We focus on finding novel approaches to improve the performance of modern computer systems without unduly increasing the complexity faced by application developers, compiler writers, or computer architects.
We research novel approaches to improve performance and increase the capabilities of modern computer systems without unduly increasing the complexity faced by application developers, compiler writers, or computer architects. We look to understand and predict the changes in user requirements and underlying technology, extrapolate their stresses on the current evolutionary path, identify possible paradigm shifts that can drastically simplify, improve and expand the system, and test these hypotheses by developing a working prototype within the current constraints. We also believe that building real systems is critical in conducting credible systems research. Over the years, the Commit group has created StreamIt, PetaBricks, StreamJIT, Halide, Simit, MILK, GraphIt and TACO programming languages and compilers, DynamoRIO dynamic instrumentation system, Superword level parallelism for SIMD vectorization, Program Shepherding to protect programs against external attacks, the OpenTuner extendable autotuner, and the Kendo deterministic execution system.
Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady-state (the throughput) is important for both compiler designers and performance engineers.
However, building an analytical model to do so is especially complicated in modern x86-64 Complex Instruction Set Computer (CISC) machines with sophisticated processor microarchitectures in that it is tedious, error-prone, and must be performed from scratch for each processor generation.
Ithemal is the first tool that learns to predict the throughput of a set of instructions. It does so more accurately than state-of-the-art hand-written tools currently used in compiler backends and static machine code analyzers. In particular, Ithemal has less than half the error of state-of-the-art analytical models (LLVM's llvm-mca and Intel's IACA).
Simit is a new programming language that makes it easy to efficiently compute on sparse systems using linear algebra. Simit programs are typically shorter than MATLAB programs and yet are comparable in performance to hand-optimized codes.
