Reducing the Read-Miss Penalty for Flat COMA Protocols
نویسندگان
چکیده
In #at cache-only memory architectures (COMA), an attraction-memory miss must "rst interrogate a directory before a copy of the requested data can be located, which often involves three network traversals. By keeping track of the identity of a potential holder of the copycalled a hint one network traversal can be saved which reduces the read penalty. We have evaluated the reduction of the read-miss penalty provided by hints using detailed architectural simulations and four benchmark applications. The results show that a previously proposed protocol using hints can actually make the read-miss penalty larger because when the hint is not correct, an extra network traversal is needed. This has motivated us to study a new protocol, using hints, that simultaneously sends a request to the potential holder and to the directory. This protocol reduces the read-miss penalty for all applications but the protocol complexity does not seem to justify the performance improvement.
منابع مشابه
Using hints to reduce the read miss penalty for flat COMA protocols
In pat COMA architectures, an attraction-memory miss must first interrogate a directory before a copy of the requested data can be located which often involves three network traversals. By keeping track of the identity of a potential holder of the copy-called a hint-one network traversal can be saved which reduces the read penalty. We have evaluated the reduction of the read miss penalty provid...
متن کاملEvaluation of a Competitive-Update Cache Coherence Protocol with Migratory Data Detection
col is very simple. Instead of a copy of the block being invalidated at the first write by another processor, the copy is updated. If the local processor does not access the copy, it is invalidated after a number of global updates determined by a competitive threshold. As a result, only those copies regularly accessed are updated. Although competitive-update protocols have better performance, t...
متن کاملArchitectural Support for an Efficient Implementation of a Software-Only Directory Cache Coherence Protocol
Software-only directory cache coherence protocols emulate directory management by handlers executed on the compute processor in shared-memory multiprocessors. While their potential lies in lower implementation cost and complexity than traditional hardware-only directory protocols, the miss penalty for cache misses induced by application data accesses as well as directory accesses is a critical ...
متن کاملEE 8365 , Advanced Computer Architecture , Spring 2000 ORL – Modified L 2 Cache Replacement Algorithm
A limit to computer system performance is the miss penalty for fetching data and instructions from lower levels in the memory hierarchy. There are two approaches to reducing this penalty. The first approach is to reduce the miss rates of the higher cache levels by utilizing an effective replacement policy that does not replace data that is going to be needed. The second approach is to reduce th...
متن کاملExcel-NUMA: Toward Programmability, Simplicity, and High Performance
ÐWhile hardware-coherent scalable shared-memory multiprocessors are relatively easy to program, they still require substantial programming effort to deliver high performance. Specifically, to minimize remote accesses, data must be carefully laid out in memory for locality and application working sets carefully tuned for caches. It has been claimed that this programming effort is less necessary ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Comput. J.
دوره 40 شماره
صفحات -
تاریخ انتشار 1997