A concurrent system consists of processes communicating via shared objects, such as shared variables, queues, etc. The concept of wait.freedom was introduced to cope with proce3J failures: each process that accesses a wait-free object is guaranteed to get a response even if all the other processes crash. But what if these wait-free objects themselves fail? For example, if a wait-free object "crashes", all the processes that access that object are prevented from making progress. In this paper, we introduce the concept of fault.tolerant wait-flee objects, and study the problem of implementing them. We give a universal method to construct fault-tolerant waR-free objects, for all types of "responsive " failures (including one in which faulty objects may "lie"). In sharp con-trast, we prove that many common and interesting object types (such as queues, sets, and test&set) have no fault-tolerant wait-free implementations even under the most benign of the "non-responsive " types of failure. We also introduce several concepts and techniques that are central to the design of fault-tolerant concurrent systems: the con-cepts of self-implementation and graceful degradation, and techniques to automatically increase the fault-tolerance of implementations. We prove matching lower bounds on the resource complexity of most of our algorithms.

To exploit instruction level parallelism, it is important not only to execute multiple memory references per cycle, but also to reorder memory references, especially to execute loads before stores that precede them in the sequential instruction stream. To guarantee correctness of execution in such situations, memory reference addresses have to be disambiguated. This paper presents a novel hardware mechanism, called an Address Resolution Buffer (ARB), for performing dynamic reordering of memory references. The ARB supports the following features: (i) dynamic memory disambiguation in a decentralized manner, (ii) multiple memory references per cycle, (iii) out-of-order execution of memory references, (iv) unresolved loads and stores, (v) speculative loads and stores, and (vi) memory renaming. The paper presents the results of a simulation study that we conducted to verify the efficacy of the ARB for a superscalar processor. The paper also shows the ARB's application in a multiscalar proce...

this paper was supported by a gift from the Systems Development Foundation. The research was done as part of my doctoral thesis [19], which was supported by a gift from the Systems Development Foundation, by an IBM Graduate Fellowship, by the Defense Advanced Research Projects Agency under Contract

, California 94704 z Computer Science Division, U.C. Berkeley, Berkeley, California 94720, Supported by an IBM Graduate Fellowship and NSF Grant No. CCR 88-13632. the original running time of P . We present general techniques for constructing simple to program selftesting /correcting pairs for a variety

highly effective algorithms for solution of various classes of linear programs. Linear programming represents one of the major achievements of the operations research and mathematical programming community. Supported in part by a National Science Foundation Graduate Fellowship. In this paper we

and Rose Goldman Career Development Chair, Dept. of Applied Mathematics and Computer Science, Weizmann Institute of Science, Rehovot 76100, Israel. Work performed while at the IBM Almaden Research Center. Research supported by an Alon Fellowship and a grant from the Israel Science Foundation administered

Twenty years ago, Calabi and Yau each proved that a complete noncompact Riemannian manifold with nonnegative Ricci curvature must have at least linear volume growth [Yau]. This was proven by studying the Busemann function, b = bγ, associated with a ray, γ, b(x) = lim (R − d(x,γ(R)). R→∞ In [So2], the author proved that if such a manifold has linear volume growth then its Busemann functions are proper. The simplest examples of manifolds with linear volume growth are the metric product manifolds, X × RI, whose cross sections, X × {r}, are level sets of the Busemann functions. In this paper we prove that a complete noncompact manifold with nonnegative Ricci curvature and linear volume growth must be close to being such a metric product manifold asymptotically [Theorem 34]. That is, as r → ∞, the set b −1 ([r,r +L]) becomes close to b −1 (r) ×[r,r +L] in the Gromov-Hausdorff topology where the closeness depends linearly on diam(b −1 (r)). See Section 2. The proof involves a careful analysis of the Busemann function using the recently-developed Cheeger-Colding Almost Rigidity Theory [ChCo]. We also use this method to prove the following theorem. Theorem 1 If M n is a manifold with nonnegative Ricci curvature and linear volume growth, then it has sublinear diameter growth, diam(b lim R→∞ −1 (R))

, stability issues are discussed, and improvements are proposed for numerical computation. Numerical examples are presented which demonstrate the improvement in accuracy over standard methods. 1 Supported by an IBM graduate fellowship (grote@cims.nyu.edu). 2 Supported in part by AFOSR, NSF, and ONR (keller

The author is supported by an IBM Graduate Fellowship. 1 Introduction Since their introduction by Elman [...

@icsi.berkeley.edu. Research supported by an IBM Graduate Fellowship and by the Internati...