The present invention relates generally to serial interfaces for integrated circuits.
Serial interfaces are used to communicate data between devices in a system. Serial interface protocols have been developed at increasingly fast data rates. Examples of industry-standard protocols for serial interfaces include PCI Express® (Peripheral Component Interconnect Express), XAUI (X Attachment Unit Interface), sRIO (serial Rapid IO), and many others.
One embodiment relates to an integrated circuit which includes a method of dynamically adjusting a receive buffer in an integrated circuit. A fixed-size buffer circuit of the receive buffer is used to buffer data received by way of a serial interface circuit. The performance of the serial interface circuit is monitored. The receive buffer is dynamically extended based on said performance.
Other embodiments, aspects and features are also disclosed.
In order to achieve high performance in a serial receiver on an integrated circuit, a typical system will provide a receive buffer formed using a standard cell or a structured array cell within the confines of “hard” logic circuitry. The size of this “hard” buffer is generally determined up-front in the design phase.
Applicants have determined that such a buffer is disadvantageous in that it cannot be re-sized after the design is fabricated and used in an actual system. In order to overcome this disadvantage, the present disclosure provides a dynamically-adjustable buffer space for an integrated circuit.
The core circuit region 102 may comprise, for example, a programmable (configurable) array of circuit blocks and interconnects. The core circuit region 102 may be, for example, the core fabric of a programmable logic device (PLD), such as a complex PLD or a field programmable gate array (FPGA). Note that the specific circuitry in the core circuit region 502 will vary, and the present invention is not intended to be limited to a specific implementation or layout of a core circuit region 102.
As further shown, an I/O circuit array 104 may include an array of high-speed serial interface (HSSI) channel circuits 106. An HSSI channel circuit 106 may include, for example, a physical coding sublayer (PCS) channel circuit and a physical media access (PMA) channel circuit.
The PCS channel circuit generally provides digital logic functions which implement data communication protocols (or portions thereof), while the PMA channel circuit generally provides mixed (analog/digital) signal functionality for the data communications. For example, a PMA channel circuit may be configured to perform, among other operations, serialization of data to be transmitted (conversion from parallel to serial) and de-serialization of received data (conversion from serial to parallel). For certain protocols, a PCS channel circuit may be configured to perform, among other functions, 64-bit to 66-bit encoding for data to be sent to the PMA channel circuit and 66-bit to 64-bit decoding for data received from the PMA channel circuit. A PCS channel may also perform phase compensation using FIFO (first-in first-out) buffers.
As shown, the integrated circuit 100 may include a “hard intellectual property” (HIP) circuitry 110. The HIP circuitry 110 may be “hard-wired” in that the circuitry is formed during the fabrication of the integrated circuit 100 and is not configured by subsequent programming. The HIP circuitry 110 may include, for example, digital circuitry designed to implement communications protocols used by the HSSI channel circuits 106. A high-performance buffer circuit 112 may be formed in the HIP circuits 110. The high-performance buffer 112 may be used to buffer data received by one or more of the HSSI channel circuits 106. The integrated circuit 100 may also include a memory column 114 of programmable memory blocks within the core circuit region 102.
In accordance with this embodiment of the invention, the integrated circuit 100 may include an extensible-range low-performance buffer 116 formed in the memory column 114 and “soft” (programmable) control logic 118 formed in the core circuit region 102. The operation of the extensible-range low-performance buffer 116 and the soft control logic 118 is described further below in relation to
As shown in
During the configuration (programming) 220 of the integrated circuit 100, the HSSI channels 106 and the soft control logic 118 are programmed 222, and the fixed-size low-performance buffer 116 is configured 224 (not necessarily in that order). The HSSI channels 106 may be programmed, for example, to utilize select HIP circuitry 110 to implement a particular serial interface protocol. As described above, the fixed-size low-performance buffer 116 may be configured within a memory column 114 on a same side of the integrated circuit 100 as the high-speed buffer 112. As described further below, the soft control logic 118 may be programmed 222 to dynamically control the accessible range of the low-performance buffer 116 as the serial data is received during the operation 230 of the integrated circuit 100.
During the operation 230 of the integrated circuit 100 in a system, serial data may be received 232 by the high-speed buffer 112 from one or more of the HSSI channel circuits, and the performance of the buffering may be monitored 232. In particular, the flow control and maximum payload performance of the buffer may be monitored 234 (for example, by monitoring circuitry in the HIP circuitry 110).
Based on the monitored performance of the buffering, the soft control logic 118 dynamically controls 236 the accessible (i.e. used) range in the low-performance buffer 116. The soft control logic 118 may also dynamically power-down 238 the un-used circuitry in the low-performance buffer 116. This advantageously conserves power consumed by the integrated circuit 100 with minimal performance impact.
The core circuit region 102 may comprise, for example, a programmable (configurable) array of circuit blocks and interconnects. The core circuit region 102 may be, for example, the core fabric of a programmable logic device (PLD), such as a complex PLD or a field programmable gate array (FPGA). Note that the specific circuitry in the core circuit region 102 will vary, and the present invention is not intended to be limited to a specific implementation or layout of a core circuit region 102.
As further shown, an I/O circuit array 104 may include an array of high-speed serial interface (HSSI) channel circuits 106. As described above, an HSSI channel circuit 106 may include, for example, a PCS channel circuit and a PMA channel circuit.
Also similar to the integrated circuit 100 of
The integrated circuit 300 includes a control block 302 which may be configured in “soft” (programmable) logic. (In an alternate embodiment, the control block may be implemented in “hard” logic.) In accordance with an embodiment of the invention, the control block 302 may be configured to utilize a partial reconfiguration 304 capability during operation of the integrated circuit 300. In particular, the control block 302 may use partial reconfiguration 304 to dynamically instantiate an appropriately-sized low-performance buffer 316 and associated “soft” (programmable) control logic 318. For example, the low-performance buffer 316 may be instantiated within the memory column 114, and the soft control logic 318 may be instantiated within the core region 102. The operation of the integrated circuit 300 of
As shown in
During the configuration (programming) 420 of the integrated circuit 300, the HSSI channels 106 and the control block 302 may be programmed 422. In addition, a low-performance buffer 316 that is reconfigurable to alternate sizes may be programmed 424. (Note that steps 422 and 424 need not be performed in the order shown.) The HSSI channels 106 may be programmed, for example, to utilize select HIP circuitry 110 to implement a particular serial interface protocol. As described above, the low-performance buffer 316 may be configured within a memory column 114 on a same side of the integrated circuit 100 as the high-speed buffer 112. As described further below, the control block 302 may be programmed 222 to use partial reconfiguration 304, if needed, to dynamically instantiate a low-performance buffer 316 of a selected size and associated soft control logic 318 as the serial data is received during the operation 430 of the integrated circuit 300.
During the operation 430 of the integrated circuit 300 in a system, serial data may be received 432 by the high-speed buffer 112 from one or more of the HSSI channel circuits, and the performance of the buffering may be monitored 432. In particular, the flow control and maximum payload performance of the buffer may be monitored 434 (for example, by monitoring circuitry in the HIP circuitry 110).
Based on the monitored performance of the buffering, the control block 302 may use partial reconfiguration 304 to dynamically instantiate 436 a low-performance buffer 316 of an appropriate size and associated soft control logic 318. This advantageously conserves power consumed by the integrated circuit 300 with minimal performance impact.
In this illustrative example, the core circuit region is shown to include logic array blocks (LABs) 502, memory blocks 504, digital signal processing (DSP) blocks 506, and memory logic array blocks (MLABs) 508. Note that the specific circuitry in the core circuit region will vary, and the present invention is not intended to be limited to a specific implementation or layout of a core circuit region.
As further shown, at least one I/O circuit array 104 may include an array of high-speed serial interface (HSSI) channel circuits 106. As described above, an HSSI channel circuit 106 may include, for example, a PCS channel circuit and a PMA channel circuit. Similar to
In accordance with an embodiment of the invention, a bounded area 510 of the integrated circuit 500 includes the HIP circuitry 110, select HSSI channels 106 being used by the HIP circuitry 110, the high-performance buffer 112 being used for data received via the HSSI channels 106, IP memory 512, and a buffer-extension range 514 within the IP memory 512. The IP memory 512 may include memory blocks (typically, higher-speed memory blocks) which are accessible to (i.e. used by) the HIP circuitry 110. The IP memory 512 may utilize structured gate array technology and may be located adjacent to the high-performance buffer 112. As described further below in relation to the method 600
As shown in
During the configuration (programming) 620 of the integrated circuit 500, the size of the buffer-extension range 514 may be allocated 622 from the IP memory 512. The buffer-extension range 514 may include, for example, select blocks of the IP memory 512 which are located adjacent to the high-performance buffer 112 and which are not necessarily needed by the communications protocol being used.
During the operation 630 of the integrated circuit 500 in a system, serial data may be received 632 by the high-speed buffer 112 from one or more of the HSSI channel circuits, and the performance of the buffering may be monitored 634 (for example, by monitoring circuitry in the HIP circuitry 110). In particular, the flow control and maximum payload performance of the buffering may be monitored.
Based on the monitored performance of the buffering, the high-performance buffer 112 may be dynamically extended 636 using one or more memory blocks from the buffer-extension range 514. For example, in order to meet new performance requirements of HIP circuitry 110 implementing PCI Express® (or another high-speed protocol), the size of the “hard” high-performance buffer 112 may be effectively extended using one or more 10 kilobyte (or other size) buffers available in the buffer-extension range 514.
The present invention can also be implemented in a system that has an integrated circuit, such as an FPGA, as one of several components.
System 700 includes a processing unit 702, a memory unit 704, and an input/output (I/O) unit 706 interconnected together by one or more buses. According to this exemplary embodiment, FPGA 708 is embedded in processing unit 702. FPGA 708 can serve many different purposes within the system 700. FPGA 708 can, for example, be a logical building block of processing unit 702, supporting its internal and external operations. FPGA 708 is programmed to implement the logical functions necessary to carry on its particular role in system operation. FPGA 708 can be specially coupled to memory 704 through connection 710 and to I/O unit 706 through connection 712.
Processing unit 702 may direct data to an appropriate system component for processing or storage, execute a program stored in memory 704, receive and transmit data via I/O unit 706, or other similar function. Processing unit 702 may be a central processing unit (CPU), microprocessor, floating point coprocessor, graphics coprocessor, hardware controller, microcontroller, field programmable gate array programmed for use as a controller, network controller, or any type of processor or controller. Furthermore, in many embodiments, there is often no need for a CPU.
For example, instead of a CPU, one or more FPGAs 708 may control the logical operations of the system. As another example, FPGA 708 acts as a reconfigurable processor that may be reprogrammed as needed to handle a particular computing task. Alternately, FPGA 708 may itself include an embedded microprocessor. Memory unit 704 may be a random access memory (RAM), read only memory (ROM), fixed or flexible disk media, flash memory, tape, or any other storage means, or any combination of these storage means.
In the above description, numerous specific details are given to provide a thorough understanding of embodiments of the invention. However, the above description of illustrated embodiments of the invention is not intended to be exhaustive or to limit the invention to the precise forms disclosed. One skilled in the relevant art will recognize that the invention can be practiced without one or more of the specific details, or with other methods, components, etc.
In other instances, well-known structures or operations are not shown or described in detail to avoid obscuring aspects of the invention. While specific embodiments of, and examples for, the invention are described herein for illustrative purposes, various equivalent modifications are possible within the scope of the invention, as those skilled in the relevant art will recognize. These modifications may be made to the invention in light of the above detailed description.
The present application is a continuation of U.S. patent application Ser. No. 13/245,790, filed Sep. 26, 2011, the disclosure of which is hereby incorporated by reference.
Number | Name | Date | Kind |
---|---|---|---|
5515329 | Dalton et al. | May 1996 | A |
5581736 | Smith et al. | Dec 1996 | A |
5867663 | McClure et al. | Feb 1999 | A |
6094695 | Kornher | Jul 2000 | A |
6385672 | Wang et al. | May 2002 | B1 |
6401147 | Sang et al. | Jun 2002 | B1 |
6987775 | Haywood | Jan 2006 | B1 |
7051306 | Hoberman | May 2006 | B2 |
7337026 | Lin | Feb 2008 | B2 |
7395385 | Cobley | Jul 2008 | B2 |
7535254 | Manula | May 2009 | B1 |
7761620 | Iwata | Jul 2010 | B2 |
7764614 | Wang | Jul 2010 | B2 |
8046597 | Dawkins | Oct 2011 | B2 |
20050209718 | Lin | Sep 2005 | A1 |
20070110086 | Wang | May 2007 | A1 |
20070112995 | Manula | May 2007 | A1 |
20110320651 | Poublan et al. | Dec 2011 | A1 |
Number | Date | Country | |
---|---|---|---|
Parent | 13245790 | Sep 2011 | US |
Child | 15496519 | US |