Arm Limited

Riverlane Ltd.

Kauser Yakub Johar

Antony John Penton

Jatin Bhartia

Simon John Craske

Balaji Venu

Max John Batley

Alexander Alfred Hornung

Loïc Pierron

Ben Andrew Barber

Luka Skoric

Michele Riga

Francisco João Feliciano Gaspar

Mohammadi Shabbirhussain Bharmal

Kenton Michael Barnes

Marco Bonino

Zemian Hughes

A decoder apparatus for decoding syndromes of a quantum error correction code, the syndromes comprising measurement data from a quantum computer, the quantum computer comprising an array of qubits including syndrome qubits and data qubits. The decoder apparatus is configured to: receive syndrome index data representative of: a physical location of each of the array of qubits within the quantum computer; and lattice dimensions of the array of qubits; receive the syndromes of the quantum error correction code from the quantum computer; determine physical co-ordinate positions for each of the array of qubits based on the syndrome index data; decode the syndromes of the quantum error correction code using the determined physical co-ordinate positions. The decoder apparatus has a first syndrome capacity for decoding quantum error correction code syndromes and the syndromes of the quantum error correction code occupy a second syndrome capacity less than or equal to the first syndrome capacity of the decoder apparatus.

Quantum error correction code decoder and associated methods

A processing systemincludes a processing pipelinewhich includes fetch circuitryfor fetching instructions to be executed from a memory. Buffer control circuitryis responsive to a programmable trigger, such as explicit hint instructions delimiting an instruction burst, or predetermined configuration data specifying parameters of a burst together with a synchronising instruction, to trigger the buffer control circuitry to stall a stallable portion of the processing pipeline (e.g. issue circuitry), to accumulate within one or more buffersfetched instructions starting from a predetermined starting instruction, and, when those instructions have been accumulated, to restart the stallable portion of the pipeline.

Programmable instruction buffering for accumulating a burst of instructions

An error protection key generation method and system are provided, the method being used to generate a key for use in computing an error protection code for an input data value according to a chosen error protection scheme. The method comprises inputting a plurality of desired data value sizes, and then applying a key generation algorithm to generate a key for use in computing the error protection code for a maximum data value size amongst the plurality of data value sizes. The key generation algorithm is arranged so that it generates the key so as to comprise a plurality of sub-keys, where each sub-key is associated with one of the input data value sizes, and where each sub-key conforms to a key requirement of the error protection scheme. As a result, a generic key is produced containing a plurality of sub-keys, where each sub-key is associated with a particular desired data value size, and can be extracted and used independently given that each sub-key conforms to the error protection scheme requirements. This provides significant benefits in the design and verification of error protection circuits using such keys.

Error protection key generation method and system

An apparatus is described comprising a cluster of processing elements. The cluster having a split mode in which the processing elements are configured to process independent processing workloads, and a lock mode in which the processing elements comprise at least one primary processing element and at least one redundant processing element, each redundant processing element configured to perform a redundant processing workload for checking correctness of a primary processing workload performed by the primary processing element. Each processing element has an associated local memory comprising a plurality of memory locations. A local memory access control mechanism is configured, during the lock mode, to allow the at least one primary processing element to access memory locations within the local memory associated with the at least one redundant processing element.

Efficient memory utilisation in a processing cluster having a split mode and a lock mode

Circuitry comprises data handling circuitry having a memory, the data handling circuitry being operable in a primary mode in which the data handling circuitry performs a data handling function by accessing the memory and in a secondary mode in which the data handling circuitry performs the data handling function independently of the memory; test circuitry to control a test operation during execution of a set of data processing instructions by a data processor configured to execute data processing instructions by reference to the data handling function performed by the data handling circuitry; in which: the test circuitry is configured to control the data handling circuitry to transition from the primary mode to the secondary mode in response to initiation of a test operation on the memory so that the data processor executes one or more of the set of data processing instructions by reference to the data handling function performed by the data handling circuitry in the secondary mode at least while the test operation is performed on the memory; and the test circuitry is configured to control the data handling circuitry to return to the primary mode in response to completion of the test operation on the memory.

Data handling circuitry performing memory data handling function and test circuitry performing test operation during execution of memory data processing

A processing system includes a processing pipeline which includes fetch circuitry for fetching instructions to be executed from a memory. Buffer control circuitry is responsive to a programmable trigger, such as explicit hint instructions delimiting an instruction burst, or predetermined configuration data specifying parameters of a burst together with a synchronising instruction, to trigger the buffer control circuitry to stall a stallable portion of the processing pipeline (e.g. issue circuitry), to accumulate within one or more buffers fetched instructions starting from a predetermined starting instruction, and, when those instructions have been accumulated, to restart the stallable portion of the pipeline.

Programmable instruction buffering of a burst of instructions including a pending data write to a given memory address and a subsequent data read of said given memory address

A data processing apparatus has prefetch circuitry for prefetching instructions from a data store into an instruction queue. Branch prediction circuitry is provided for predicting outcomes of branch instructions and the prefetch circuitry may prefetch instructions subsequent to the branch based on the predicted outcome. Instruction identifying circuitry identifies whether a given instruction prefetched from the data store is a predetermined type of program flow altering instruction and if so then controls the prefetch circuitry to halt prefetching of subsequent instructions into the instruction queue.

Instruction prefetch halting upon predecoding predetermined instruction types

Apparatus for data processing and a method of data processing are provided, according to which the processing circuitry of the apparatus can access a memory system and execute data processing instructions in one context of multiple contexts which it supports. When the processing circuitry executes a barrier instruction, the resulting access ordering constraint may be limited to being enforced for accesses which have been initiated by the processing circuitry when operating in an identified context, which may for example be the context in which the barrier instruction has been executed. This provides a separation between the operation of the processing circuitry in its multiple possible contexts and in particular avoids delays in the completion of the access ordering constraint, for example relating to accesses to high latency regions of memory, from affecting the timing sensitivities of other contexts.

Context sensitive barriers with an implicit access ordering constraint for a victim context

Apparatus and a method for processor core self-testing are disclosed. The apparatus comprises processor core circuitry to perform data processing operations by executing data processing instructions. Separate self-test control circuitry causes the processor core circuitry to temporarily switch from a first state of executing the data processing instructions to a second state of executing a self-test sequence of instructions, before returning to the first state of executing the data processing instructions without a reboot of the processor core circuitry being required. There is also self-test support circuitry, wherein the processor core circuitry is responsive to the self-test sequence of instructions to cause an export of at least one self-test data item via the self-test support circuitry to the self-test control circuitry.

Self-testing in a processor core

An apparatus and method are provided for generating an error code for a block comprising a plurality of data bits and a plurality of address bits. The apparatus has block generation circuitry to generate a block comprising a plurality of data bits and a plurality of address bits, and error code generation circuitry for receiving that block and a mask array comprising a plurality of mask rows, and for then applying an error code generation algorithm to generate an error code for the block. The error code comprises a plurality of check bits, where each check bit is determined using the block and a corresponding mask row of the mask array. Each mask row comprises a plurality of mask bits, each mask bit being associated with a corresponding bit of the block. At least one mask row has its mask bit values constrained so as to ensure that when all of the data bits of the block have the same value, the error code generated by the error code generation circuitry has at least one check bit having a different value to the value of the data bits irrespective of the value of the address bits. In addition to supporting detection and/or correction of errors in the data bits, such an approach also allows memory address decode errors to be detected while in addition allowing detection of stuck at zero or stuck at one errors in a memory's output.

Apparatus and method for generating an error code for a block comprising a plurality of data bits and a plurality of address bits

A data processing apparatus and method of processing data are disclosed according to which a processor unit is configured to issue write access requests for memory which are buffered and handled by a memory access buffer. A cache unit is configured, in dependence on an allocation policy defined for the cache unit, to cache accessed data items. Memory transactions are constrained to be carried out so that all of a predetermined range of memory addresses within which one or more memory addresses specified by the buffered write access requests lie must be written by the corresponding write operation. If the buffered write access requests do not comprise all memory addresses within at least two predetermined ranges of memory addresses, and the cache unit is configured to operate with a no-write allocate policy, the data processing apparatus is configured to cause the cache unit to subsequently operate with a write allocate policy.

Growing community of inventors

Cambridge, United Kingdom

Kauser Yakub Johar