Parallel caching architecture and methods for block-based data processing

Patent metadata

Field	Value
Publication number	US-2016357668-A1
Application number	US-201514728973-A
Country	US
Kind code	A1
Filing date	Jun 2, 2015
Priority date	Jun 2, 2015
Publication date	Dec 8, 2016
Grant date	—

Abstract

A multi-processor computer system with shared memory resources includes a first plurality of sensors configured to acquire inertial and positional data related to a mobile platform. The system further includes a first plurality of co-processors having a hardware logic configured to control the acquisition of the inertial and positional data and configured to analyze the acquired data. The system also includes a second plurality of sensors configured to acquire input data related to the mobile platform connected to a second plurality of co-processors having a hardware logic configured to receive a plurality of streams of input data from the second plurality of sensors and configured to segment the input data into a plurality of discrete data segments. The system also includes a plurality of hardware processing units configured to perform calculations related to the input data using the plurality of data segments.

First claim

What is claimed is:

1 . A multi-processor computer system with shared memory resources, the computer system comprising:
a first plurality of sensors configured to acquire inertial and positional data related to a mobile platform; a first plurality of co-processors communicatively coupled to the first plurality of sensors, the first plurality of co-processors including a hardware logic configured to control the acquisition of the inertial and positional data and configured to analyze the acquired inertial and positional data; a second plurality of sensors configured to acquire input data related to the mobile platform; a second plurality of co-processors communicatively coupled to the second plurality of sensors, the second plurality of co-processors including a hardware logic configured to to receive a plurality of streams of input data from the second plurality of sensors and configured to segment the input data into a plurality of discrete data segments, wherein each data segment includes an element of input data stream and a subset of the inertial and positional data related to the element of input data stream; and a plurality of hardware processing units configured to perform calculations related to the input data using the plurality of data segments.

2 . The multi-processor computer system of claim 1 , wherein the input data comprises image data and wherein the performed calculations comprise image processing calculations.

3 . The multi-processor computer system of claim 1 , wherein the data segment comprises a stride.

4 . The multi-processor computer system of claim 3 , further comprising at least one array of Double Data Rate (DDR) memory blocks and a memory controller communicatively coupled to the at least one array of DDR memory blocks and the plurality of hardware processing units, the memory controller configured to store and retrieve the plurality of strides in/from the at least one array of DDR memory blocks.

5 . The multi-processor computer system of claim 1 , wherein each of the plurality of hardware processing units comprises at least one of an Application Specific Integrated Circuit (ASIC) and a Field Programmable Gate Array (FPGA).

6 . The multi-processor computer system of claim 4 , wherein a stride size is substantially equal to integer multiples of DDR memory block size.

7 . The multi-processor computer system of claim 1 , further comprising a state vector memory comprising a plurality of addressable memory locations and a state vector memory management interface communicatively coupled to the state vector memory and the first and second co-processors, the state vector management interface configured to provide read and write control to the state vector memory.

8 . The multi-processor computer system of claim 4 , further comprising at least one sequencer communicatively coupled to each of the plurality of hardware processing units and to the memory controller, wherein the sequencer is configured to control the calculations performed by the plurality of the hardware processing units.

9 . The multi-processor computer system of claim 2 , further comprising an embedded Ethernet packet routing controller.

10 . A method comprising:
storing state vector data received from a first plurality of sensors in a state vector memory using a state vector component; receiving a plurality of streams of input data from a second plurality of sensors; segmenting the input data into a plurality of discrete data segments, wherein each data segment includes an element of input data stream and a subset of state vector data related to the element of input data stream; and performing concurrent calculations related to the input data using the plurality of data segments.

11 . The method of claim 10 , wherein the data segment comprises a stride.

12 . The method of claim 11 , further comprising storing results of the performed calculations in at least one array of Double Data Rate (DDR) memory blocks as a plurality of strides, wherein one memory block stores at least one stride.

13 . The method claim 10 , wherein the input data comprises image data and wherein the performed concurrent calculations comprise concurrent image processing calculations.

14 . The method of claim 13 , wherein the concurrent calculations are performed by a plurality of hardware processing units and wherein each of the plurality of hardware processing units comprises at least one of an Application Specific Integrated Circuit (ASIC) and a Field Programmable Gate Array (FPGA).

15 . The method of claim 14 , wherein each of the plurality of hardware processing units is pre-programmed to perform a particular image processing function.

Assignees

• Goodrich Corp

Inventors

• Rencs Erik V
• Ramsey Scott W

Classifications

• G06F9/462 Primary

  with multiple register sets · CPC title

• G06F3/0604

  Improving or facilitating administration, e.g. storage management · CPC title

• G06F12/0842

  for multiprocessing or multitasking · CPC title

• G06F12/0875

  with dedicated cache, e.g. instruction or stack · CPC title

• G06F2212/60

  Details of cache memory · CPC title