Exocompilation for productive programming of hardware accelerators

Y Ikarashi, GL Bernstein, A Reinking, H Genc… - Proceedings of the 43rd …, 2022‏ - dl.acm.org
High-performance kernel libraries are critical to exploiting accelerators and specialized
instructions in many applications. Because compilers are difficult to extend to support …

babble: Learning better abstractions with e-graphs and anti-unification

D Cao, R Kunkel, C Nandi, M Willsey… - Proceedings of the …, 2023‏ - dl.acm.org
Library learning compresses a given corpus of programs by extracting common structure
from the corpus into reusable library functions. Prior work on library learning suffers from two …

Better together: Unifying datalog and equality saturation

Y Zhang, YR Wang, O Flatt, D Cao, P Zucker… - Proceedings of the …, 2023‏ - dl.acm.org
We present egglog, a fixpoint reasoning system that unifies Datalog and equality saturation
(EqSat). Like Datalog, egglog supports efficient incremental execution, cooperating …

Coyote: A compiler for vectorizing encrypted arithmetic circuits

R Malik, K Sheth, M Kulkarni - Proceedings of the 28th ACM International …, 2023‏ - dl.acm.org
Fully Homomorphic Encryption (FHE) is a scheme that allows a computational circuit to
operate on encrypted data and produce a result that, when decrypted, yields the result of the …

A tensor compiler with automatic data packing for simple and efficient fully homomorphic encryption

A Krastev, N Samardzic, S Langowski… - Proceedings of the …, 2024‏ - dl.acm.org
Fully Homomorphic Encryption (FHE) enables computing on encrypted data, letting clients
securely offload computation to untrusted servers. While enticing, FHE has two key …

Rewrite rule inference using equality saturation

C Nandi, M Willsey, A Zhu, YR Wang, B Saiki… - Proceedings of the …, 2021‏ - dl.acm.org
Many compilers, synthesizers, and theorem provers rely on rewrite rules to simplify
expressions or prove equivalences. Develo** rewrite rules can be difficult: rules may be …

Pure tensor program rewriting via access patterns (representation pearl)

GH Smith, A Liu, S Lyubomirsky, S Davidson… - Proceedings of the 5th …, 2021‏ - dl.acm.org
Tensor kernels in machine learning (ML) often correspond to pure mathematical
expressions, making term rewriting an attractive strategy for optimization and map** to …

Equality saturation theory exploration á la carte

A Pal, B Saiki, R Tjoa, C Richey, A Zhu, O Flatt… - Proceedings of the …, 2023‏ - dl.acm.org
Rewrite rules are critical in equality saturation, an increasingly popular technique in
optimizing compilers, synthesizers, and verifiers. Unfortunately, develo** high-quality …

Impress: Large integer multiplication expression rewriting for fpga hls

E Ustun, I San, J Yin, C Yu… - 2022 IEEE 30th Annual …, 2022‏ - ieeexplore.ieee.org
Large integer multiplication is becoming a major challenge for FPGA-based acceleration of
many cryptographic applications. Existing techniques for decomposing and optimizing large …

Automatic generation of vectorizing compilers for customizable digital signal processors

S Thomas, J Bornholt - Proceedings of the 29th ACM International …, 2024‏ - dl.acm.org
Embedded applications extract the best power-performance trade-off from digital signal
processors (DSPs) by making extensive use of vectorized execution. Rather than …