USPatent applicationPatented

Method and apparatus for concurrently communicating with multiple embedded dynamic random access memory devices

Granted 3 Jun 2003 · no office action yet

Current assignee: Google · originally International Business Machines

Law firm: Law firm · Log in to unlock

Attorney: Attorney · Log in to unlock

Inventors: William John Starke, Praveen S. Reddy, Ravi Kumar Arimilli, James Stephen Fields, Jr. +1 · Examiner: Reginald G. Bragdon · AU 2186 · TC 2100

Life of the application

8 dated events
⤢ drag to zoom20022004200620082010201220142016201820202022ProsecutionOwnershipTerm & fees
ProsecutionOwnershipTerm & feeshover for detail · click to open

Abstract

An apparatus for providing concurrent communications between multiple memory devices and a processor is disclosed. Each of the memory device includes a driver, a phase/cycle adjust sensing circuit, and a bus alignment communication logic. Each phase/cycle adjust sensing circuit detects an occurrence of a cycle adjustment from a corresponding driver within a memory device. If an occurrence of a cycle adjustment has been detected, the bus alignment communication logic communicates the occurrence of a cycle adjustment to the processor. The bus alignment communication logic also communicates the occurrence of a cycle adjustment to the bus alignment communication logic in the other memory devices. There are multiple receivers within the processor, and each of the receivers is designed to receive data from a respective driver in a memory device. Each of the receivers includes a cycle delay block. The receiver that had received the occurrence of a cycle adjustment informs the other receivers that did not receive the occurrence of a cycle adjustment to use their cycle delay block to delay the incoming data for at least one cycle.

Description

6 parts
›BACKGROUND OF THE INVENTION

1. Technical Field

The present invention relates to memory devices in general, and in particular to embedded dynamic random access memory devices. Still more particularly, the present invention relates to a method and apparatus for concurrently communicating with multiple embedded dynamic random access memory devices.

2. Description of the Prior Art

Embedded dynamic random access memory (DRAM) is a type of integrated circuit having both memory cells and their control circuits formed on a single semiconductor chip. Specifically, a complete embedded DRAM includes a transfer field effect transistor (FET), a capacitor coupled to the transfer FET, and a group of control circuitry. The transfer FET acts as a switch between the lower electrode of the capacitor and a bit line. Therefore, data within the capacitor can be written in or read out.

Embedded DRAMs are capable of transferring a large quantity of data at a very high speed. Because of their relatively high processing speed and storage capacity, embedded DRAMs have been commonly employed in various high-end integrated circuits, such as graphic processors. In addition, embedded DRAMs can provide a processor a faster access to larger capacities of memory at a lower cost than that currently available using conventional static random access memories (SRAMs) and/or electrically erasable programmable read only memories (EEPROMs).

The present disclosure provides a method and apparatus for concurrently communicating with multiple embedded DRAM devices by a processor.

›SUMMARY OF THE INVENTION

In accordance with a preferred embodiment of the present invention, a data processing system includes a processor and multiple memory devices coupled to the processor. Each of the memory device includes a driver, a phase/cycle adjust sensing circuit, and a bus alignment communication logic. Each phase/cycle adjust sensing circuit detects an occurrence of a cycle adjustment from a corresponding driver within a memory device. If an occurrence of a cycle adjustment has been detected, the bus alignment communication logic communicates the occurrence of a cycle adjustment to the processor. The bus alignment communication logic also communicates the occurrence of a cycle adjustment to the bus alignment communication logic in the other memory devices. There w are multiple receivers within the processor, and each of the receivers is designed to receive data from a respective driver in a memory device. Each of the receivers includes a cycle delay block. The receiver that had received the occurrence of a cycle adjustment informs the other receivers that did not receive the occurrence of a cycle adjustment to use their cycle delay block to delay the incoming data for at least one cycle.

All objects, features, and advantages of the present invention will become apparent in the following detailed written description.

›BRIEF DESCRIPTION OF THE DRAWINGS

The invention itself, as well as a preferred mode of use, further objects, and advantages thereof, will best be understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying drawings, wherein:

FIG. 1 is a block diagram of a data processing system having multiple embedded DRAM devices, in accordance with a preferred embodiment of the present invention;

FIG. 2 is a block diagram of one of the embedded DRAM devices within the data processing system from FIG. 1, in accordance with a preferred embodiment of the present invention;

FIG. 3 is a block diagram of a communication controller within the embedded DRAM device from FIG. 2, in accordance with a preferred embodiment of the present invention;

FIG. 4 is a block diagram of a block diagram of the processor from FIG.1 for controlling multiple embedded DRAM devices, in accordance with a preferred embodiment of the present invention;

FIG. 5 is a block diagram of a primary receiver within the processor from FIG. 4, in accordance with a preferred embodiment of the present invention;

FIG. 6 is a block diagram of a secondary receiver within the processor from FIG. 4, in accordance with a preferred embodiment of the present invention; and

FIGS. 7 a - 7 b are high-level logic flow diagrams depicting a method for providing concurrent communication between multiple embedded DRAM devices and a processor, in accordance with a preferred embodiment of the present invention.

›DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT · 1 of 3

In order for a processor to take further advantage of the relatively fast speed of embedded dynamic random access memory (DRAM) devices, it may be desirable for the processor to be able to access multiple embedded DRAM devices concurrently. For example, when implementing an off-chip cache memory for a processor with four separate embedded DRAM devices, each cache line of the cache memory can be divided among each of the four embedded DRAM devices. With such configuration, each cache line data transfer from the off-chip cache memory can be accomplished via one transfer cycle instead of four separate transfer cycles. Sometimes, however, data from each of the four embedded DRAM devices may not arrive at the same time because of line delays caused by temperature increase or some other disturbances. Thus, it is desirable to provide a method and apparatus to ensure that data from each of the four embedded DRAM devices can be arrived at the processor simultaneously.

Referring now to the drawings and in particular to FIG. 1, there is depicted a block diagram of a data processing system having multiple embedded DRAM devices, in accordance with a preferred embodiment of the present invention. As shown, a data processing system 10 includes a processor 11 coupled to embedded DRAM devices 12 a - 12 d . Each of embedded DRAM devices 12 a - 12 d is controlled by processor 11 via control lines 13 a - 13 d , respectively. Similarly, data communications are provided between each of embedded DRAM devices 12 a - 12 d and processor 11 via data buses 14 a - 14 d , respectively. In this example, embedded DRAM device 12 a is denoted as a primary embedded DRAM device while embedded DRAM devices 12 b - 12 d are denoted as secondary embedded DRAM devices. As the primary embedded DRAM device, embedded DRAM device 12 a is responsible to communicate data transfer status to processor 11 via a response bus 15 . Each of embedded DRAM devices 12 b - 12 d is also equipped with a separate response bus, but those response buses are preferably left unconnected.

With reference now to FIG. 2, there is illustrated a block diagram of one of embedded DRAM devices 12 a - 12 d from FIG. 1, in accordance with a preferred embodiment of the present invention. Using embedded DRAM device 12 a as an example, embedded DRAM device 12 a includes a general function control logic 21 and a communication controller 22 . General function control logic 21 includes multiple sets of data buses 23 a - 23 h . Each set of data buses 23 a - 23 h is connected to a corresponding bank of memories within embedded DRAM device 12 a , as is well-known to those skilled in the art. In this example, embedded DRAM device 12 a has a total of eight banks of memories, from bank 0 to bank 7 . Communication controller 22 synchronizes data transfers between an embedded DRAM device, such as embedded DRAM device 12 a , and all other connected embedded DRAM devices, such as embedded DRAM devices 12 b - 12 d , and processor 11 .

Referring now to FIG. 3, there is illustrated a detailed block diagram of communication controller 22 within embedded DRAM device 12 a from FIG. 2, in accordance with a preferred embodiment of the present invention. As shown, communication controller 22 includes a phase/cycle adjust sensing circuit 31 , a bus alignment communication logic block 32 , and a driver logic block 33 . Phase/cycle adjust sensing circuit 31 monitors response signals on response bus 15 as well as data being driven out of embedded DRAM 12 a from driver logic block 33 (to processor 11 in FIG. 1 ). Bus alignment communication logic block 32 includes a mode register 34 for indicating the identification number (ID no.) of an embedded DRAM device to which mode register 34 belongs. For example, an ID no. “0” may be assigned to embedded DRAM device 12 a , an ID no. “1” may be assigned to embedded DRAM device 12 b , an ID no. “2” may be assigned to embedded DRAM device 12 c , and an ID no. “3” may be assigned to embedded DRAM device 12 d . Based on the assigned ID numbers, bus alignment communication logic block 32 can update a corresponding line of a communication ring bus 35 connecting all embedded DRAM devices 12 a - 12 d , as necessary. In the present example, communication ring bus 35 is preferably four lines wide, and each line corresponds to one of four embedded DRAM devices 12 a - 12 d . Similarly, response bus 15 is at least four lines wide, and each of the four lines corresponds to one of four embedded DRAM devices 12 a 12 d .

With reference now to FIG. 4, there is depicted a detailed block diagram of processor 11 for controlling embedded DRAM devices 12 a - 12 d , in accordance with a preferred embodiment of the present invention. As shown, processor 11 includes a processor core 40 having various logic circuits for processing data, a set of drivers 41 a - 41 d for driving control signals through control lines 13 a - 13 d , respectively, and a set of receivers 42 a - 42 d for driving data signals through data buses 14 a - 14 d , respectively. In the present example, receiver 42 a is denoted as a primary receiver while receivers 42 b - 42 d are denoted as secondary receivers. As the primary receiver, receiver 42 a is responsible to receive data transfer status from the primary embedded DRAM device (i.e., embedded DRAM device 12 a from FIG. 1) via response bus 15 . Receivers 42 a - 42 d may communicate with each other via a communication channel 45 .

Referring now to FIG. 5, there is depicted a detailed block diagram of primary receiver 42 a from FIG. 4, in accordance with a preferred embodiment of the present invention. As shown, primary receiver 42 a includes a decode alignment move block 51 and a cycle delay block 52 . Decode alignment move block 51 includes a mode register 53 for indicating the ID no. of an embedded DRAM device to which primary receiver 42 a is coupled. Decode alignment move block 51 detects any occurrence of a cycle adjustment in embedded DRAM device 12 a , and cycle delay block 52 provides a cycle delay on incoming data from embedded DRAM device 12 a , as will be further explained infra.

›DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT · 2 of 3

With reference now to FIG. 6, there is depicted a detailed block diagram of secondary receiver 42 b from FIG. 4, in accordance with a preferred embodiment of the present invention. As shown, secondary receiver 42 b includes an identify alignment move block 61 and a cycle delay block 62 . Identify alignment move block 61 includes a mode register 63 for indicating the ID no. of an embedded DRAM device to which secondary receiver 42 b is coupled. Identify alignment move block 61 detects any occurrence of a cycle adjustment in embedded DRAM device 12 b , and cycle delay block 62 provides a cycle delay on incoming data from embedded DRAM device 12 b , as will be further explained infra.

Sometimes a need for a cycle adjustment (i.e., a delay or an alignment move) may occur in one of embedded DRAM devices 12 a - 12 d . The occurrence of a cycle adjustment in embedded DRAM device 12 a can be initially detected by decode alignment move block 51 within primary receiver 42 a . The occurrence of a cycle adjustment in any one of embedded DRAM devices 12 b - 12 d can be detected by an identify alignment move block within a corresponding secondary receiver, such as identify alignment move block 61 in an associated one of secondary receivers 42 b - 42 d . The information of the occurrence of a cycle adjustment are then sent from an embedded DRAM device to a corresponding receiver in the processor. The receiver that had received the occurrence of a cycle adjustment informs the other receivers (in the processor) that did not receive the occurrence of a cycle adjustment to use their cycle delay block to delay the incoming data for at least one cycle.

Referring now to FIGS. 7 a - 7 b , there are illustrated several high-level logic flow diagrams depicting a method for providing concurrent communication between multiple embedded DRAM devices and a processor, in accordance with a preferred embodiment of the present invention. Within each embedded DRAM device, a phase/cycle adjust sensing circuit, such as phase/cycle adjust sensing circuit 31 from FIG. 3, monitors a response signal and data being driven out of the embedded DRAM from a driver logic block, such as driver logic block 33 from FIG. 3, to a processor, such as processor 11 from FIG. 1, as shown in block 71 . The phase/cycle adjust sensing circuit may monitor the ramp rate of the data being driven out of a corresponding driver logic block. Based on the ramp rate, the phase/cycle adjust sensing circuit determines whether or not a cycle adjustment has occurred, as depicted in block 72 . For example, a driver logic block of within one of the embedded DRAM devices may have a problem in driving data out to the processor because of temperature increase or some other disturbances (as a result, its ramp rate is slower than normal), and a cycle adjustment is needed. If a cycle adjustment has been detected, the phase/cycle adjust sensing circuit sends a single cycle pulse to a bus alignment communication logic block of the embedded DRAM device, such as bus alignment communication logic block 32 in FIG. 3, as shown in block 73 . The bus alignment communication logic block merges the pulse on the appropriate line location of a communication ring bus, such as communication ring bus 35 from FIG. 3, to generate a merged communication ring status, as depicted in block 74 . The bus alignment communication logic block then forwards the merged communication ring status on the communication ring bus, as shown in block 75 , such that all embedded DRAM device will aware of the cycle adjustment. Next, the bus alignment communication logic block sends the merged communication ring status onto the driver logic block, and the driver logic block then drives the merged communication ring status to the processor via a response bus, such as response bus 15 , as depicted in block 76 .

As mentioned previously, each embedded DRAM device is assigned with an ID no. For example, an ID no. “0” is assigned to embedded DRAM device 12 a , an ID no. “1” is assigned to embedded DRAM device 12 b , an ID no. “2” is assigned to embedded DRAM device 12 c , and an ID no. “3” is assigned to embedded DRAM device 12 d . The same assignment also applies to each line on communication ring bus 35 and response bus 15 . For example, line 0 of communication ring bus 35 and response bus 15 correspond to embedded DRAM device 12 a , line 1 of communication ring bus 35 and response bus 15 correspond to embedded DRAM device 12 b , etc. Based on the above-mentioned assigned ID numbers, the bus alignment communication logic block may provide the cycle adjustment information to other embedded DRAM devices and to the receivers within the processor by updating the corresponding line of communication ring bus 35 and response bus 15 .

In the present embodiment, the bus alignment communication logic block communicates the information of an occurrence of a cycle adjustment within an embedded DRAM device in the form of the above-mentioned merged communication ring status to other embedded DRAM devices via the communication ring bus 35 and to the processor via response bus 15 , as follows. A logical “1” in a line within communication ring bus 35 and response bus 15 imply a cycle adjustment has occurred in the embedded DRAM device associated with that line, and a logical “0” in a line within communication ring bus 35 and response bus 15 imply no cycle adjustment has occurred in the embedded DRAM device associated with that line. For example, a logical “0,” “1,” “0, and “0” in lines 0 , 1 , 2 , and 3 , respectively, within communication ring bus 35 and response bus 15 imply a cycle adjustment has occurred in embedded DRAM device 12 b only. As another example, a logical “1,” “0,” “1,” and “0” in lines 0 , 1 , 2 , and 3 , respectively, within communication ring bus 35 and response bus 15 imply cycle adjustments have occurred in embedded DRAM devices 12 a and 12 c .

The merged response from embedded DRAM device 12 a is seen only by the primary receiver, such as primary receiver 42 a from FIG. 4 . Within the primary receiver, the merged response is decoded, as shown in block 81 . A determination is then made as to whether or not a cycle adjustment has been detected for any of the receivers, such as receivers 42 a - 42 d , as depicted in block 82 . If a cycle adjustment has been detected, another determination is made as to whether or not the cycle adjustment is for the primary receiver, as shown in block 83 . If the cycle adjustment is for the primary receiver, then the decode alignment move block within the primary receiver, such as decode alignment move block 51 from FIG. 5, sends a single cycle pulse to a cycle delay block in all the secondary receivers, such as cycle delay block 61 from FIG. 6, to delay the incoming data by one cycle, as depicted in block 84 .

›DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT · 3 of 3

Otherwise, if the cycle adjustment is not for the primary receiver, then an alignment move request is forward to the appropriate secondary receiver on communication channel 45 , as shown in block 85 . The appropriate secondary receiver would be the secondary receiver having a cycle adjustment occurred at its associated embedded DRAM device. Within the appropriate secondary receiver, an identify alignment move block, such lay as identify alignment move block 62 from FIG. 6, sends a single cycle pulse to the cycle delay block in the primary receiver and to the cycle delay block in the remaining secondary receivers to delay the incoming data by one cycle, as depicted in block 84 .

If a cycle adjustment occurred in only one of the four embedded DRAM devices, then the incoming data in the remaining three embedded DRAM devices will be delayed by one cycle. If cycle adjustments occurred in two of the four embedded DRAM devices, then the incoming data in the remaining two embedded DRAM devices will be delayed by one cycle. If cycle adjustments occurred in all four embedded DRAM devices, then no delay in the incoming data will be necessary.

As has been described, the present invention provides a method and apparatus for concurrently communicating with multiple embedded DRAM devices. Although data are shown to be driven from embedded DRAM devices to a processor, it is understood by those skilled in the art that the present invention is also applicable to data driven from the processor to the embedded DRAM devices. Furthermore, even though embedded DRAM devices are used to illustrate the present invention, it is understood by those in the art that the present invention can also be utilized in other similar memory devices such as DRAM devices.

While the invention has been particularly shown and described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention.

Claims as granted

16 claims

Log in to read the claims of this application.

Log in to unlock

Classifications

10 codes
IPC · International Patent Classification
Section G — Physics
  • G06F13/28
  • G06F12/00
  • G06F13/00
  • G06F13/42
USPC · US Patent Classification
711/167365/144713/503713/401711/168365/233

Claim changes

Soon
Coming soonHow the claims changed between publication and grant

See which claims were amended, added or cancelled during examination, with every added and removed word marked.

AmendedAddedCancelledUnchanged

The published claims of this application are not paired with the granted ones in what we hold.

File wrapper

⤢ drag to zoomJul 2001Oct 2001Jan 2002Apr 2002Jul 2002Oct 2002Jan 2003Apr 2003Jul 2003USPTOApplicantNotice of allowance
USPTOApplicanthover for detail · click to open
Pendency
1.9 y
691 days filing → grant
Office actions
0
none on record
Examiner
Reginald G. Bragdon
art unit 2186 · TC 2100
Citations: 2 back · 1 forward

See the full prosecution history — every USPTO and applicant action on this file, in order.

Log in to unlock

Documents

Log in to open the documents of this file: the application as filed, every office action and response, the notice of allowance.

Log in to unlock

Chain of title

⤢ drag to zoom20022004200620082010201220142016201820202022Owner 1Owner 2Owner 3
Titlehover for detail · click to open

See the full assignment history — every owner this patent has passed through, with recordation dates and reel/frame numbers.

Log in to unlock