WO1999023551A1 - Microprocesseur pipeline a saut conditionnel en un cycle d'horloge - Google Patents
Microprocesseur pipeline a saut conditionnel en un cycle d'horloge Download PDFInfo
- Publication number
- WO1999023551A1 WO1999023551A1 PCT/FR1998/002265 FR9802265W WO9923551A1 WO 1999023551 A1 WO1999023551 A1 WO 1999023551A1 FR 9802265 W FR9802265 W FR 9802265W WO 9923551 A1 WO9923551 A1 WO 9923551A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- jmpc
- sector
- jump instruction
- microprocessor
- decoding
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/30003—Arrangements for executing specific machine instructions
- G06F9/3005—Arrangements for executing specific machine instructions to perform operations for flow control
- G06F9/30058—Conditional branch instructions
Definitions
- the present invention relates to a microprocessor of the pipeline type comprising an ordinal counter, an incrementing circuit of the ordinal counter, means for decoding a set of instructions of the microprocessor comprising means for decoding a conditional jump instruction of the counter ordinal, and a calculation unit comprising a first output for delivering a result and a second output for delivering status bits of the result.
- the present invention also relates to the execution of a conditional jump instruction in a pipeline microprocessor of the above type.
- conditional jump instruction is meant in the present request a JMPc instruction to jump from the ordinal counter PC ("program counter") of a microprocessor, the execution of which depends on the result of a previous instruction.
- PC program counter
- steps 1.x and 2.x above should be processed simultaneously with a row offset.
- step 2.3 cannot overlap with step 1.4 because the result of the previous operation is not yet available in the register Rf.
- a new instruction following the JMPc instruction cannot be read before step 2.3 of the JMPc instruction is completed because it is not yet known at which address this instruction must be read.
- the conventional solution to overcome this drawback consists in suspending the injection of a new instruction in the pipeline chain until the flags C, N, P, Z are loaded in the register Rf (step 1.5) and that the conditional jump instruction is processed (step 2.3).
- This method has the drawback of temporarily blocking the operation of the pipeline chain and slowing down the execution of a program.
- the time lost for the execution of a program is proportional to the number of conditional jumps that the program contains.
- the main object of the present invention is to provide a pipeline microprocessor structure making it possible to execute a conditional jump instruction without interrupting the pipeline processing chain.
- a secondary objective of the present invention is to provide a pipeline microprocessor structure which is simple and consumes little current.
- the present invention provides a microprocessor of the type mentioned above in which the calculation unit and the means for decoding the conditional jump instruction are arranged in two neighboring pipeline stages, and the means for decoding the conditional jump instruction are connected to the second output of the computing unit.
- the microprocessor comprises a first and a second pipeline stage each comprising a first and a second sector, the first sector of each stage being active during a first half clock cycle and the second sector of each active stage during a second clock half cycle.
- the means for decoding the conditional jump instruction are arranged in the first sector of the first pipeline stage, the calculation unit is arranged in the first sector of the second pipeline stage.
- the means for decoding the conditional jump instruction are arranged in the second sector of the first pipeline stage, the calculation unit is arranged in the first sector of the second pipeline stage.
- the incrementation circuit of the ordinal counter is arranged in the second sector of the first pipeline stage.
- the microprocessor comprises a bank of registers belonging to read in the first sector and in write to the second sector of the second pipeline stage.
- This characteristic makes it possible to simplify the structure of the microprocessor which can comprise only two pipeline stages, for the benefit of electrical consumption, and execute compact instructions comprising only an operation code, the address of a source register and the address of a destination register of the result of the operation designated by the instruction.
- the sectors of the second pipeline stage are delimited by a buffer to three states arranged between the first and second outputs of the calculation unit and write inputs of the register bank.
- conditional jump instruction is a relative jump instruction
- the means for incrementing the ordinal counter comprise an adder circuit receiving as input the current value of the ordinal counter and the output of a multiplexer circuit
- the multiplexer circuit is controlled by the means for decoding the conditional jump instruction and receives as input a normal incrementation value of the ordinal counter and a relative jump value of the ordinal counter, contained in the conditional jump instruction.
- the conditional jump instruction is an absolute jump instruction
- the means of incrementation of the ordinal counter include a multiplexer circuit receiving as input the output of an adder circuit and an absolute jump value of the ordinal counter contained in the conditional jump instruction
- the multiplexer circuit is controlled by the means for decoding the conditional jump instruction
- the adder circuit receives as input the current value of the ordinal counter and a normal incrementation value of the ordinal counter .
- the decoding means are arranged to decode a compact conditional jump instruction comprising an operation code, a condition and a jump value.
- FIG. 1 shows an alternative embodiment of the microprocessor of Figure 1.
- FIG. 1 represents an MPI pipeline microprocessor according to the invention.
- the microprocessor is arranged to execute compact instructions in the following format:
- These instructions include a field CODE 0P coded on 4 bits, giving the operation OP to be executed by the unit ALU, a field ADRd of 4 bits giving the address of a register Rd of destination of the result, and a 4-bit ADRs field giving the address of a source register Rs.
- the last field, of 4 bits, is not used here.
- the operation OP is for example the addition ADD
- the microprocessor is also arranged to execute a compact conditional jump instruction JMPc, which is for example in the following format:
- This instruction includes the operation code CODE JMPC of the instruction, on 4 bits, the code CODE CON D of the jump condition COND, on 4 bits, and the jump value VAL, on 8 bits.
- the microprocessor includes the following elements: a REGBANK register bank; an ALU arithmetic and logic calculation unit; an ordinal counter PC for giving the address for reading, in a program memory MEM, the instructions to be processed; a circuit SPC for incrementing, with each new instruction received, the ordinal counter PC; means DECl, DEC2, PREDEC for decoding a set of instructions from the microprocessor.
- the unit ALU represented in the form of a block, comprises conventional circuits for processing data necessary for the execution of instructions of arithmetic and logical type, and may in particular comprise a bit shift register.
- the unit ALU here has two inputs INI, IN2, a result output SI and an output S2 delivering the status bits C, N, P, Z of the result.
- the REGBANK register bank contains various working registers, for example sixteen registers RI to Rl ⁇ (not shown), and hardware registers among which there is a register Rf containing the status bits C, N, P, Z.
- the bank REGBANK comprises an input ADRs to receive the address of the source register Rs, an input ADRd to receive the address of the destination register Rd, an output OUTRs for reading the source register Rs, an output OUTRd for reading the destination register Rd, an input INRd for writing a result into the destination register Rd, and an input INRf for writing into the register Rf the status bits C, N, P, Z of a result.
- the decoding means DECl, DEC2, PREDEC comprise a main decoder DECl, a decoder DEC2 to control the ALU, and a PREDEC decoder of the conditional jump instruction JMPc.
- the microprocessor data path is divided into two contiguous pipeline stages ST1, ST2 delimited by registers B1, B3.
- registers B1, B3, for example lock registers (or "latch") are triggered by the passage to 1 of a clock signal H1.
- the two stages ST1 ST2 are each divided into two sectors ST11 / ST12, ST21 / ST22 by means of a register B2 arranged between the registers B1, B3 and a register B4 arranged after the register B3 according to the direction of circulation of data.
- the register B2 is for example a locking register triggered by the passage to 1 of a signal H2 phase shifted by 180 ° relative to the signal Hl.
- the register B4 is preferably a three-state buffer register ("buffer tristate”) transparent when the signal H2 is at 1 and in the high impedance state when H2 is at 0.
- the clock signal Hl synchronizes the transfer of data from stage STl to stage ST2, and signal H2 synchronizes the transfer of data from the first sector ST11, ST21 to the second sector ST12, ST22 of each stage STl , ST2.
- the clock signals Hl and transfer H2 are represented at the bottom of FIG. 1.
- the unit ALU is arranged in the sector ST21 of the pipeline stage ST2 and the decoder PREDEC arranged in the sector ST11 of the stage STl.
- the output S2 of the unit ALU which delivers the bits C, N, P, Z, is applied via a bus 10 to the PREDEC decoder.
- this connection of the PREDEC decoder to the output of the ALU unit delivering the bits C, N, P, Z, makes it possible to immediately process the conditional jump instruction, without waiting for these bits to be loaded in the register Rf.
- the PREDEC decoder receives the first 8 bits of the instructions present at the output of the memory MEM. In the case of the JMPc instruction, these first eight bits include the operation code CODE JMPC and the condition code CODE C OND-
- the outputs OUTRs and OUTRd of the REGBANK bank are applied to the inputs INI and IN2 of the ALU unit, respectively.
- the outputs SI and S2 of the ALU unit are returned via the register B4 to the inputs INRd and INRf of the REGBANK bank, respectively.
- the decoder DECl arranged in the sector ST12, receives at least the first 12 bits of the instructions, that is to say the fields OP , ADRd and ADRs, via the register B2.
- the addresses ADRs, ADRd are transferred to the corresponding entries in the REGBANK bank via the register B3.
- the REGBANK bank is commanded to read by a signal RDH1 and to write by a signal RH2, these signals being transmitted by the decoder DECl.
- the read signal RDH1 is emitted when the clock signal Hl is at 1 and the write signal WRH2 is emitted when the signal H2 is at 1.
- the bank REGBANK thus belongs, in reading, to the first sector ST21 of the stage ST2 and, in writing, in the second sector ST22 of stage ST2.
- the decoder DEC2 arranged in the sector ST21, receives from the decoder DECl the codes CODE 0P of the operations which the unit ALU must carry out.
- the ordinal counter PC is produced by means of a transparent buffer register RPC on reception of the write signal RH2.
- the buffer RPC receives from the circuit SPC a new address value, or PC address.
- the PC address is found at the output of the RPC buffer and is returned to the address input of the memory MEM by means of the register Bl.
- the memory MEM delivers at the output the instruction contained at the address PC, on reception of the read signal RDH1 (for the sake of consistency of the diagram of FIG. 1, the memory MEM is represented in the sector STll, although being a peripheral element of the microprocessor).
- the PC address can be applied to the memory MEM via an MMU converter
- the MMU converter receives an OFFAD offset address on one input and converts the PC address by addition (as shown in the figure) or by concatenation of the OFFAD address with the PC address.
- the circuit SPC comprises a multiplexer MUX receiving on a first input a normal incrementation value of the ordinal counter, here the value +1, and on a second input the last eight bits of the instruction present at the output of the memory MEM, through the B2 register. According to the instruction format described above, these last eight bits correspond to the VAL area of the JMPc instruction.
- the output of the multiplexer MUX is applied to an input of an adder ADD whose output is applied to the input of the buffer SPC.
- the PC address at the output of the SPC buffer is returned via the B1 register to the second input of the adder ADD.
- the memory MEM receives the read signal RDH1 and delivers the instruction SUB RI, R2,
- the PREDEC decoder receives the CODE SUB code and sets the SEL signal to 0 because the JMPC CODE code is not recognized.
- 1.2 - Hl 0, H2 1, register B2 triggered, sector ST12 active: the DECl decoder receives the instruction, sends the ADRI and ADR2 addresses towards the REGBANK bank and the CODE SUB operation code towards the DEC2 decoder (these data are intercepted and blocked by register B3), the multiplexer MUX receives the signal SEL (at 0), selects the input "+1", so that the output of the adder ADD delivers the value ADD ⁇ + 1, ie ADD 2 , - the RPC register receives the write signal WRH2 and the PC address (formerly ADDp becomes ADD 2 .
- the RPC register receives the signal RH2 and the PC address becomes equal to ADD n or to ADD 3 depending on the value of the signal
- the steps described above overlap as summarized in the table below. It can be seen that the first step 2.1 of the JMPc instruction begins at the same time as step 1.3 of the SUB instruction, that is to say in the second pipeline cycle of this instruction. It can also be seen that the first step 3.1 of reading a new instruction following the JMPc instruction can begin after step 2.2, the JMPc instruction having been executed and the PC address updated in a single pipeline cycle. At the end of step 2.2, the PC address is indeed equal to ADD n or to ADD 3 depending on whether the jump condition has been achieved or not. Board
- the particularly simple structure of the microprocessor according to the invention guarantees low current consumption.
- An additional measure making it possible to limit the electrical consumption consists in having on the outputs OUTRs, OUTRd of the REGBANK bank of the buffer registers controlled by a signal Hl 'shifted by duration ⁇ t relative to the signal Hl. In this way, the ALU unit is only connected to the outputs OUTRs, OUTRd when these are stabilized.
- the shifted signal Hl ' can be obtained simply by means of a delay circuit, for example cascading reverse gates.
- FIG. 2 represents a variant MP2 of the microprocessor according to the invention which differs from the previous embodiment by the fact that the decoding of the instruction JMPc is entrusted to a decoder DEC3 which is located at the location of the decoder DECl previously described and performs the operations of the PREDEC and DECl decoders.
- the decoder DEC3 notably delivers the signal SEL during step 2.2, with reference to the sequence described above. It should be noted that the signal SEL can be delivered quickly by the decoder DEC3 during step 2.2 since the bits C, N, P, Z are validated at the end of step 2.1 (which corresponds to the 'calculation step 1.3 of the previous instruction).
- the bits C, N, P, Z delivered by the unit ALU can be applied to the decoder DEC3 via the register B4, for example by means of a bus 11 shown in dotted lines.
- this does not necessarily mean that the output of the ALU unit is connected directly to the means of decoding.
- the bits C, N, P, Z are taken at the output of the unit ALU or on the bus which transports them to the register Rf instead of being read in the register Rf as in the prior art .
- microprocessor may include various other means, for example means for decoding complex instructions, means for writing data into the memory, etc.
- a variant consists in carrying out an "absolute" jump operation by reversing the arrangement of the MUX multiplexer and of the adder ADD (not shown).
- the ADD adder receives the PC address and the +1 increment value as input.
- the output of the adder ADD is applied to an input of the multiplexer MUX which always receives on its other input the jump value VAL.
- the output of the MUX multiplexer is applied to the RPC register.
- the PC address is incremented by the value + 1 or is brought to the value VAL depending on whether the signal SEL is at 1 or at 0.
- the value VAL represents, in this case, an absolute jump address and not a relative jump value.
- the operation of the pipeline sectors in half clock cycles can be obtained in an equivalent manner by providing a clock signal having a double speed, the pipeline sectors not being activated only one clock cycle out of two by shifting by one clock cycle the activation of each sector relative to the next sector.
- the selection of one clock bit out of two as control signal can be made in a simple manner by means of a frequency divider circuit, for example a flip-flop D whose output / Q is brought back to input D.
- the control signals of the pipeline registers can moreover be combined in exclusive OR gates before being applied, in order to guarantee that two complementary sectors of the same pipeline stage will not be activated simultaneously.
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Executing Machine-Instructions (AREA)
- Advance Control (AREA)
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
DE69828305T DE69828305T2 (de) | 1997-11-03 | 1998-10-23 | Pipelinemikroprozessor mit verarbeitung von bedingten verzweigungungen in einem einzigen taktzyklus |
AU21185/99A AU2118599A (en) | 1997-11-03 | 1998-10-23 | Pipeline microprocessor with conditional jump in a clock cycle |
EP98951538A EP1027643B1 (fr) | 1997-11-03 | 1998-10-23 | Microprocesseur pipeline a saut conditionnel en un cycle d'horloge |
US09/564,477 US6678819B1 (en) | 1997-11-03 | 2000-05-03 | Pipeline microprocessor with conditional jump in one clock cycle |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR97/13757 | 1997-11-03 | ||
FR9713757A FR2770662B1 (fr) | 1997-11-03 | 1997-11-03 | Microprocesseur pipeline a saut conditionnel en un cycle d'horloge |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/564,477 Continuation US6678819B1 (en) | 1997-11-03 | 2000-05-03 | Pipeline microprocessor with conditional jump in one clock cycle |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1999023551A1 true WO1999023551A1 (fr) | 1999-05-14 |
Family
ID=9512938
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/FR1998/002265 WO1999023551A1 (fr) | 1997-11-03 | 1998-10-23 | Microprocesseur pipeline a saut conditionnel en un cycle d'horloge |
Country Status (6)
Country | Link |
---|---|
US (1) | US6678819B1 (fr) |
EP (1) | EP1027643B1 (fr) |
AU (1) | AU2118599A (fr) |
DE (1) | DE69828305T2 (fr) |
FR (1) | FR2770662B1 (fr) |
WO (1) | WO1999023551A1 (fr) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE19945940C2 (de) * | 1999-09-24 | 2002-01-17 | Infineon Technologies Ag | Verfahren und Vorrichtung zur Bearbeitung bedingter Sprungbefehle in einem Prozessor mit PIPELINE-Rechnerarchitektur |
DE10061064B4 (de) * | 2000-12-08 | 2004-02-26 | Conti Temic Microelectronic Gmbh | Verfahren zum Betrieb eines motorangetriebenen Kraftfahrzeugs |
CN104331268B (zh) * | 2014-10-27 | 2017-05-03 | 杭州中天微系统有限公司 | 一种用于低功耗处理器的加快条件跳转执行的装置 |
CN113946540B (zh) * | 2021-10-09 | 2024-03-22 | 深圳市创成微电子有限公司 | 一种dsp处理器及其判断跳转指令的处理方法 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0378816A2 (fr) * | 1988-12-19 | 1990-07-25 | Bull HN Information Systems Inc. | Méthode et dispositif d'exécution d'instructions en pipeline |
US5349671A (en) * | 1989-03-23 | 1994-09-20 | Matsushita Electric Industrial Co., Ltd. | Microprocessor system generating instruction fetch addresses at high speed |
US5485629A (en) * | 1993-01-22 | 1996-01-16 | Intel Corporation | Method and apparatus for executing control flow instructions in a control flow pipeline in parallel with arithmetic instructions being executed in arithmetic pipelines |
US5617574A (en) * | 1989-05-04 | 1997-04-01 | Texas Instruments Incorporated | Devices, systems and methods for conditional instructions |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4943915A (en) * | 1987-09-29 | 1990-07-24 | Digital Equipment Corporation | Apparatus and method for synchronization of a coprocessor unit in a pipelined central processing unit |
JPH081599B2 (ja) * | 1988-02-24 | 1996-01-10 | 三菱電機株式会社 | データ処理装置 |
US5136696A (en) * | 1988-06-27 | 1992-08-04 | Prime Computer, Inc. | High-performance pipelined central processor for predicting the occurrence of executing single-cycle instructions and multicycle instructions |
US4974155A (en) * | 1988-08-15 | 1990-11-27 | Evans & Sutherland Computer Corp. | Variable delay branch system |
-
1997
- 1997-11-03 FR FR9713757A patent/FR2770662B1/fr not_active Expired - Fee Related
-
1998
- 1998-10-23 WO PCT/FR1998/002265 patent/WO1999023551A1/fr active IP Right Grant
- 1998-10-23 DE DE69828305T patent/DE69828305T2/de not_active Expired - Lifetime
- 1998-10-23 AU AU21185/99A patent/AU2118599A/en not_active Abandoned
- 1998-10-23 EP EP98951538A patent/EP1027643B1/fr not_active Expired - Lifetime
-
2000
- 2000-05-03 US US09/564,477 patent/US6678819B1/en not_active Expired - Lifetime
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0378816A2 (fr) * | 1988-12-19 | 1990-07-25 | Bull HN Information Systems Inc. | Méthode et dispositif d'exécution d'instructions en pipeline |
US5349671A (en) * | 1989-03-23 | 1994-09-20 | Matsushita Electric Industrial Co., Ltd. | Microprocessor system generating instruction fetch addresses at high speed |
US5617574A (en) * | 1989-05-04 | 1997-04-01 | Texas Instruments Incorporated | Devices, systems and methods for conditional instructions |
US5485629A (en) * | 1993-01-22 | 1996-01-16 | Intel Corporation | Method and apparatus for executing control flow instructions in a control flow pipeline in parallel with arithmetic instructions being executed in arithmetic pipelines |
Non-Patent Citations (3)
Title |
---|
JOUPPI N P: "THE NONUNIFORM DISTRIBUTION OF INSTRUCTION-LEVEL AND MACHINE PARALLELISM AND ITS EFFECT ON PERFORMANCE", IEEE TRANSACTIONS ON COMPUTERS, vol. 38, no. 12, 1 December 1989 (1989-12-01), pages 1645 - 1658, XP000096380 * |
MONTANARO J ET AL: "A 160-MHZ, 32-B, 0.5-W CMOS RISC MICROPROCESSOR", IEEE JOURNAL OF SOLID-STATE CIRCUITS, vol. 31, no. 11, November 1996 (1996-11-01), pages 1703 - 1714, XP000691455 * |
PARKER AND VENESKI: "Control structure for making fast conditional branch decisions", IBM TECHNICAL DISCLOSURE BULLETIN, vol. 27, no. 2, July 1984 (1984-07-01), ARMONK,US, pages 1239 - 1240, XP002070445 * |
Also Published As
Publication number | Publication date |
---|---|
DE69828305D1 (de) | 2005-01-27 |
AU2118599A (en) | 1999-05-24 |
EP1027643B1 (fr) | 2004-12-22 |
FR2770662A1 (fr) | 1999-05-07 |
EP1027643A1 (fr) | 2000-08-16 |
FR2770662B1 (fr) | 1999-12-10 |
DE69828305T2 (de) | 2005-12-08 |
US6678819B1 (en) | 2004-01-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6304955B1 (en) | Method and apparatus for performing latency based hazard detection | |
EP1027644B1 (fr) | Microprocesseur, notamment pour carte a puce et procede de traitement d'une instruction | |
EP0712133A1 (fr) | Procédé de lecture anticipée de mémoire à accès série et mémoire s'y rapportant | |
EP3503103A1 (fr) | Système comportant une mémoire adaptée à mettre en oeuvre des opérations de calcul | |
EP0006478A1 (fr) | Dispositif programmable de verrouillage de signaux de commande dans un système de traitement de données | |
EP0155731B1 (fr) | Dispositif d'adressage pour fournir à une mémoire des codes d'adresse | |
EP1027643B1 (fr) | Microprocesseur pipeline a saut conditionnel en un cycle d'horloge | |
EP0638904B1 (fr) | Mémoire à double accès | |
FR3101980A1 (fr) | Processeur | |
FR2678400A1 (fr) | Processeur de protocole destine a l'execution d'un ensemble d'instructions en un nombre reduit d'operation. | |
FR2604270A1 (fr) | Additionneur binaire comportant un operande fixe, et multiplieur binaire parallele-serie comprenant un tel additionneur | |
US6263424B1 (en) | Execution of data dependent arithmetic instructions in multi-pipeline processors | |
FR2770661A1 (fr) | Microprocesseur comprenant des moyens de concatenation de bits | |
EP0733977B1 (fr) | Système informatique avec mémoires hiérarchisées | |
FR2626691A1 (fr) | Multiplieur numerique generalise et filtre numerique mettant en oeuvre ce multiplieur | |
EP1338956A1 (fr) | Dispositif électronique de traitement de données, en particulier processeur audio pour un décodeur audio/vidéo | |
EP0437127B1 (fr) | Circuit décaleur avec générateur de bits de parité | |
WO2000010089A1 (fr) | Memoire auto-associative adressable par donnees d'entree a forme redondante | |
US6901490B2 (en) | Read/modify/write registers | |
EP0169089B1 (fr) | Dispositif élémentaire de traitement de données | |
FR2809835A1 (fr) | Systeme et procede d'arrondissement base sur une selection pour des operations a virgule flottante | |
FR2737589A1 (fr) | Systeme d'organisation et procede de sequencement des circuits d'un microprocesseur | |
EP0595707B1 (fr) | Séquenceur linéaire avec signaux de sortie binaires | |
EP2186093B1 (fr) | Mémoire partagée | |
US20040199563A1 (en) | Noise invariant circuits, systems and methods |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AU CA CN JP KR US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH CY 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: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 09564477 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1998951538 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 1998951538 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: CA |
|
WWG | Wipo information: grant in national office |
Ref document number: 1998951538 Country of ref document: EP |