Transactional Memory Coherence and Consistency

Hammond, Lance; Wong, Vicky; Chen, Mike Y.; Carlstrom, Brian D.; Davis, John D.; Hertzberg, Ben; Prabhu, Manohar K.; Wijaya, Honggo; Kozyrakis, Christos; Olukotun, Kunle

doi:10.1145/1028176.1006711

Cited by 197 publications

(114 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Interestingly, these features are quite similar to nondeterministic chunk-based systems such as TCC [13] and BulkSC [10]. In these systems, if a chunk is squashed, it is re-executed and retried for commit later, without affecting other processors; furthermore, conflict-free chunks can always commit regardless of the behavior of other processors.…”

Section: Bulkcompactor: Minimizing Squash Delaymentioning

confidence: 91%

“…The components of the design include constructing a deterministic postponement detector (Section 4.1), ensuring deterministic chunk building (Section 4.2), and enforcing deterministic conflict detection (Section 4.3). Supporting BulkCompactor and BulkCompactor-S in other chunkbased platforms such as TCC [13] or the software-based Grace system [5] may involve somewhat different issues. In the following, we describe the three components.…”

Section: Designmentioning

confidence: 99%

See 1 more Smart Citation

BulkCompactor: Optimized deterministic execution via Conflict-Aware commit of atomic blocks

Duan

Zhou

Ahn

et al. 2012

IEEE International Symposium on High-Performance Comp Architecture

View full text Add to dashboard Cite

Recent proposals for determinism-enforcement architectures are able to honor the dependences between threads through a commit step that often becomes a performance bottleneck. As they commit code blocks (or chunks) in a round-robin order, if one chunk gets squashed due to a conflict, its successors also observe a stall. We call this effect transitive squash delay.This paper proposes a novel, high-performance approach to deterministic execution based on Conflict-Aware commit. Rather than committing chunks in strict round-robin order, the idea is to skip those chunks with conflicts and deterministically execute them slightly later. The scheme, called BulkCompactor, largely eliminates transitive squash delay, "compacts" the chunk commits, and substantially speeds-up execution. With BulkCompactor, the squash overhead is O(N ) rather than O(N 2 ) as in round-robin. We describe BulkCompactor designs for machines with centralized or distributed commit. Finally, a simulation-based evaluation shows that BulkCompactor delivers performance comparable to nondeterministic systems. For example, for 32 processors, BulkCompactor incurs an average execution overhead of 22% over a nondeterministic system. The round-robin scheme's average overhead is 133%.

show abstract

Section: Bulkcompactor: Minimizing Squash Delaymentioning

confidence: 91%

Section: Designmentioning

confidence: 99%

BulkCompactor: Optimized deterministic execution via Conflict-Aware commit of atomic blocks

Duan

Zhou

Ahn

et al. 2012

IEEE International Symposium on High-Performance Comp Architecture

View full text Add to dashboard Cite

show abstract

“…Examples of designs that have used this policy are [Hammond et al 2004], [Ananian et al 2005], [Rajwar et al 2005], and [Herlihy and Moss 1993]. These designs mostly use their caches to store the speculative data and in some cases extra buffers or software structures to handle overflows.…”

Section: The Speculative Memorymentioning

confidence: 99%

Energy-Efficient and High-Performance Lock Speculation Hardware for Embedded Multicore Systems

Papagiannopoulou

Capodanno

Moreshet

et al. 2015

ACM Trans. Embed. Comput. Syst.

View full text Add to dashboard Cite

Embedded systems are becoming increasingly common in everyday life and like their general-purpose counterparts, they have shifted towards shared memory multicore architectures. However, they are much more resource-constrained, and as they often run on batteries, energy efficiency becomes critically important. In such systems, achieving high concurrency is a key demand for delivering satisfactory performance at low energy cost. In order to achieve this high concurrency, consistency across the shared memory hierarchy must be accomplished in a cost-effective manner in terms of performance, energy, and implementation complexity. In this paper, we propose EMBEDDED-SPEC, a hardware solution for supporting transparent lock speculation, without the requirement for special supporting instructions. Using this approach, we evaluate the energy consumption and performance of a suite of benchmarks, exploring a range of contention management and retry policies. We conclude that for resource-constrained platforms, lock speculation can provide real benefits in terms of improved concurrency and energy efficiency, as long as the underlying hardware support is carefully configured.

show abstract

“…Transactional memory was initially studied as a hardware architecture [10,23,5]. Software transactional memory [25] is the idea of implementing all transactional semantics in software.…”

Section: Related Workmentioning

confidence: 99%

A Domain Specific Language for Composable Memory Transactions in Java

Bois

Echevarria

2009

Domain-Specific Languages

View full text Add to dashboard Cite

Abstract. In this paper we present CMTJava, a domain specific language for composable memory transactions [7] in Java. CMTJava provides the abstraction of transactional objects. Transactional objects have their fields accessed only by special get and set methods that are automatically generated by the compiler. These methods return transactional actions as a result. A transactional action is an action that, when executed, will produce the desired effect. Transactional actions can only be executed by the atomic method. Transactional actions are first class values in Java and they are composable: transactions can be combined to generate new transactions. The Java type system guarantees that the fields of transactional objects will never be accessed outside a transaction. CMTJava supports the retry and orElse constructs from STM Haskell. To validate our design we implemented a simple transactional system following the description of the original Haskell system. CMTJava is implemented as a state passing monad using BBGA closures, a Java extension that supports closures in Java.

show abstract

Transactional Memory Coherence and Consistency

Abstract: In

Cited by 197 publications

References 40 publications

BulkCompactor: Optimized deterministic execution via Conflict-Aware commit of atomic blocks

BulkCompactor: Optimized deterministic execution via Conflict-Aware commit of atomic blocks

Energy-Efficient and High-Performance Lock Speculation Hardware for Embedded Multicore Systems

A Domain Specific Language for Composable Memory Transactions in Java

Contact Info

Product

Resources

About