and harris, gaussianblur from LLVM-TestSuite. Overall, on PolyBench kernels, our methods show up to 3.31\texttimes (geomean) speedup with errors below ≈ 0.08% (geomean) for the octagon sub-polyhedral approximation.