Talk Title: Unbounded Transactional Memory Speaker: Bradley C. Kuszmaul Date: 10/5/2005 Abstract: In this talk, I argue that hardware transactional memory should support *unbounded transactions*: transactions of arbitrary size and duration. Our team designed a hardware implementation of unbounded transactional memory, called UTM, which exploits the common case for performance without sacrificing correctness on transactions whose footprint can be nearly as large as virtual memory. We also designed a "smaller" version, called LTM, that scales only up to the size of physical memory but is pin-compatible with today's microprocessors. We assessed UTM and LTM through microbenchmarking and by automatically converting the SPECjvm98 Java benchmarks and the Linux 2.4.19 kernel to use transactions instead of locks. We use both cycle-accurate simulation and instrumentation to understand benchmark behavior. Our studies show that the common case is small transactions that commit, even when contention is high, but that some applications contain very large transactions. For example, although 99.9% of transactions in the Linux study touch 54 cache lines or fewer, some transactions touch over 8000 cache lines. Our studies also indicate that hardware support is required, because some applications spend over half their time in critical regions. Finally, they suggest that hardware support for transactions can make Java programs run faster than when run using locks and can increase the concurrency of the Linux kernel by as much as a factor of 4 with no additional programming work. Bio: Bradley C. Kuszmaul is a research scientist in the Supercomputing Technologies Group at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). He is one of the developers of the Cilk multithreaded programming environment, is the coauthor of several prize-winning parallel chess programs, and developed a theory of asymptotically optimal circuits for out-of-order superscalar processors. He was one of the principal architects of the Connection Machine CM-5 supercomputer at Thinking Machines Corporation, and lead the development the Akamai Technologies network usage database. He is currently investigating cache-oblivious data structures and transactional memory.