WO1998045787A1 - Pin count reduction through serialization techniques - Google Patents

Pin count reduction through serialization techniques Download PDF

Info

Publication number
WO1998045787A1
WO1998045787A1 PCT/US1997/021936 US9721936W WO9845787A1 WO 1998045787 A1 WO1998045787 A1 WO 1998045787A1 US 9721936 W US9721936 W US 9721936W WO 9845787 A1 WO9845787 A1 WO 9845787A1
Authority
WO
WIPO (PCT)
Prior art keywords
bits
serial
stream
clock signal
semiconductor device
Prior art date
Application number
PCT/US1997/021936
Other languages
French (fr)
Inventor
Russell Bell
Original Assignee
Advanced Micro Devices, Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Advanced Micro Devices, Inc. filed Critical Advanced Micro Devices, Inc.
Publication of WO1998045787A1 publication Critical patent/WO1998045787A1/en

Links

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C5/00Details of stores covered by group G11C11/00
    • G11C5/06Arrangements for interconnecting storage elements electrically, e.g. by wiring
    • G11C5/066Means for reducing external access-lines for a semiconductor memory clip, e.g. by multiplexing at least address and data signals

Landscapes

  • Information Transfer Systems (AREA)

Abstract

A method and circuit for reducing the pin count of semiconductor devices through high-speed serialization technique. Normally parallel data is bundled into a serial data package, serialized, and sent through a single pin. At the receiving semiconductor device, the serial data package is received through a single pin and deserialized.

Description

PIN COUNT REDUCTION THROUGH SERIALIZATION TECHNIQUES
Technical Field
The invention relates generally to semiconductor devices having inter-device connections and more particularly to reducing pin count in such devices.
Background Art
Modern semiconductor devices, or "chips, " form their electrical connections with the outside world through pins. Pins used for a variety of purposes such as applying power and ground reference potentials to chip and carrying signals consume space on the chip and often require expensive packages; thus it is generally desirable to minimize the number of pins.
Another purpose for pins is to provide electrical connections for transferring bit data from one chip to another chip. Conventional chips use multiple pins to input and output multiple-bit data. For example, in Fig. 1(a) , conventional IC 110 transmits eight-bit parallel data to IC 120 using transmit data pins TDO - TD7 and a transmit data strobe pin TDS . The strobe pin TDS is used for signaling that the multiple-bit data on pins TDO - TD7 is valid. IC 120 receives the eight-bit byte at pins RD0 - RD7 , when signalled by receive data strobe RDS . Thus, transmitting and receiving an eight-bit byte require eighteen pins on each chip in a conventional design.
Other conventional chips use a set of bi-directional pins to transmit and receive parallel bit data. In Fig. 1(b) , IC 111 transmits an eight-bit byte to IC 121 using bidirectional pins DO - D7 in addition to strobe pins TDS and TRS . In this configuration, the number of pins required is ten.
Switching multiple pins on the parallel bus from high to low, however, produces a significant amount of current sinking that must pass through the package ground and I/O pins, which results in a significant amount of ground bounce and electromagnetic interference ("EMI"). The ground bounce is a voltage spike caused by the lead frame/pin inductance and is directly proportional to di/dt; where di/dfc is the change in the aggregate current per unit of time caused by the switching of all the pins. Thus, the voltage spikes become worse as the system clock rate increases and as more parallel bits are switched. Ground bounce is a problem because voltage spikes can cause internal chip elements to fail. EMI is also a concern because it creates noisy signals, causing the device to misoperate. Therefore, a need exists to reduce the pin count of semiconductor devices and to curb the electrical problems of ground bounce and EMI associated with switching multiple pins simultaneously.
Disclosure of the Invention A principal objective of the present invention is to provide a method for transmitting a multiple bit signal between two chips that uses fewer pins than conventional technology. A further objective of the invention is to provide a method for transmitting a multiple bit signal between two chips that reduces the amount of ground bounce and EMI.
According to the present invention, the foregoing and other objects and advantages are attained by a circuit for converting parallel bits into a bit stream comprising a serial output buffer and a digital, high-speed transceiver. The serial output buffer receives the parallel bits and outputs a nybble stream to the transceiver. The transceiver receives the nybble stream and encodes the nybbles into the bit stream. In another aspect, the serial output buffer concatenates addressing information to the parallel bits. In accordance with one embodiment of the invention, a circuit for converting a bit stream into parallel bits comprises a digital, high-speed transceiver and a serial input buffer. The transceiver receives the bit stream and decodes the bit stream into a nybble stream and a recovered clock signal. The serial input buffer receives the nybble stream and the recovered clock signal and outputs the parallel bits based upon the nybble stream and the recovered clock signal. Preferrably, the serial input buffer outputs the parallel bits in response to addressing information embedded in the bit stream.
In accordance with another aspect of the invention, provided is a method for transmitting a data bits from a source semiconductor device to a target semiconductor device. The source semiconductor device serializes the data bits into a bit stream and outputs the bit stream through a single pin of the device. The target semiconductor device receives the bit stream through a single one of its pins. Finally, the target semiconductor device deserializes the bit stream into the data bits. Preferably, addressing information is added to each serial data package to identify the target semiconductor device.
Brief Description of the Drawings The present invention is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:
Figs. 1(a) and 1(b) are block diagrams of a connection between two conventional semiconductor devices.
Fig. 2 is a block diagram showing connections between two semiconductor devices according to the present invention.
Fig. 3(a) is a block diagram showing a high-speed, all-digital transceiver implemented in the invention. Fig. 3(b) is a timing diagram of the signals of the high-speed all-digital transceiver of Fig. 3(a).
Figs. 4(a) and 4(b) are block diagrams showing an embodiment of a serial output circuit in accordance with the invention. Figs. 5(a) and 5(b) are block diagrams showing an embodiment of a serial input buffer in accordance with the invention.
Best Mode for Carrying out the Invention
A circuit and method for reducing the pin count of an integrated circuit are described. In the following description, for the purposes of explanation, numerous specific details are set forth to provide a thorough understanding of the present invention. It will be apparent, however, that the present invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form to avoid unnecessarily obscuring the present invention.
With reference to Fig. 2, provided is a serial link connection, using minimal switching, between IC 200 and IC 201, allowing serial information to represent the normally parallel data that are provided by multiple pins. The parallel bus structure is replaced by a serial link that places data bits TD into a serial format.
IC 200 comprises transceiver 210 and serial output buffer 230; IC 201 comprises transceiver 211 and serial input buffer 221. IC 200 sends parallel data TD to IC 201 by latching parallel data TD into serial output buffer 230 with load signal LOAD. Serial output buffer 230 packages parallel data TD into data package TDATA and clocks it out to transceiver 210 using local clock LCLK. As described in more detail below, transceiver 210 serializes data package TDATA using local clock LCLK into a serial data stream, output at pin SEROUT. Pin SEROUT of IC 200 is coupled to pin SERIN of IC 201, where transceiver 211 receives the serial data stream and deserializes it into data package RDATA and recovered clock RCLK.
Serial input buffer 221 receives data package RDATA and recovered clock RCLK and transforms them into parallel bit data RD.
Both transceivers 210 and 211 are described with reference to transceiver 300 of Fig. 3(b). A preferred embodiment of the invention uses for transceiver 300 a highspeed, all-digital serializer and deserializer, described in U.S. Patent No. 5,400,370 issued to Guo on Mar. 21, 1995, incorporated herein by reference. Transceiver 300 is all- digital, in that it does not implement large, lumped, analog components, such as capacitors, for data storage, timing, or other functionsm and hence be readily implemented in large scale integrated circuitry. b
Transceiver 300 has three major components, deserializer 310, digital phase adjustment circuit 320, and serializer 330. Serializer 330, described in U.S. Patent No. 5,349,612, issued to Guo et al . on September 20, 1994, incorporated herein by reference, receives parallel data package TDATA, in N-bit nybbles clocked by local clock LCLK. The operation of serializer 330 in serializing an N-bit nybble TDATA is described with reference to the timing diagram in Fig. 3(b) . For each cycle of local clock LCLK, serializer 330 generates N cycles of bit clock BCLK, the first bit clock cycle being aligned with an edge of local clock LCLK. The period of bit clock BCLK is at most the period of local clock LCLK divided by N; the bit clock period does not have to be an integral fraction of the local clock period. The cycles of bit clock BCLK are produced by a cascade of adjustable, digital delay elements, which are preferably calibrated by an on-chip digital servo subcomponent of serializer 330, described in U.S. Patent No. 5,457,719, issued to Guo et al . on October 10, 1995. The result of the calibration circuit is sent as signal ADJ from serializer 330 to also calibrate the adjustable, digital delay elements of digital phase adjustment circuit 320.
Given parallel data nybble TDATA and generated bit clock BCLK, serializer 330 encodes nybble TDATA into a bit stream SEROUT, which may be an ECL signal for high-speed transmission. One encoding, as shown in Fig. 3(b) is NRZI (Non-Return to Zero, Invert on one) , where a transition signifies a one and a lack of a transition signifies a zero. Other encodings such as NRZ or Manchester may be used as well. When transceiver 300 receives an incoming serial data stream SERIN, the clock signal for incoming serial data stream SERIN is generally out of phase with respect to bit clock BCLK. Therefore, digital phase adjustment circuit 320 shifts the phase of the bit clock BCLK to create shifted bit clock SBCLK, having a timing edge aligned with the center of the data eye of the incoming serial data stream SERIN. Deserializer 310 takes shifted bit clock SBCLK and decodes incoming serial data stream SERIN into received data nybble RDATA. Deserializer 310 is described in more detail in U.S. Patent No. 5,367,542, D issued to Guo on November 22, 1994, incorporated herein by reference.
According to one embodiment of the invention, serial output buffer 230 is implemented by serial output shift register 410 and transceiver 400, shown in Fig. 4(a). Serial output shift register 410 receives parallel bit data TD and shifts out a series of N-bit nybbles TDATA in response to load signal LOAD using local clock LCLK. Upon each cycle of local clock LCLK, N-bit nybble TDATA is serialized by transceiver 400 and sent out through a single pin.
In Fig. 5(a), transceiver 500 receives serial data stream SERIN through a single pin and recovers each N-bit nybble RDATA and recovered clock RCLK. Using recovered clock RCLK, each N-bit nybble RDATA is shifted into serial input shift register 520. Serial input shift register outputs parallel data DATA to latch 550. Also receiving recovered clock RCLK is counter 530 which counts the total number of nybbles for the serial data package, and outputs the result to match logic 540. When the correct count of nybbles is received, match logic 540 output transfer signal XFER to latch 550. Upon receiving transfer signal XFER, latch 550 will latch parallel data DATA so that received parallel bit data RD becomes available to the rest of the chip.
Hence, what was parallel data at one chip is converted to a serial format, transmitted serially to another chip, recovered, and converted back to parallel data, effectively eliminating the requirement for the parallel bus. Therefore, only two pins are required to send and receive a plurality of bits, whereas, for example, nine pins are required by conventional technology to send eight bits in parallel. Accordingly, since only one pin on each chip is being switched, the ground bounce and other undesirable electrical effects associated with switching multiple pins simultaneously are eliminated. Finally, the use of a high-speed serializer and deserializer that encodes and recovers a faster clock signal ensures comparable performance.
In another embodiment of the invention, referring to Fig. 4(b), additional information, including destination address DEST ADDR, is concatenated to transmit data -TD to create a data package. Destination address DEST ADDR indicates which chip is to receive the data package. In Fig. 5(b), serial input buffer 521 receives deserialized data package RDATA, decodes an embedded address, and outputs it as received address ADDR. Match logic 541 compares received address ADDR to the address of the chip. Only if the two addresses match and counter 531 has reached the correct value, match logic 541 signals latch 551 with transfer signal XFER to latch parallel data DATA. In this manner, a chip directs a data package to a specific chip, even though there are other chips attached to the same 1-bit bus.
In this disclosure there is shown and described only the preferred embodiment of the invention, but, as aforementioned, it is to be understood that the invention is capable of use in various other combinations and environments and is capable of changes or modifications within the scope of the inventive concept as expressed herein.

Claims

1. A serial output circuit for converting a plurality of bits into a bit stream, comprising: a serial output buffer for receiving a load signal, a local clock signal and said plurality of bits and for outputting a nybble stream containing said plurality of bits; and a digital, high-speed transceiver, coupled to said serial output buffer, for receiving the nybble stream and the local clock signal and for encoding the nybble stream with the local clock signal as said bit stream.
2. The circuit of claim 1, wherein: said serial output buffer is configured to concatenate a destination address to the plurality of bits and output a nybble stream containing said plurality of bits and the destination address.
3. A serial input circuit for converting a bit stream into a plurality of bits, comprising: a digital, high-speed transceiver for receiving said bit stream and for generating a recovered nybble stream and a recovered clock signal based on said bit stream; and a serial input buffer, coupled to said digital, highspeed transceiver, for receiving the recovered nybble stream and the recovered clock signal and for outputting said plurality of bits based on the recovered nybble stream and the recovered clock signal.
4. The circuit of claim 3, wherein said serial input buffer comprises: a serial input shift register, coupled to said transceiver, for shifting the recovered nybble stream with the recovered clock signal, and for outputting a received plurality of bits based upon the nybble stream; a counter, coupled to said transceiver, for counting the cycles of the recovered clock signal and for outputting a clock signal count; match logic, coupled to said counter, for receiving the clock signal count and generating a transfer signal based upon the clock signal count; and a latch, coupled to said serial input shift register and said match logic, for latching the received plurality of bits in response to the transfer signal and outputting said plurality of bits.
5. The circuit of claim 4, wherein: said serial input shift register is configured to output a destination address based upon the nybble stream; and said match logic is coupled to said serial input shift register and is configured to generate the transfer signal based upon the clock signal count and the destination address.
6. The circuit of claim 5, wherein: said match logic is configured to compare the destination address with a chip address and generate the transfer signal based upon the clock signal count based upon the comparison of the destination address and the chip address.
7. A method for transmitting a plurality of data bits from a source semiconductor device to a target semiconductor device, comprising the steps of: serializing said plurality of data bits into a serial data package with a first digital, high-speed transceiver at said source semiconductor device; outputting said serial data package through a single pin of said source semiconductor device; transmitting the serial data package from said signal pin of said source semiconductor device to a single pin of said target semiconductor device; receiving said serial data package through the single pin of the target semiconductor device; and deserializing said serial data package from the single pin of said target semiconductor device into said plurality of bits with a second digital, high-speed transceiver at said target semiconductor device.
8. The method of claim 7, wherein the step of serializing said plurality of data bits into a serial package with a digital, high-speed transceiver at said source semiconductor device, comprises the steps of: concatenating said plurality of data bits with an address signal into a data package, and serializing the data package into said serial data package with the digital, high-speed transceiver; and wherein the step of deserializing said serial data package into said plurality of bits with a digital, high-speed transceiver at said target semiconductor device, comprises the steps of: deserializing the amplified serial data package with a digital, high-speed transceiver into a stream of serial bits, a recovered clock, and a decoded address signal, latching the plurality of loaded bits based upon the recovered clock signal and the decoded address signal into a latch, and generating said plurality of bits from the latch.
PCT/US1997/021936 1997-04-10 1997-11-25 Pin count reduction through serialization techniques WO1998045787A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US83881397A 1997-04-10 1997-04-10
US08/838,813 1997-04-10

Publications (1)

Publication Number Publication Date
WO1998045787A1 true WO1998045787A1 (en) 1998-10-15

Family

ID=25278110

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1997/021936 WO1998045787A1 (en) 1997-04-10 1997-11-25 Pin count reduction through serialization techniques

Country Status (1)

Country Link
WO (1) WO1998045787A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7336209B2 (en) * 2002-06-27 2008-02-26 Robert Bosch Gmbh Method and device for data transmission

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4148099A (en) * 1978-04-11 1979-04-03 Ncr Corporation Memory device having a minimum number of pins
US4159541A (en) * 1977-07-01 1979-06-26 Ncr Corporation Minimum pin memory device
EP0416714A2 (en) * 1985-07-12 1991-03-13 Anamartic Limited Wafer-scale integrated memory circuits
US5237322A (en) * 1990-12-08 1993-08-17 Deutsche Itt Industries Gmbh Master-slave data transmission system employing a flexible single-wire bus
EP0597583A1 (en) * 1992-10-09 1994-05-18 Nokia Mobile Phones Ltd. Serial bus between integrated circuits
EP0744752A2 (en) * 1995-05-22 1996-11-27 AT&T IPM Corp. Memory chip architecture

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4159541A (en) * 1977-07-01 1979-06-26 Ncr Corporation Minimum pin memory device
US4148099A (en) * 1978-04-11 1979-04-03 Ncr Corporation Memory device having a minimum number of pins
EP0416714A2 (en) * 1985-07-12 1991-03-13 Anamartic Limited Wafer-scale integrated memory circuits
US5237322A (en) * 1990-12-08 1993-08-17 Deutsche Itt Industries Gmbh Master-slave data transmission system employing a flexible single-wire bus
EP0597583A1 (en) * 1992-10-09 1994-05-18 Nokia Mobile Phones Ltd. Serial bus between integrated circuits
EP0744752A2 (en) * 1995-05-22 1996-11-27 AT&T IPM Corp. Memory chip architecture

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
SORGENFREI J ET AL: "SCHNITTSTELLEN-IC FUR DEN SERIELLEN KFZ-BUS", ELEKTRONIK, vol. 39, no. 10, 11 May 1990 (1990-05-11), pages 140 - 143, XP000117042 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7336209B2 (en) * 2002-06-27 2008-02-26 Robert Bosch Gmbh Method and device for data transmission

Similar Documents

Publication Publication Date Title
US5931927A (en) Device for reducing the fluctuation of power supply voltage due to inductance by inverting bits in groups of data lines
US4587445A (en) Data output circuit with means for preventing more than half the output lines from transitioning simultaneously
CN101496367B (en) Alignment and deskew for multiple lanes of serial interconnect
US20090027243A1 (en) Mcu with integrated voltage isolator to provide a galvanic isolation between input and output
US20080317106A1 (en) Mcu with integrated voltage isolator and integrated galvanically isolated asynchronous serial data link
US6894530B1 (en) Programmable and fixed logic circuitry for high-speed interfaces
US6311239B1 (en) Architecture, circuitry and method for transmitting n-bit wide data over m-bit wide media
US6741193B2 (en) Parallel in serial out circuit having flip-flop latching at multiple clock rates
US6192069B1 (en) Circuit and methodology for transferring signals between semiconductor devices
US11145340B2 (en) Data transmission code and interface
CN116153362B (en) Delay feedback method and delay feedback storage structure for read latency counter
US7796063B2 (en) Data transmission circuits and data transceiver systems
US5646543A (en) Integrated circuit having reduced inductive noise
US11169952B2 (en) Data transmission code and interface
US7397878B2 (en) Data communication method and data communication device and semiconductor device
WO1998045787A1 (en) Pin count reduction through serialization techniques
US5402430A (en) Parity inversion test system
US7519742B2 (en) Asynchronous serial data apparatus for transferring data between one transmitter and a plurality of shift registers, avoiding skew during transmission
US5886539A (en) Communication within an integrated circuit by data serialization through a metal plane
US6556645B2 (en) Multi-bit counter
CN113032321A (en) Address extension circuit, communication interface chip and communication system
TW201926908A (en) Encoding and decoding architecture for high-speed data communication system and related physical layer circuit, transmitter and receiver and communication system thereof
US6049571A (en) Encoding circuit with a function of zero continuous-suppression in a data transmission system
JP3504316B2 (en) Multi-bit counter
KR100260795B1 (en) An interface system for on-chip or off-chip interconnection

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): JP KR

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
NENP Non-entry into the national phase

Ref country code: JP

Ref document number: 1998542740

Format of ref document f/p: F

122 Ep: pct application non-entry in european phase