International Business Machines Corporation

David Allan Webber

Timothy Gerard McNamara

Michael Thomas Vaden

Fadi Y Busaba

Gary A Van Huben

Khary J Alexander

Phillip John Restle

Michael H Wood

Brian William Curran

Patrick M West, Jr

Martin Stanley Schmookler

Alina Deutsch

Peter J Camporese

Lawrence Joseph Powell, Jr

Jonathan Yang Chen

Ashutosh Goyal

Patrick James Meaney

Christopher J Berry

Tin-Chee Lo

Michael Billeci

George A Katopis

Ching-Lung L Tong

Paul R Turgeon

Jeffrey A Magee

Giorgos Stellos Tsapepas

Kutalmis Koyuncu

Joseph L Braun

According to an embodiment, a computer system for cache management includes a processor and a cache, the computer system configured to perform a method including receiving a first store request for a first address in the cache and receiving a first fetch request for the first address in the cache. The method also includes executing the first store request and the first fetch request, latching the first store request in a store write-back pipeline in the cache, detecting, in the processor, a conflict following execution of the first store request and the first fetch request and receiving the first store request from a recycle path including the store write-back pipeline and executing the first store request a second time.

Conflict resolution of cache store and fetch requests

A method of automating circuit design is provided and includes storing one or more circuit design schematics in a memory, providing, by way of an interface, a plurality of search parameters for searching for nets of the schematics in the memory, searching for nets of the schematics in the memory in accordance with search parameters input into the interface and presenting, by way of the interface, information associated with a net matching the received search parameters.

Schematic wire annotation tool

A method of discovering a fault in a circuit is disclosed. The method comprises generating a first result of a selected function by performing the selected function on an operand, wherein the selected function employs a mask. Once the function is performed, an antimask of the mask is created, and the modulo of the antimask is calculated. A modulo function of the first result of the selected function is calculated to obtain a third result. A modulo of the operand is then calculated to obtain a fourth result, and a second function is then performed on the second result and the third result to obtain a fifth result. In response to comparing the fifth result to the fourth result, a signal is propagated to indicate a fault in the circuit.

Method and system for checking rotate, shift and sign extension functions using a modulo function

A method of operating an arithmetic logic unit (ALU) by inverting a result of an operation to be executed during a current cycle in response to control signals from instruction decode logic which indicate that a later operation will require a complement of the result, wherein the result is inverted during the current cycle. The later operation may be a subtraction operation that immediately follows the first operation. The later instruction is decoded prior to the current cycle to control the inversion in the ALU. The ALU includes an adder, a rotator, and a data manipulation unit which invert the result during the current cycle in response to an invert control signal. The second operation subtracts the result during a subsequent cycle in which a carry control signal to the adder is enabled, and the rotator and the data manipulation unit are disabled. The ALU may be used in an execution unit of a microprocessor, such as a fixed-point unit.

Inverting data on result bus to prepare for instruction in the next cycle for high frequency execution units

A double data rate elastic interface in which programmable latch stages provide an elastic delay, preferably on the driving side of the elastic interface. However, the invention is not limited to the driver side/chip, it can be implemented in the receiver side/chip as well. However, since the receiver side of an elastic interface already has complicated logic, the invention will be usually implemented on the driving side. The programmable latch stages on the driving chip side of the interface, can often operate at the local clock frequency (the same frequency as the elastic interface bus clock frequency), which in turn is half of the double data rate at which the receiving latch stages operate, thereby decreasing the logic and storage resources in the interface receivers. The programmable latch stages can also be used in the case that the local clock frequency is twice the elastic interface bus clock frequency.

Programmable bus driver launch delay/cycle delay to reduce elastic interface elasticity requirements

A double data rate launch system and method in which the two-to-one multiplexer select signal delay is programmable and can be adjusted individually for each system. This allows the amount of delay to be minimized based on the actual set up time required, not the worst-case set-up time. The select signal to the multiplexer is delayed sufficiently to compensate for non-uniformity of duty cycle of data at the inputs to the multiplexer. Compensation of the non-uniformity allows the data on the wire to have a uniform duty cycle for all data transferred regardless of which latch is sourcing the data. The multiplexer that selects data from the two latches which are launching data on the edge of different clocks has a select line that is delayed by a variable amount to tune the select such that the data is clean at the input to the multiplexer on all ports.

Mechanism for windaging of a double rate driver

Design Data Management uses one copy of common data sets along with a plurality of instances, while continuing to utilize the existing design databases and existing CAD tools. Allowing a minimum amount of user intervention to create and maintain the common data set, Design Data Management employs replicating common data sets into one or more clone data sets. The method preferred provides for replicating and synchronizing one or more data sets with a master data set, comprises providing data design management of a master data set and at least one clone data set, and copying a master physical design data set into one or more physical instances to enable customization of said one or more physical instances. The master data set describes at least one of: a design component, a circuit macro, and a circuit entity, and comprises logical data sets, and it comprise physical design data sets. This permits all existing verification processes that are normally executed against the common data set to also be equally applied to the clones of said data set by way of automatic synchronization between of the common dataset and the clones.

Method for replicating and synchronizing a plurality of physical instances with a logical master

A method for determining gate array distribution includes steps or acts of: randomly placing a plurality of test boxes in a logic circuit layout; counting the number of fill cells in each of the plurality of test boxes; recording the count; grouping the plurality of test boxes into two groups: a first group with local clock buffers and a second group without local clock buffers; determining the fill cell percentage of each of the plurality of test boxes; and flagging the test boxes with a poor distribution of gate array cells.

Spare gate array cell distribution analysis

A method and apparatus for operating a clock in a processor having asymmetrically mirrored base-mirror units is disclosed. The method includes initializing a base-unit and a mirror-unit of the processor to the same state, and starting the mirror-unit-clock one clock cycle later than the base-unit-clock.

Method and apparatus for controlling clocks in a processor with mirrored units

An X-Y grid tree clock distribution network for distributing a clock signal across a VLSI chip. Tunable wiring tree networks are combined with an X-Y grid vertically and horizontally connecting all the tree end points. No drivers are necessary at connection points of the tree end points to the X-Y grid. The final X-Y grid distributes the clock signal close to every place it is needed, and reduces skew across local regions. A tuning method allows buffering of the clock signal, while minimizing both nominal clock skew and clock uncertainty. The tuned tree networks provide low skew even with variations in clock load density and non-ideal buffer placement, while minimizing the number of buffers needed. The tuning method first represents a total capacitance of one or more of clock pin loads and twig wiring as a clustered grid load. Next, a smoothing of the clustered grid loads approximates the effect of the X-Y grid. Electrical simulation models are created for network components and clustered grid loads are substituted with smoothed clustered grid loads. A set of N,electrical net lists are next created by extracting a net list with associated X-Y grid wires cut to isolate each sector net list from its neighboring sectors. Each N,electrical net list is then tuned, wherein the smoothed clustered grid loads represent an approximation of the effects of the neighboring sectors of each N,electrical net list.

X-Y grid tree clock distribution network with tunable tree and grid networks

X-Y grid tree tuning method

A method for dynamic bandwidth testing of a link between two computer subsystems useful for determining the amount of data which can be buffered in a transmission line by delay, in which at each end of the line circuit modules are provided to couple the subsystem by a bi-directional multi-bit (BiDi) link, and providing within each said circuit module a built-in circuit and logic for dynamic transmission characterization and test of a said BiDi link between computer subsystems using built-in characterization logic macros, and during a test mode, switching said said built-in circuit and logic to test mode and using the test mode to characterize the link performance, and after the completion of characterization, the switching built-in characterization logic macros back to a normal system mode after programmatically setting timing parameters for the BiDi link to ensure safe operation of data transfer before the BiDi link is switched to system mode.

Growing community of inventors

Poughkeepsie, NY, United States of America

David Allan Webber