This disclosure generally relates to power management with an idle state definition.
The demand for more powerful computers and communication products has resulted in faster processors and busses that often have higher die temperatures and consume increasing amounts of power. Consequently, design engineers struggle with reducing power consumption and die temperature.
For example, serial busses that utilize serial links, such as, PCI-Express and other link based interconnection systems, utilize an interface that is always transmitting and receiving data. Therefore, one disadvantage of high speed serial links is increased power consumption. Specifically, the physical layer transmitter and receiver can consume a constant DC current, regardless whether the interface is actually transmitting or receiving data.
Various attempts for power management schemes for addressing the increased power consumption for serial links and serial busses have been proposed and implemented. However, they require time to put the serial link to sleep and time to wake the link up again. Consequently, the latency required for transitioning between power management states adversely impacts performance in a microprocessor environment.
Subject matter is particularly pointed out and distinctly claimed in the concluding portion of the specification. The claimed subject matter, however, both as to organization and method of operation, together with objects, features, and advantages thereof, may best be understood by reference to the following detailed description when read with the accompanying drawings in which:
In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the claimed subject matter. However, it will be understood by those skilled in the art that the claimed subject matter may be practiced without these specific details. In other instances, well-known methods, procedures, components and circuits have not been described in detail so as not to obscure the claimed subject matter.
An area of current technological development relates to achieving longer battery life for communication products and computer or computing systems by reducing power consumption. Presently, serial busses that utilize serial links, such as, PCI-Express and other link based interconnection systems, utilize an interface that is always transmitting and receiving data. Therefore, one disadvantage of high speed serial links is increased power consumption. However, they require time to put the serial link to sleep and time to wake the link up again. Consequently, the latency required for transitioning between power management states adversely impacts performance in a microprocessor environment.
In contrast, the claimed subject matter facilitates reduced power consumption by a physical interface layer transmitting and/or receiving a predetermined network packet when another layer does not need the physical interface layer for transmitting or receiving data. In one embodiment, the predetermined network packet is designated as an “idle packet”, wherein the data is constant for a respective lane. Upon transmitting or receiving the idle packets, the predetermined network pack is detected. Consequently, various circuit portions of the physical interface's transmitter and/or receiver may be clock gated or powered down. Therefore, the transmitter and/or receiver utilize less power relative to the condition when the transmitter and receiver are receiving normal network packets with data for other protocol layers.
In one embodiment, the predetermined network packet is generated based at least in part on conforming to a predetermined ECC (Error Correcting Code). In another embodiment, the predetermined network packet is generated based at least in part on conforming to predetermined ECC (Error Correcting Code) and Code Recovery Codes (CRC). In both previous embodiments, any known ECC and CRC may be used. In yet another embodiment, the predetermined network packet is generated based at least in part maintaining constant data in each lane regardless of ECC and CRC.
In one embodiment, the network packet comprises a plurality of lanes. The network packet contains control, data, and may also include clock information. In this embodiment, the lane is a single bit wide (serial) conduit of control/data or clock information. For example, this lane is physically represented by a differential pair of copper conductors.
As previously described, the claimed subject matter facilitates reduced power consumption by a physical interface layer transmitting and/or receiving a predetermined network packet when another layer does not need the physical interface layer for transmitting or receiving data. In one embodiment, the cache-coherent, link-based interconnect scheme utilizes a plurality of layers, such as, protocol layer, transport layer, routing layer, link layer, and physical layer. In this embodiment, the physical layer 202 is considered the “lowest” layer since it involves actually transmitting and/or receiving data and is directly interacts with the Link Layer. Thus, it isolates the other layers from electrical and physical implementation details.
In another embodiment, the cache-coherent, link-based interconnect scheme utilizes a plurality of layers, such as, protocol layer, link layer, and physical layer. For example, the functionality of a layer may be incorporated within another layer for certain platforms, such as, desktop/mobile and dual processor systems. Specifically, the functionality of the routing layer may be embedded in a link layer. Hence, the routing layer is not needed for such systems.
As previously described, the physical interface layer transmits and/or receives a predetermined network packet when another layer does not need the physical interface layer for transmitting or receiving data. In one embodiment, the predetermined network packet is designated as an “idle packet”, wherein the data is constant for a respective lane. For example, this proposed packet is one example of an idle packet that may be used by the physical interface layer to transmits and/or receives when another layer does not need the physical interface layer for transmitting or receiving data. Upon transmitting or receiving the idle packets, various circuit portions of the physical interface's transmitter and/or receiver may be clock gated or power down, which is discussed further in connection with
In this embodiment, the proposed idle packet meets Cyclic Redundancy Code (CRC) requirements of the cache-coherent, link-based interconnect scheme. The first example of a packet depicts 22 lanes with two additional lanes for CRC. In this embodiment, the first row is the header of the packet and may be repeated for the remaining words in the packet. In this embodiment, the packet contains four words and the binary signature is repeated for each row so as to maintain constant binary values for each respective lane. In another embodiment, a packet depicts 18 lanes with two additional lanes for CRC. For both previous embodiments, the idle packet conforms to a CRC and the CRC was generated based on the following polynomial: 0x185, i.e., x8+x7+x2+1.
As previously described, a predetermined network packet is generated based either on ECC, CRC, or neither, to arrive at a particular signature of data with each lane maintaining the same binary value. During normal operation, a network packet is detected a being an “idle packet” by analyzing the first row, designated as a header, to determine whether it is similar to a header of an idle packet. Consequently, various circuit portions of the physical interface's transmitter and/or receiver may be clock gated or power down, which is discussed further in connection with
However, the claimed subject matter is not limited to the present example depicted in
In one embodiment, the network packet comprises a plurality of lanes. The network packet contains control, data, and may also include clock information. In this embodiment, the lane is a single bit wide (serial) conduit of control/data or clock information. For example, this lane is physically represented by a differential pair of copper conductors. One skilled in the art appreciates utilizing network packets of different lane sizes and with different combinations of clock information. In another embodiment, a network packet comprises 20 lanes with 1 lane for clock information. For example, in one embodiment for a 20 bit wide link utilizes 42 lanes: 20 lanes for carrying control/data and 1 lane carrying the forwarded clock in one direction and 20 lanes for carrying control/data and 1 lane carrying the forwarded clock in the other direction.
The receiver comprises a sampler, interpolator, a variable offset comparator (VOC), trim and deskew control, a multiplexer coupled to the DLL and interpolator. As previously described, the transmitter would forward a predetermined network packet (idle packet) to the receiver in the event that actual data/control information is not needed. For example, the idle packet may be sent when a particular communication layer, such as, a physical layer is not needed to send or receive actual data that is needed by an integrated device or system. Upon transmitting the idle packet, the transmitter data path circuitry may be powered down or clock gated, such as, the TX EQ and pre-driver circuitry. The receiver analyzes at least one row (header) of the network packet and if the particular header matches a predetermined signature of the idle packet, various portions of the receiver may be clock gated or power down. In this embodiment, the sampler, VOC, alignment register, and interpolator, are clock gated or power down for a duration of at least this particular network packet.
Subsequently, if another idle packet is detected, the receiver portions may remain powered down. Eventually, when a network packet that does not match the predetermined idle packet is received, the various portions of the receiver circuit that are needed for this network packet may be powered on or disabling of the clock gating.
The claimed subject matter is not limited to powering down or clock gating the previously described portions of the receiver and transmitter. One skilled in the art appreciates powering down or clock gating different portions of the receiver and transmitter based at least in part on their particular design, communication protocol, etc.
For embodiment 406, the uni-processor P is coupled to graphics and memory control, depicted as IO+M+F, via a network fabric link that corresponds to a layered protocol scheme. The graphics and memory control is coupled to memory and is capable of receiving and transmitting via PCI Express Links. Likewise, the graphics and memory control is coupled to the ICH. Furthermore, the ICH is coupled to a firmware hub (FWH) via a LPC bus. Also, for a different uni-processor embodiment, the processor would have external network fabric links. The processor may have multiple cores with split or shared caches with each core coupled to a Xbar router and a non-routing global links interface. Thus, the external network fabric links are coupled to the Xbar router and a non-routing global links interface.
Although the claimed subject matter has been described with reference to specific embodiments, this description is not meant to be construed in a limiting sense. Various modifications of the disclosed embodiment, as well as alternative embodiments of the claimed subject matter, will become apparent to persons skilled in the art upon reference to the description of the claimed subject matter. It is contemplated, therefore, that such modifications can be made without departing from the spirit or scope of the claimed subject matter as defined in the appended claims.
While the present invention has been described with respect to a limited number of embodiments, those skilled in the art will appreciate numerous modifications and variations therefrom. It is intended that the appended claims cover all such modifications and variations as fall within the true spirit and scope of this present invention.
This application is a continuation of U.S. patent application Ser. No. 10/833,964 filed Apr. 27, 2004 now U.S. Pat. No. 7,586,951, entitled “METHOD, APPARATUS, AND SYSTEM FOR IDLE STATE DEFINITION FOR POWER MANAGEMENT,” the content of which is hereby incorporated by reference.
Number | Name | Date | Kind |
---|---|---|---|
5371734 | Fischer | Dec 1994 | A |
5754436 | Walsh et al. | May 1998 | A |
5778419 | Hansen et al. | Jul 1998 | A |
6009488 | Kavipurapu | Dec 1999 | A |
6065077 | Fu | May 2000 | A |
6108322 | Kotzin et al. | Aug 2000 | A |
6246725 | Vanzieleghem et al. | Jun 2001 | B1 |
6457081 | Gulick | Sep 2002 | B1 |
6457087 | Fu | Sep 2002 | B1 |
6498936 | Raith | Dec 2002 | B1 |
6516442 | Wang et al. | Feb 2003 | B1 |
6795450 | Mills et al. | Sep 2004 | B1 |
7007103 | Pinkerton et al. | Feb 2006 | B2 |
7031258 | Frisch et al. | Apr 2006 | B1 |
7035294 | Dove et al. | Apr 2006 | B2 |
7095798 | Hunton | Aug 2006 | B2 |
7117292 | Kelly | Oct 2006 | B2 |
7131020 | Moll et al. | Oct 2006 | B2 |
7165531 | Bratvold et al. | Jan 2007 | B2 |
7170952 | Hunton | Jan 2007 | B2 |
7254159 | Lavelle et al. | Aug 2007 | B1 |
7313712 | Cherukuri et al. | Dec 2007 | B2 |
7586951 | Frodsham et al. | Sep 2009 | B2 |
7610500 | Cherukuri et al. | Oct 2009 | B2 |
20030206564 | Mills et al. | Nov 2003 | A1 |
20040153676 | Krantz et al. | Aug 2004 | A1 |
20050238055 | Frodsham et al. | Oct 2005 | A1 |
20050262368 | Cherukuri et al. | Nov 2005 | A1 |
20060034295 | Cherukuri et al. | Feb 2006 | A1 |
20060180123 | Bratvold et al. | Aug 2006 | A1 |
20070041405 | Navada et al. | Feb 2007 | A1 |
20080075107 | Cherukuri et al. | Mar 2008 | A1 |
20080077814 | Cherukuri et al. | Mar 2008 | A1 |
20080091963 | Cherukuri et al. | Apr 2008 | A1 |
20100027564 | Frodsham et al. | Feb 2010 | A1 |
Number | Date | Country | |
---|---|---|---|
20100027564 A1 | Feb 2010 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10833964 | Apr 2004 | US |
Child | 12511239 | US |