About: BackgroundWork   Sponge Permalink

An Entity of Type : owl:Thing, within Data Space : 134.155.108.49:8890 associated with source dataset(s)

Many influences have been drawn upon in this work. Around 1987 Jason Cozens worked with Ivor Catt at Sinclair Research. The work involved looking at ways to program the Catt Spiral. Some of the things looked at were the use of CSP (Hoare, 1985) to implement Fast Fourier Transforms. Catt split the wafer into cells. A cell would be selected on the edge of the wafer and this would probe its neighbouring cells to find one that was working. When a working cell was found this cell would then probe its neighbours. In this way bad cells would be isolated and a chain of good cells would be established.

AttributesValues
rdfs:label
  • BackgroundWork
rdfs:comment
  • Many influences have been drawn upon in this work. Around 1987 Jason Cozens worked with Ivor Catt at Sinclair Research. The work involved looking at ways to program the Catt Spiral. Some of the things looked at were the use of CSP (Hoare, 1985) to implement Fast Fourier Transforms. Catt split the wafer into cells. A cell would be selected on the edge of the wafer and this would probe its neighbouring cells to find one that was working. When a working cell was found this cell would then probe its neighbours. In this way bad cells would be isolated and a chain of good cells would be established.
dcterms:subject
abstract
  • Many influences have been drawn upon in this work. Around 1987 Jason Cozens worked with Ivor Catt at Sinclair Research. The work involved looking at ways to program the Catt Spiral. Some of the things looked at were the use of CSP (Hoare, 1985) to implement Fast Fourier Transforms. The Catt Spiral was a way of addressing the problem of flaws in silcon wafers. A major factor limiting the size of silicon chips is the fact that the probability of a flaw on a silicon chip increases with square of the width of the chip (assuming a square chip). Therefore doubling the width of a chip increases the probability of the chip failing by a factor of 4. This makes the chance of using a whole wafer with today's manufaturing techniques very small. {NOTE: This is a very naive model of flaws , see (Mead, Conway, 1980) for a less naive model.} Catt split the wafer into cells. A cell would be selected on the edge of the wafer and this would probe its neighbouring cells to find one that was working. When a working cell was found this cell would then probe its neighbours. In this way bad cells would be isolated and a chain of good cells would be established. The major problem in programming the Catt Spiral was the limitations of the architecture and the I/O bottleneck. Catt later (Circa 1990) went on to propose The Kernel Logic Machine. This extended the spiral idea to connect up a mesh network of processors. There seems to have been limited interest in this at the time. Also during the late 1980s Jason Cozens was researching into Adaptable Processor Systems. This work was influenced by SUN's Network extensible Window System NeWS and looked at developing formal models of such systems. As part of this work systems were modelled using CHOCS, a Calculus of Higher Order Communicating Systems, developed by Bent Thomsens (Thomsen, 1989). A number of fruitful discussions and meetings were had with Bent Thomsen. Eager Queue Protocols were started around 2004 (See OpenEd-Lab4 for some early ideas) as a way of monitoring distributed processes as part of the Child Trust Fund project.
Alternative Linked Data Views: ODE     Raw Data in: CXML | CSV | RDF ( N-Triples N3/Turtle JSON XML ) | OData ( Atom JSON ) | Microdata ( JSON HTML) | JSON-LD    About   
This material is Open Knowledge   W3C Semantic Web Technology [RDF Data] Valid XHTML + RDFa
OpenLink Virtuoso version 07.20.3217, on Linux (x86_64-pc-linux-gnu), Standard Edition
Data on this page belongs to its respective rights holders.
Virtuoso Faceted Browser Copyright © 2009-2012 OpenLink Software