Kabilan Mahathevan

PhD Student @VT CS

prof_pic.jpg

Hi! I’m Kabilan Mahathevan, a PhD student at Virginia Tech’s Computer Science department. I am fortunate enough to be advised by Prof. Kirshanthan (“Krish”) Sundararajah. My research focuses on building tools and techniques that make tensor computations efficient, reliable, and practical for real-world computational workloads.

I’m currently focused on sparse tensor computation, where most entries of a tensor are zero, as is common in scientific simulations, graph analytics, and machine learning workloads. Because so much of the data can be skipped, sparse tensors are stored in compressed formats and processed with irregular, data-dependent control flow, which makes them notoriously difficult for compilers and hardware to optimize. My work explores how to make the compilers targeting these workloads faster and more trustworthy.

If you’re interested in this area or would like to discuss related research, feel free to reach out.

Outside of research, I watch an unhealthy amount of animeCompleted Anime List (Incomplete)Attack on TitanKimetsu no Yaiba(Demon Slayer)*JUJUTSU KAISENDeath NoteMy Hero AcademiaHunter x HunterOne-Punch ManOne PieceTokyo GhoulAttack on Titan Season 2Fullmetal Alchemist: BrotherhoodNarutoA Silent VoiceYour Name.Chainsaw ManAssassination ClassroomMob Psycho 100Your lie in AprilNaruto: ShippuudenSteins;GateBlack CloverHaikyuuSPY x FAMILYDr. STONEVinland SagaThe Seven Deadly SinsBleachCode Geass: Lelouch of the RebellionNeon Genesis EvangelionFrieren: Beyond Journey’s EndFire ForceSpirited AwayCowboy BebopSolo LevelingTokyo RevengersMy Dress-Up DarlingDAN DA DANFairy TailMonsterBlue LockBANANA FISHDragon BallGrave of the FirefliesSuzumeWolf ChildrenA Whisker AwayPonyoBerserkThe Girl Who Leapt Through TimeTrigunHajime no Ippo: The Fighting!BAKIRecord of RagnarokWhisper of the HeartSailor MoonRurouni Kenshin and play a little badminton!

News

Nov 02, 2026 🏆 Splyce’s artifact won the Reproducibility Challenge by Chameleon Cloud at PACT’26 and was featured in their October newsletter.
Oct 28, 2026 🧪 Joined the artifact evaluation committee of POPL’27 and PPoPP’27
Oct 09, 2026 🎤 Gave an invited talk on Splyce: SIMD Vectorization of Sparse Coiteration at the PurPL seminar, Purdue University.
Aug 05, 2026 📄 Our paper Splyce: SIMD Vectorization of Sparse Coiteration got accepted to PACT 2026.
May 27, 2026 🏅 I have been awarded with Pratt Fellowship by the Computer Science Department at Virginia Tech for the 2026–2027 academic year.
May 11, 2026 📄 Our Vectorization for Sparse Coiteration Work has been accepted at ARRAY'26 (Co-located \w PLDI'26)
Feb 27, 2026 🎤 Presented TENSURE: Fuzzing Sparse Tensor Compilers at the Fuzzing Workshop (FUZZING’26), co-located with NDSS 2026 in San Diego.
Feb 23, 2026 📄 Our paper Dialect-Agnostic SQL Parsing via LLM-Based Segmentation has been accepted to SIGMOD 2026.
Jan 29, 2026 🎤 Gave an invited talk on TenSure: Fuzzing Sparse Tensor Compilers at the PurPL seminar, Purdue University.
Jan 16, 2026 📄 Our paper TENSURE: Fuzzing Sparse Tensor Compilers got accepted to Fuzzing’26 (NDSS) 2026.

Latest Posts

Selected Publications

  1. PACT 2026
    ACM badge: Artifacts Available ACM badge: Artifacts Evaluated - Functional ACM badge: Results Reproduced
    Splyce: SIMD Vectorization of Sparse Coiteration
    Kabilan Mahathevan, Poorna Gunathilaka, and Kirshanthan Sundararajah
    In Proceedings of the 35th ACM International Conference on Parallel Processing Techniques and Applications (PACT 2026), 2026
    🏆 Winner, Chameleon Reproducibility Challenge
Lived & worked Conferences & events Fun visits