Home | Links | Contact Us | More About Intellectual Property | Bookmark
Search patents:
Home File Sharing Distributed-processing-memory-chip-with-embedded-logic-having-both-data-memory-and-broadcast-memory

 Creation and use of control information associated with packetized network data by protocol drivers and device drivers
OF THE PREFERRED EMBODIMENTS As used herein, the term "software component" refers to any set of ...


 Reconfigurable computing architecture for providing pipelined data paths
The present invention is a reconfigurable data path whose functionality is controlled by a ...


 Multiple thread multiple data predictive coded parallel processing system and method
Accordingly, it is an object of the present invention to process large quantities of data with a ...


 Distributed data dependency stall mechanism
The invention resides in allowing each probe queue in a multiprocessor computer system to be ...


 Method and apparatus for arbitrating between command streams
The present invention provides a method and apparatus for arbitrating between command streams. In ...


 Computer with remote wake up and transmission of a status packet when the computer fails a self test
The failure of a component of a computer upon the awakening of the computer is brought to the ...


 Low power, high speed communications bus
The solution to the problem of data communication between a CPU and a memory where the internal ...


 Apparatus and method for handling multiple mergeable misses in a non-blocking cache
In light of the above, therefore, according to a broad aspect of the invention, disclosed herein is ...


 Method and apparatus for maintaining one or more queues of elements such as commands using one or more token queues
The present invention provides a method and apparatus for maintaining one or more queues of ...


 Method and system for executing a program within a multiscalar processor by processing linked thread descriptors
The multiscalar processing paradigm disclosed herein overcomes numerous deficiencies of the ...


 Distributed processing memory chip with embedded logic having both data memory and broadcast memory

Details
Inventors: Mahant-Shetti, Shivaling S.; Smith, Derek J.; Pawate, Basavaraj I.; Doddington, George R.; Bean, Warren L.; Harward, Mark G.; Aton, Thomas J.;
Assignee: Texas Instruments Incorporated (Dallas, TX)
Primary Examiner: Robertson; David L.
Assistant Examiner:
Attorney, Agent or Firm: Garner; Jacqueline J., Brady, III; W. James, Donaldson; Richard L.

Memory chips with data memory (202), embedded logic (206) and broadcast memory (204) for two modes of operation are disclosed. A first mode of operation is the usual memory mode expected of a data RAM. The second mode of operation allows localized computation and/or processing of the data in data memory (202) by the embedded logic (206) with minimal handshaking with a remote CPU. In a functioning system, the memory chips are organized in a hierarchical manner and include address-associative memory systems.

DETAILED DESCRIPTION The present invention provides computer memory chips which include internal computational logic and secondary memory that may be broadcast to in all chips simultaneously and systems with such memories.
The memory chips may have a standard package pin configuration.
The architecture allows easy expansion of the number of parallel Processing Elements by simply increasing the number of chips and permits massively parallel processing.
The chips may also operate as conventional RAMs.
(Random Access Memories).
This solves known system problems by minimizing the need for board level data movement.
The chips achieve this by combining storage with computation and reducing the amount of data sent to the main CPU.
A first order estimate of system throughput increase is given by the ratio of raw data to "information" data.
For example, in the case of matrix multiplication (to be described in detail later), assume that two bytes of information is the result obtained from 256 raw bytes.
Thus, the system improvement is evident in that only two bytes of information are brought to the CPU rather than 256 raw numbers.
That is, an improvement by a factor of 128.
Depending on the amount of data path logic included in the chips and the number of chips included in the system, the number of MIPS and the throughput improvement may be orders of magnitude higher than the 128 of the previous example.
The design of the chip allows easy expansion to provide the desired amount of distributed parallel processing power by simply increasing the number of chips.
The economics of the commodity memory business force participants to maximize device density while minimizing process and package costs.
The transfer of computational workload from the CPUs to the memories takes advantage of memory manufacturing discipline while greatly reducing the required CPU bus bandwidth.
The preferred embodiment design also allows the system designer flexibility in setting processing power vs cost as the number of processing elements is defined by the memory chip count



Related patents
  Method and apparatus for locally generating addressing information for a memory access
The present invention overcomes many of the disadvantages of the prior art by providing a system whereby a processor or the like need not provide an address to a memory ...
  System for managing the transfer of data between FIFOs within pool memory and peripherals being programmable with identifications of the FIFOs
Thus, a general purpose of the present invention is to provide a peripheral I/O controller supporting multiple, parallel variable bandwidth data streams over a high ...
  Interface device connected between a LAN and a printer for outputting formatted debug information about the printer to the printer
The aforementioned problems are addressed by the present invention in which formatted debug information is output by a network device to an image forming apparatus to ...
  Asynchronous transfer mode based service consolidation switch
From the foregoing, a need has arisen for a telecommunications switch that integrates a variety of services through an asynchronous transfer mode based operation. In ...
  Queue ordering for memory and I/O transactions in a multiple concurrent transaction computer system
Embodiments of the present invention provide a transaction ordering mechanism for processor-based computing systems which ensures proper ordering of transactions between ...
  Method and apparatus for controlling configuration memory contexts of processing elements in a network of multiple context processing elements
A method and apparatus for providing local control of processing elements in a network of multiple context processing element are provided. According to one aspect of ...
  System and method for queuing of tasks in a multiprocessing system
OF THE INVENTION The block diagram of FIG. 1 illustrates a multi-node network 10 configured as a disk drive controller array. Nodes A and D are data storage nodes that ...
  System for flushing queued memory write request corresponding to a queued read request and all prior write requests with counter indicating requests to be flushed
The problems outlined above are in large part solved by a device and method for improving memory bus efficiency. The device includes an interface unit coupled between a ...
  Load and store unit for a vector processor
A load/store unit within a vector processor services memory requests to load or store vectors (multiple data elements of a substantially similar type). Such an apparatus ...
  Smart buffer size adaptation apparatus and method
The present invention is directed toward a data communication system, comprising a frame processing device having a driver and a processor for executing the driver; a LAN...

0.024

Archive: All patents - Links

Copyright (c)2006 Eipa-patents.org - All rights reserved