About

Vikram Adve is the Donald B. Gillies Professor in Computer Science at the University of Illinois Urbana-Champaign. His research areas include compilers, architecture, and parallel computing, as well as programming languages, formal methods, and software engineering. He has contributed to the development of the LLVM compiler infrastructure and has been recognized with awards such as the NSF CAREER Award, ACM Software System Award, and fellowships with the ACM. Adve has also been involved in advancing secure virtual architecture environments and improving program analysis and transformation frameworks. His academic background includes a Ph.D. in Computer Science from the University of Wisconsin-Madison, obtained in 1993.

Research topics

• Computer Science
• Data Mining
• Mathematics
• Arithmetic
• Programming language
• Embedded system
• Algorithm
• Human–computer interaction
• Theoretical computer science
• Operating system
• Database
• Data science
• Engineering

Selected publications

• VTC: DNN Compilation with Virtual Tensors for Data Movement Elimination

  arXiv (Cornell University) · 2026-02-11

  preprintOpen access

  With the widening gap between compute and memory operation latencies, data movement optimizations have become increasingly important for DNN compilation. Current optimizations such as layout transformations and operator fusion only target a subset of tensor operators and consequently miss important opportunities for reducing data movement in contemporary DNN workloads, including large language models. We introduce VTC, a novel tensor compilation framework that for the first time eliminates all unnecessary data movement by targeting the full spectrum of data movement operators. VTC proposes the concept of virtual tensors to track data movement between compute operators via index mappings rather than expensive physical data transfers to and from global memory, which can seamlessly interoperate with existing computation kernels and handle arbitrary tensor operator compositions. We also introduce a novel data movement elimination algorithm to automatically identify a profitable virtual tensor creation strategy. Evaluation on a variety of DNNs shows that VTC can outperform existing ML compilers by up to 1.93x (1.28x on average) on NVIDIA GPUs with up to 60% (17.5% on average) inference memory savings.

  PublisherDOI

• VTC: DNN Compilation with Virtual Tensors for Data Movement Elimination

  arXiv (Cornell University) · 2026-02-11

  articleOpen access

  With the widening gap between compute and memory operation latencies, data movement optimizations have become increasingly important for DNN compilation. Current optimizations such as layout transformations and operator fusion only target a subset of tensor operators and consequently miss important opportunities for reducing data movement in contemporary DNN workloads, including large language models. We introduce VTC, a novel tensor compilation framework that for the first time eliminates all unnecessary data movement by targeting the full spectrum of data movement operators. VTC proposes the concept of virtual tensors to track data movement between compute operators via index mappings rather than expensive physical data transfers to and from global memory, which can seamlessly interoperate with existing computation kernels and handle arbitrary tensor operator compositions. We also introduce a novel data movement elimination algorithm to automatically identify a profitable virtual tensor creation strategy. Evaluation on a variety of DNNs shows that VTC can outperform existing ML compilers by up to 1.93x (1.28x on average) on NVIDIA GPUs with up to 60% (17.5% on average) inference memory savings.

  PublisherOA PDF

• Towards Formal Verification of LLM-Generated Code from Natural Language Prompts

  ArXiv.org · 2025-07-17 · 1 citations

  preprintOpen accessSenior author

  In the past few years LLMs have emerged as a tool that can aid programmers by taking natural language descriptions and generating code based on it. However, the reliability of LLM code generation and current validation techniques for it are far from strong enough to be used for mission-critical or safety-critical applications. In this work we explore ways to offer formal guarantees of correctness to LLM generated code; such guarantees could improve the quality of general AI Code Assistants and support their use for critical applications. To address this challenge we propose to incorporate a Formal Query Language that can represent a user's intent in a formally defined but natural language-like manner that a user can confirm matches their intent. We then have a formal specification of the user intent which we can use to verify that LLM-generated code matches the user's intent. We implement these ideas in our system, Astrogator, for the Ansible programming language, widely used for system administration, including for critical systems. The system includes an intuitive formal query language, a calculus for representing the behavior of Ansible programs, and a symbolic interpreter and a unification algorithm which together are used for the verification. A key innovation in Astrogator is the use of a Knowledge Base to capture system-specific implementation dependencies that greatly reduce the need for system knowledge in expressing formal queries. On a benchmark suite of 21 code-generation tasks, our verifier is able to verify correct code in 83% of cases and identify incorrect code in 92%.

  PublisherOA PDFDOI

• Neptune: Advanced ML Operator Fusion for Locality and Parallelism on GPUs

  ArXiv.org · 2025-10-09

  preprintOpen access

  Operator fusion has become a key optimization for deep learning, which combines multiple deep learning operators to improve data reuse and reduce global memory transfers. However, existing tensor compilers struggle to fuse complex reduction computations involving loop-carried dependencies, such as attention mechanisms. This paper introduces Neptune, a tensor compiler for advanced operator fusion for sequences of reduction operators. Neptune presents a new approach for advanced operator fusion, which intentionally breaks some existing dependencies and compensates by constructing algebraic correction expressions that allow the kernel to produce the correct result. Applying Neptune's advanced operator fusion to a plain attention operator generates operators equivalent to FlashAttention and FlashDecoding. On ten attention-based benchmarks, Neptune, starting from a plain attention code and a high-level scheduling template, outperforms existing compilers like Triton, TVM, and FlexAttention, including Triton-based implementations of FlashAttention. Across four different GPU architectures from NVIDIA and AMD, Neptune-generated kernels have an average speedup of $1.35\times$ over the next best alternative, with up to $2.65\times$ speedup on Nvidia GPUs and up to $3.32\times$ on AMD GPUs, demonstrating its effectiveness for deep learning workloads.

  PublisherOA PDFDOI

• HPVM-HDC: A Heterogeneous Programming System for Accelerating Hyperdimensional Computing

  2025-06-20 · 2 citations

  articleSenior author
  PublisherDOI

• Hercules: A Compiler for Productive Programming of Heterogeneous Systems

  ArXiv.org · 2025-03-13

  preprintOpen accessSenior author

  Modern computing systems increasingly rely on composing heterogeneous devices to improve performance and efficiency. Programming these systems is often unproductive: algorithm implementations must be coupled to system-specific logic, including device-specific optimizations, partitioning, and inter-device communication and synchronization, which requires developing different programs for different system configurations. We propose the Juno language, which represents general purpose applications in an imperative form that can be transformed into parallel, optimized, system-specific code using an expressive and granular imperative scheduling language. We also introduce the Hercules compiler, which uses a novel intermediate representation to represent general and device-specific parallel code in a manner that is easy to analyze and manipulate using schedules. Our system achieves competitive performance with hand-optimized device-specific code (geomean speedups of $1.25\times$ and $1.48\times$ on the CPU and GPU) and significantly outperforms a prior general purpose heterogeneous programming system (geomean speedups of $9.31\times$ and $16.18\times$ on the CPU and GPU).

  PublisherOA PDFDOI

• MISAAL: Synthesis-Based Automatic Generation of Efficient and Retargetable Semantics-Driven Optimizations

  Proceedings of the ACM on Programming Languages · 2025-06-10

  articleOpen accessSenior author

  Using program synthesis to select instructions for and optimize input programs is receiving increasing attention. However, existing synthesis-based compilers are faced by two major challenges that prohibit the deployment of program synthesis in production compilers: exorbitantly long synthesis times spanning several minutes and hours; and scalability issues that prevent synthesis of complex modern compute and data swizzle instructions, which have been found to maximize performance of modern tensor and stencil workloads. This paper proposes MISAAL, a synthesis-based compiler that employs a novel strategy to use formal semantics of hardware instructions to automatically prune a large search space of rewrite rules for modern complex instructions in an offline stage. MISAAL also proposes a novel methodology to make term-rewriting process in the online stage (at compile-time) extremely lightweight so as to enable programs to compile in seconds. Our results show that MISAAL reduces compilation times by up to a geomean of 16x compared to the state-of-the-art synthesis-based compiler, HYDRIDE. MISAAL also delivers competitive runtime performance against the production compiler for image processing and deep learning workloads, Halide, as well as HYDRIDE across x86, Hexagon and ARM.

  PublisherDOI

• MIRAGE: A Benchmark for Multimodal Information-Seeking and Reasoning in Agricultural Expert-Guided Conversations

  ArXiv.org · 2025-06-25

  preprintOpen accessSenior author

  We introduce MIRAGE, a new benchmark for multimodal expert-level reasoning and decision-making in consultative interaction settings. Designed for the agriculture domain, MIRAGE captures the full complexity of expert consultations by combining natural user queries, expert-authored responses, and image-based context, offering a high-fidelity benchmark for evaluating models on grounded reasoning, clarification strategies, and long-form generation in a real-world, knowledge-intensive domain. Grounded in over 35,000 real user-expert interactions and curated through a carefully designed multi-step pipeline, MIRAGE spans diverse crop health, pest diagnosis, and crop management scenarios. The benchmark includes more than 7,000 unique biological entities, covering plant species, pests, and diseases, making it one of the most taxonomically diverse benchmarks available for vision-language models, grounded in the real world. Unlike existing benchmarks that rely on well-specified user inputs and closed-set taxonomies, MIRAGE features underspecified, context-rich scenarios with open-world settings, requiring models to infer latent knowledge gaps, handle rare entities, and either proactively guide the interaction or respond. Project Page: https://mirage-benchmark.github.io

  PublisherOA PDFDOI

• AgMMU: A Comprehensive Agricultural Multimodal Understanding Benchmark

  ArXiv.org · 2025-04-14

  preprintOpen access

  We present AgMMU, a challenging real-world benchmark for evaluating and advancing vision-language models (VLMs) in the knowledge-intensive domain of agriculture. Unlike prior datasets that rely on crowdsourced prompts, AgMMU is distilled from 116,231 authentic dialogues between everyday growers and USDA-authorized Cooperative Extension experts. Through a three-stage pipeline: automated knowledge extraction, QA generation, and human verification, we construct (i) AgMMU, an evaluation set of 746 multiple-choice questions (MCQs) and 746 open-ended questions (OEQs), and (ii) AgBase, a development corpus of 57,079 multimodal facts covering five high-stakes agricultural topics: insect identification, species identification, disease categorization, symptom description, and management instruction. Benchmarking 12 leading VLMs reveals pronounced gaps in fine-grained perception and factual grounding. Open-sourced models trail after proprietary ones by a wide margin. Simple fine-tuning on AgBase boosts open-sourced model performance on challenging OEQs for up to 11.6% on average, narrowing this gap and also motivating future research to propose better strategies in knowledge extraction and distillation from AgBase. We hope AgMMU stimulates research on domain-specific knowledge integration and trustworthy decision support in agriculture AI development.

  PublisherOA PDFDOI

• AIFARMS: Artificial intelligence for future agricultural resilience, management, and sustainability

  AI Magazine · 2024-02-22 · 12 citations

  articleOpen access1st authorCorresponding

  Abstract The AIFARMS Artificial Intelligence for Future Agricultural Resilience, Management, and Sustainability national AI institute brings together over 40 world‐class AI and agriculture researchers, with the common mission to develop foundational advances in AI and use them to ensure that future agriculture is environmentally friendly, sustainable, affordable, and accessible to diverse farming communities. Since its establishment in 2020, AIFARMS has advanced the state of the art in autonomous farming, cover crop planting, machine learning for improved outcomes from remote sensing, dynamic estimation of yield loss from weeds, and livestock management. The institute has prioritized the creation and utilization of high‐quality, openly available data sets for advancing foundational AI and tackling agricultural challenges. AIFARMS leverages a close partnership between UIUC and Tuskegee University to build programming for a skilled and diverse next‐generation workforce in digital agriculture. We are expanding the reach of AIFARMS outside of the current partners to collaborate with national AI institutions and international partners.

  PublisherOA PDFDOI

Recent grants

• CSR: Small: Automated Software Fault Localization via Static and Dynamic Analysis

  NSF · $490k · 2012–2016

• CRI: CRD: A Compiler-Based Virtual Machine Infrastructure for System Software

  NSF · $500k · 2007–2014

• SHF: Medium: Programmability, Portability, Performance and Energy Efficiency for Heterogeneous Systems

  NSF · $900k · 2013–2017

• Cooperative Hardware/Software Designs for Virtual Instruction Set Computers

  NSF · $150k · 2004–2007

• TWC: Small: Trustworthy Code Generation for the LLVM Compiler Infrastructure

  NSF · $500k · 2016–2020

Frequent coauthors

• Sarita V. Adve

  29 shared

• Chris Lattner

  Google (United States)

  21 shared

• Stephen Heumann

  University of Illinois Urbana-Champaign

  16 shared

• John Criswell

  University of Rochester

  14 shared

• Dinakar Dhurjati

  Qualcomm (United States)

  14 shared

• Robert L. Bocchino

  Jet Propulsion Laboratory

  13 shared

• Swarup Kumar Sahoo

  13 shared

• Mohsen Vakilian

  11 shared

Labs

• Siebel School of Computing and Data SciencePI

Education

• Ph.D., Computer Science

  Massachusetts Institute of Technology

  1992

• M.S., Computer Science

  Massachusetts Institute of Technology

  1988

• B.S., Electrical Engineering and Computer Science

  University of California, Berkeley

  1986

Awards & honors

• NSF CAREER Award (2001)
• C. W. Gear Outstanding Junior Faculty Award, Computer Scienc…
• ACM Software System Award (2012)
• Fellow of the Association for Computing Machinery (ACM) (201…
• Illinois Innovation Award (2026)