Qualcomm Technologies, Inc.

Arteris, Inc.

Qualcomm Incorporated

Arteris S.a.

T. Sqware Inc.

Philippe Boucard

Jean-Jacques Lecler

Laurent Rene Moll

Luc Montperrus

Cesar Douady

Philippe Martin

César Douady

Boris Boutillier

Vincent Vacquerie

Rakesh Kumar Gupta

Christophe Avoinne

Christophe Layer

Pascal Godet

Olivier Loison

Ameline Le Rouzic

C&eacute;sar Douady

Nordine Chaibelaine

This disclosure provides systems, methods, and devices for memory systems that support enhanced exclusive access fairness operations for scalable exclusive monitor architectures. In a first aspect, device includes a processing system that includes one or more shared memory devices and one or more request nodes. The processing system also includes one or more network interface units (NIUs), each NIU including an exclusive monitor configured to monitor exclusive accesses to shared memory addresses for a corresponding request node and including a timeout register for the exclusive monitor configured to control exclusive access fairness. The processing system includes one or more home nodes coupled to a corresponding shared memory device of the one or more shared memory devices. The processing system further includes an interconnect configured to couple the one or more request nodes to the one or more shared memory devices. Other aspects and features are also claimed and described.

Method and apparatus for exclusive access fairness in memory systems with distributed exclusive access management

Data may be communicated from a sender device to a receiver device over enabled or selected byte positions or other data bit groups of a data bus. The sender device may determine data values to be sent over the data bus and may determine which byte positions are enabled or selected and which are not selected. The sender device may also determine a code. The code may be a value that is not included in the data values to be sent over the data bus. The sender device may then send the selected data values in selected byte positions of the data bus and send the code in non-selected byte positions of the data bus. The sender device may also send the code to the receiver device separately from the data bit lanes of the data bus.

Encoding byte information on a data bus with separate code

Aspects include computing devices, apparatus, and methods implemented by the apparatus for implementing multiple split snoop directories on a computing device having any number of processors, any number of power domains, and any number of processor caches. For example, various aspects may include enabling a first split snoop directory for a first power domain and a second split snoop directory for a second power domain, wherein the first power domain includes a first plurality of processor caches and the second power domain includes at least one processor cache, determining whether all of the first plurality of processor caches are in a low power state, and disabling the first split snoop directory in response to determining that the first plurality of processor caches are in a low power state. Similar operations may be performed for N number of power domains and M number of processor caches.

Coherent interconnect power reduction using hardware controlled split snoop directories

Packet format configurability is extended for packets transported on physical links of an Intellectual Property (IP) core interconnect by using at least two independent parameters: one parameter governing data-width and one parameter governing latency penalty. The at least two independent parameters allow creation of transport protocol packets without additional latency insertion, which is useful for low-latency applications. The at least two independent parameters also allow creation of narrow packets with multi-cycle additional latency, which is useful for latency tolerant, area sensitive applications.

Zero-latency network on chip (NoC)

Systems and methods relate to a network on chip (NoC) which includes one or more channels configured to carry data packets in a first direction, the first direction having an upstream end and a downstream end. A tunnel is configured between an upstream element at the upstream end and a downstream element at the downstream end. The tunnel includes common wires which are shared by the one or more channels. The tunnel is configured to transmit data packets of two or more formats on the common wires based on common signals. common signals comprise data signals to transmit one or more of data, control, or debug information belonging to the data packets on the common wires, and framing signals to control transmission of the data signals on the common wires.

Tunneling within a network-on-chip topology

The invention is a transaction interface protocol wherein the interface protocol has a transaction identifier signal in each of the request and response channels. It is used between a target network interface unit (NIU) master and an initiator NIU slave that are directly connected through a transaction interface. The target NIU response channel uses the transaction ID signal to identify the entry in a context array associated with the corresponding request. The coupling of target NIU and initiator NIU enable the formation of an on-chip interconnect comprising multiple network-on-chip (NoCs) wherein the topology of the interconnect is simpler, smaller, faster, and has lower latency.

Network on a chip socket protocol

A system with a prefetch address generator coupled to a system translation look-aside buffer that comprises a translation cache. Prefetch requests are sent for page address translations for predicted future normal requests. Prefetch requests are filtered to only be issued for address translations that are unlikely to be in the translation cache. Pending prefetch requests are limited to a configurable or programmable number. Such a system is simulated from a hardware description language representation.

DMA engine with STLB prefetch capabilities and tethered prefetching

System TLBs are integrated within an interconnect, use a and share a transport network to connect to a shared walker port. Transactions are able to pass STLB allocation information through a second initiator side interconnect, in a way that interconnects can be cascaded, so as to allow initiators to control a shared STLB within the first interconnect. Within the first interconnect, multiple STLBs share an intermediate-level translation cache that improves performance when there is locality between requests to the two STLBs.

System translation look-aside buffer integrated in an interconnect

A power disconnect unit within a data transport topology of a NoC includes an asynchronous clock domain adapter unit inserted between a master side manager unit and a slave side manager unit. This configuration allows for the master and slave side managers of the power disconnect unit to be placed physically far apart on the chip, relieving the need to route long power rail signals on the chip. A response data path and associated asynchronous clock domain adapter unit is optionally included on the chip. A path to bypass the asynchronous clock domain adapter units is optionally included on the chip to enable a fully synchronous mode of operation without the data latency cost of the asynchronous adapter unit.

Power disconnect unit for use in data transport topology of network on chip design having asynchronous clock domain adapter sender and receiver each at a separate power domain

Method of managing priority during the transmission of a message, in an interconnections network comprising at least one transmission agent which comprises at least one input and at least one output, each input comprising a means of storage organized as a queue of messages. A message priority is assigned during the creation of the message, and a queue priority equal to the maximum of the priorities of the messages of the queue is assigned to at least one queue of messages of an input. A link priority is assigned to a link linking an output of a first transmission agent to an input of a second transmission agent, equal to the maximum of the priorities of the queues of messages of the inputs of said first agent comprising a first message destined for that output of said first agent which is coupled to said link, and the priority of the link is transmitted to that input of said second agent which is coupled to the link.

Method and device for managing priority during the transmission of a message

A system TLB accepts translation prefetch requests from initiators. Misses generate external translation requests to a walker port. Attributes of the request such as ID, address, and class, as well as the state of the TLB affect the allocation policy of translations within multiple levels of translation tables. Translation tables are implemented with SRAM, and organized in groups.

System translation look-aside buffer with request-based allocation and prefetching

A Network-on-Chip (NoC) is provided that performs reordering of transaction responses such as those with requests that cross address mapping boundaries. Ordering is ensured by filtering logic in reorder buffers, some of which include storage to allow multiple simultaneously pending transactions. Transactions are transported by a packet-based transport protocol. The reorder buffering is done at packet level, within the transport topology. Reorder buffers are distributed physically throughout the floorplan of the chip, they have localized connectivity to initiators, and they operate in separate power and clock domains.

System and method of distributed initiator-local reorder buffers

A cache coherency controller, a system comprising such, and a method of its operation are disclosed. The coherency controller ensures that target-side security checking rules are not violated by the performance-improving processes commonly used in coherency controllers such as dropping, merging, invalidating, forwarding, and snooping. This is done by ensuring that requests marked for target-side security checking and any other requests to overlapping addresses are forwarded directly to the target-side security filter without modification or side effects.

Growing community of inventors

Le Chesnay, France

Philippe Boucard