About Me
I am a Senior Runtime Engineer at Intel, where I am a member of the Communications Library team, developing and optimizing runtime software for AI workloads and high-performance computing. Before that, I was a research scientist in Intel's Tech, Pathfinding and Innovation (TPI) group.
I was a postdoctoral scholar in Computer Science at Stanford University, and worked with Professor Alex Aiken on scalable task-based programming systems.
I received my Ph.D. in Computer Science from the University of Illinois at Urbana-Champaign in 2020, advised by Professors David Padua and William Gropp. During the Ph.D., I developed Locus, a new system, and a language for optimizing complex, long-lived applications for different platforms. It can ameliorate the difficulty of optimizing applications using a methodology based on optimization programming and automated empirical search. The Locus language allows the definition of a search space combined with the programming of optimization sequences separated from the application’s reference code. Based on a Locus program, the system automatically selects, generates, and executes candidate implementations to find the one with the best performance. I was also part of the XPACC project.
During the summer of 2017, I headed off to Seattle, where I joined the Programming Systems and Applications Research Group at Nvidia Research. At Nvidia, I worked on a code optimizer for deep neural networks.
In 2015, I spent the summer at the Lawrence Livermore National Laboratory. My research mostly focused on understanding how to use machine learning to explore the space of code optimizations efficiently.
I have also spent some good years researching and developing parallel geophysical algorithms at Petrobras. I have been part of the Geophysical Technology research group and, in 2009, helped to successfully deploy in production one of the first heterogeneous clusters with GPUs for seismic processing.
All the journey started off at UFMG in Brazil, where I have received my BSc and MSc in Computer Science advised by Professor Wagner Meira. There, I have had the opportunity to join the Speed Lab and develop efficient parallel runtime systems for data mining algorithms, also mentored by Professors Dorgival Guedes, Renato Ferreira.
Experience
Senior Runtime Engineer, Communications Library team, Intel (current)
Research Scientist, Tech, Pathfinding and Innovation (TPI), Intel
Postdoctoral Scholar, Computer Science, Stanford University
Ph.D. in Computer Science, University of Illinois at Urbana-Champaign, 2020
Research Intern, Nvidia Research, 2017
Summer Intern, Lawrence Livermore National Laboratory, 2015
Geophysical Technology Research Group, Petrobras
B.Sc. and M.Sc. in Computer Science, UFMG, Brazil
Awards
Bronze Medal in the ACM Student Research Competition at PACT 2017.
Best Master's Thesis Award in 2011 selected by the Commission of Computer Architecture and High Performance Computing of Brazilian Computer Society.
Top Ten Best Master's Thesis of 2010 selected by the Brazilian Computer Society.
Second place on Second Marathon of Parallel Programming at SBAC-PAD'07.
WSCAD'05 Best paper award.
Selected Publications
VeriCoder: Enhancing LLM-Based RTL Code Generation through Functional Correctness Validation
Anjiang Wei, Huanmi Tan, Tarun Suresh, Daniel Mendoza, Thiago S. F. X. Teixeira, Ke Wang, Caroline Trippel, Alex Aiken
DL4C Workshop @ NeurIPS, 2025, San Diego, CA, USA
CodeARC: Benchmarking Reasoning Capabilities of LLM Agents for Inductive Program Synthesis
Anjiang Wei, Tarun Suresh, Jiannan Cao, Naveen Kannan, Yuheng Wu, Kai Yan, Thiago S. F. X. Teixeira, Ke Wang, Alex Aiken
COLM, 2025, Montreal, Canada
EquiBench: Benchmarking Large Language Models' Reasoning about Program Semantics via Equivalence Checking
Anjiang Wei, Jiannan Cao, Ran Li, Hongyu Chen, Yuhui Zhang, Ziheng Wang, Yuan Liu, Thiago S. F. X. Teixeira, Diyi Yang, Ke Wang, Alex Aiken
EMNLP, 2025, Suzhou, China (also DL4C Workshop @ NeurIPS, 2025, Oral)
Improving Parallel Program Performance with LLM Optimizers via Agent-System Interface
Anjiang Wei, Allen Nie, Thiago S. F. X. Teixeira, Rohan Yadav, Wonchan Lee, Ke Wang, Alex Aiken
ICML, 2025, Vancouver, Canada
Automated Mapping of Task-Based Programs onto Distributed and Heterogeneous Machines
PDF file
Slides
/code repo/
Thiago S. F. X. Teixeira, Alexandra Henzinger, Rohan Yadav, Alex Aiken
Supercomputing, 2023, Denver, CO, USA
Task-based Framework for Physics-based Ensemble Simulation and in Situ Data Processing
Paper Link
K Maeda, T Teixeira
Center for Turbulence Research Annual Research Briefs, 2022, Stanford, CA, USA
Online Multimedia Retrieval on CPU-GPU Platforms with Adaptive Work Partition
Rafael Souza, André Fernandes, Thiago S. F. X. Teixeira, George Teodoro, Renato Ferreira
JPDC, 2021
A Language and a System for Program Optimization
Thiago S. F. X. Teixeira
Ph.D. Dissertation, 2020, University of Illinois at Urbana-Champaign, IL, USA
Managing code transformations for better performance portability
PDF file
Thiago S. F. X. Teixeira, William Gropp, David Padua
IJHPCA, 2019
Locus: A System and a Language for Program Optimization
PDF file
Slides
Video
/code repo/
/loop optimizations repo/
Thiago S. F. X. Teixeira, Corinne Ancourt, David Padua, William Gropp
CGO, 2019, Washington DC, USA.
A DSL for Performance Orchestration (short paper)
Thiago S. F. X. Teixeira, David Padua, William Gropp
26th PACT, 2017, Portland, OR, USA
Large-Scale Distributed Locality-Sensitive Hashing for General Metric Data
E. Silva, T. Teixeira, G. Teodoro, E. Valle
7th SISAP, 2014, Los Cabos, Mexico
Reverse Time Migration with Manycore Coprocessors
T. Teixeira, P. Souza, L. Borges, A. Neto, T. Philippe, C. Andreolli
76th EAGE Conference And Exhibition, 2014, Amsterdam, Netherlands
Accelerating Time and Depth Seismic Migration by CPU and GPU Cooperation
J. Panetta, T. S. F. X. Teixeira, P. Souza, et al
IJPP, 2011
A Scalable Parallel Deduplication Algorithm
PDF file
W. Santos, T. Teixeira, C. Machado, W. Meira Jr., et al
19th SBAC-PAD, 2007, Gramado, RS, Brazil