Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
Research Interests:
Recently, much research has gone into wide-area message passing systems, but more work is necessary so that message passing systems can adapt to wide-area environments by themselves and stop requiring manual configuration. Thus, in this... more
Recently, much research has gone into wide-area message passing systems, but more work is necessary so that message passing systems can adapt to wide-area environments by themselves and stop requiring manual configuration. Thus, in this paper, I make two proposals concerning the design and implementation of adaptive message passing systems for wide-area, distributed environments. My first proposal is a method for message passing systems to establish connections and perform routing in a wide-area-aware manner. Given just the list of end points, each process autonomously discovers which connections are blocked by firewalls, and establishes a subset of the connections that are not blocked. In determining which subset to establish, processes use latency information obtained at run-time to limit the number of wide-area connections. Moreover, processes perform routing so that all processes can communicate with each other, even when some nodes are behind firewalls or only have private IP a...
ABSTRACT
Research Interests:
Research Interests:
Research Interests:
Computational complexity theories have been playing central roles in all areas of computer science. Traditionally, computational complexity is based on the random access memory (RAM) model, in which unit amount of data at an arbitrary... more
Computational complexity theories have been playing central roles in all areas of computer science. Traditionally, computational complexity is based on the random access memory (RAM) model, in which unit amount of data at an arbitrary location in the memory can be accessed with some fixed constant cost. Recent advances in information technologies made this assumption unrealistic. The speed gap between the processing units and the main memory has been widening dramatically, demanding for deep memory hierarchies. Recent parallel processing systems have more processors than that can costeffectively share the same memory without cache mechanism. Distributed computation is getting more and more popular. All these demand for a computational complexity model that is more aware of locality. In this paper, we propose a new framework for computational complexity, named access complexity, in which the cost is in the data transfer than in computation itself. The model is designed so that it nat...
In the first three-year development stage of the fifth generation computer systems project, a series of high-performance personal computers called sequential inference machines are being developed at ICOT Research Center. The machines... more
In the first three-year development stage of the fifth generation computer systems project, a series of high-performance personal computers called sequential inference machines are being developed at ICOT Research Center. The machines have a high-level machine language called KL0, which is a PROLOG-based logic language with various extensions. In the software development of the sequential inference machines. ESP, a software-supported
Research Interests:
Research Interests:

And 88 more