Stream data processor

Patent metadata

Field	Value
Publication number	US-9448967-B2
Application number	US-201213687102-A
Country	US
Kind code	B2
Filing date	Nov 28, 2012
Priority date	Oct 31, 2012
Publication date	Sep 20, 2016
Grant date	Sep 20, 2016

Abstract

Techniques are provided aimed at improving the flexibility and reducing the area and power consumption of digital baseband integrated circuits by using stream data processor based modem architecture. Semiconductor companies offering baseband ICs for handsets, face the challenges of improving die size efficiency, power efficiency, performance, time to market, and coping with evolving standards. Software defined radio based implementations offer a fast time to market. Dedicated hardware designs give the best die size and power efficiency. To combine the advantages of dedicated hardware with the advantages of conventional software defined radio solutions the stream data processor is partitioned into a stream processor unit, which implements processing functions in dedicated hardware and is hence die size and power efficient, and a flexible stream control unit which may be software defined to minimize the time to market of the product.

First claim

What is claimed is:

1. A stream data processor, comprising:
one or more stream input interfaces;
one or more stream output interfaces;
one or more memories;
one or more stream processing units that implement data processing operations, each stream processing unit comprising a plurality of processing elements that implement processing operations and each processing element comprising one or more read-blocks and one or more write-blocks; and
one stream control unit that provides configuration data to the stream processing units to perform sequencing of the data processing operations,
wherein the stream control unit provides a separate configuration, a separate synchronous enable and a common trigger to each read-block and write-block, and
wherein the stream control unit provides a new configuration to each read-block and write-block once each of the one or more read-blocks and one or more write-blocks has asserted a respective done flag indicating that a configured operation has completed.

2. The stream data processor according to claim 1 , wherein the stream control unit comprises:
one or more processing element configuration interfaces that interface to the stream processing unit;
one or more top level configuration interfaces that interface to at least one higher level processor; and
a finite state machine that executes a task by sequencing the data processing operations.

3. The stream data processor according to claim 1 , wherein the stream control unit comprises:
one or more processing element configuration interfaces that interface to the stream processing unit;
one or more top level configuration interfaces that interface to at least one higher level processor; and
a plurality of finite state machines configured to execute a plurality of independent tasks.

4. The stream data processor according to claim 1 , wherein the stream control unit comprises:
one or more processing element configuration interfaces that interface to the stream processing unit;
one or more top level configuration interfaces that interface to at least one higher level processor; and
a configuration register which is fed by a dedicated microprocessor that provides configuration sequences.

5. The stream data processor according to claim 1 , wherein the stream control unit comprises:
one or more processing element configuration interfaces that interface to the stream processing unit;
one or more top level configuration interfaces that interface to at least one higher level processor; and
a configuration register, fed by a finite state machine coupled with a FIFO containing regularly updated configuration sequences.

6. The stream data processor according to claim 1 , wherein the stream control unit comprises:
one or more processing element configuration interfaces that interface to the stream processing unit;
one or more top level configuration interfaces that interface to at least one higher level processor; and
a configuration register, fed by a finite state machine coupled with a ROM containing predetermined configuration sequences.

7. The stream data processor according to claim 1 , wherein the stream processor unit comprises:
one or more memory interfaces, each having one memory arbitration unit;
each processing element corresponding to one processing element configuration interface;
one stream input multiplexer; and
one stream output multiplexer.

8. The stream data processor according to claim 7 , wherein each memory arbitration unit arbitrates memory accesses requested from concurrent data streams, the concurrent data streams flow to and from the associated memory, and wherein the stream input multiplexer routes one or more concurrent data streams from the stream input interfaces, and from the memory arbitration units, to respective processing elements of the plurality of processing elements, and then from the respective processing elements to the stream output interfaces and to the memory arbitration units.

9. The stream data processor according to claim 7 , wherein each processing element comprises:
one or more processing functions that perform data stream operations,
wherein the one or more read-blocks generate processing function input data streams by performing a sequence of read operations; and
the one or more write-blocks forward processing function output data streams by performing a sequence of write operations.

10. The stream data processor according to claim 9 , wherein each processing function comprises:
one or more stream input data interfaces;
one or more stream output interfaces; and,
a configurable stream operation.

11. The stream data processor according to claim 10 , wherein each input stream interface and output stream interface respectively comprise a relative-address stream interface, each relative-address stream interface having one relative-address generator.

12. The stream data processor according to claim 1 , wherein each memory has a bit width tailored to application of the stream data processor.

Assignees

• Mstar Semiconductor Inc
• Mstar Semiconductor Inc
• Mstar Software R&D (Shenzhen) Ltd

Inventors

• Kuechler Wolfgang

Classifications

• G06F15/8015 Primary

  One dimensional arrays, e.g. rings, linear arrays, buses · CPC title