Abstract: Devices and techniques for memory access bounds checking for a programmable atomic operator are described herein. A processor can execute a programmable atomic operator with a base memory address. The processor can obtain a memory interleave size indicator corresponding to the programmable atomic operator and calculate a contiguous memory address range from the base memory address and the memory interleave size. The processor can then detect that a memory request from the programmable atomic operator is outside the contiguous memory address range and deny the memory request when it is outside of the contiguous memory address range and allow the memory request otherwise.

Abstract: Devices and techniques for thread scheduling control and memory splitting in a barrel processor are described herein. An apparatus includes a barrel processor, which includes thread scheduling circuitry, where the barrel processor is configured to perform operations through use of the thread scheduling circuitry, the operations including those to: place a thread to be scheduled in one of two groups: a first group and a second group, wherein the first group is associated with a first processor storage device, and the second group is associated with a second processor storage device; and schedule a current thread to place into a pipeline for the barrel processor, the scheduling performed by alternating between threads in the first group and threads in the second group.

Abstract: Devices and techniques for on-demand programmable atomic kernel loading are described herein. A programmable atomic unit (PAU) of a memory controller can receive an invocation of a programmable atomic operator by the memory controller. The PAU can then perform a verification on a programmable atomic operator partition for the programmable atomic operator. Here, the programmable atomic operator partition is located in a memory of the PAU. The PAU can then signal a trap in response to the verification indicating that the programmable atomic operator partition is not prepared.

Abstract: Devices and techniques for rescheduling a failed memory request in a processor are described herein. When a memory request for a thread is denied at a point in the execution pipeline of the processor beyond a thread rescheduling point, the thread can be placed into a memory response path of the processor. An indicator that a register write-back will not occur for the thread can also be provided. Then, the thread can be rescheduled with other threads in the memory response path.

Abstract: A chiplet system can include a Serial Peripheral Interface (SPI) bus for communication. A controller or primary device coupled to the SPI bus can generate a message with read or write instructions for one or more secondary devices. In an example, the primary device can be configured to use information on a data input port or data input bus to determine a communication status of one or multiple secondary devices on the bus.

Abstract: A chiplet system can include a Serial Peripheral Interface (SPI) bus for communication. A controller or primary device coupled to the SPI bus can generate a message with read or write instructions for one or more secondary devices. In an example, the primary device can be configured to read information from a secondary device about whether the secondary device supports parity-protected data communications. The primary device can be configured to selectively send or receive parity-protected data communications depending on a capability of the secondary device to support parity.

Abstract: Devices and techniques for thread replay to preserve state in a barrel processor are described herein. An apparatus includes a barrel processor, which includes a temporary memory; and a thread scheduling circuitry; wherein the barrel processor is configured to perform operations through use of the thread scheduling circuitry, the operations including those to: schedule a current thread to place into a pipeline for the barrel processor on a clock cycle, the barrel processor to schedule threads on each clock cycle; store the current thread in the temporary memory; detect that no thread is available on a clock cycle subsequent to the cycle that the current thread is scheduled; and in response to detecting that no thread is available on the subsequent clock cycle, repeat scheduling the current thread based on the contents of the temporary memory.

Abstract: A chiplet system comprises an interposer including interconnect and multiple chiplets arranged on the interposer and interconnected using the interconnect of the interposer. The multiple chiplets include a throttle level bus source chiplet including a throttle level bus drive interface configured to place a throttle level value onto the throttle level bus, and one or more throttle level bus receiver chiplets operatively coupled to the throttle level bus. Each chiplet of the multiple chiplets includes throttling logic circuitry configured to set a throttle level of a chiplet according to the throttle level value.

Abstract: Devices and techniques for thread-based processor halting are described herein. A processor monitors control-status register (CSR) values that correspond to a halt condition for a thread. The processor then compares the halt condition to a current state of the thread and halts in response to the current state of the thread meeting the halt condition.

Abstract: A system comprises an interposer including interconnect and multiple chiplets arranged on the interposer. Each chiplet includes multiple chiplet input-output (I/O) channels interconnected to I/O channels of other chiplets by the interposer; a chiplet I/O interface for the chiplet I/O channels that includes multiple interface layers; and initialization logic circuitry configured to advance initialization of the chiplet interface sequentially through the interface layers starting with a lowest interface layer.

Abstract: Devices and techniques for short-thread rescheduling in a processor are described herein. When an instruction for a thread completes, a result is produced. The condition that the same thread is scheduled in a next execution slot and that the next instruction of the thread will use the result can be detected. In response to this condition, the result can be provided directly to an execution unit for the next instruction.

Abstract: A chiplet system can include a Serial Peripheral Interface (SPI) bus for communication. A controller or primary device coupled to the SPI bus can generate a message with read or write instructions for one or more secondary devices. Secondary devices on the SPI bus can be configured to include or use respective static identifiers that uniquely identify or address each device. In an example, the primary device can communicate messages using the SPI bus, and the messages can include or use a device identification field. In an example, secondary devices on the SPI bus can be configured to monitor the device identification fields of incoming messages. If a message includes an identification field that corresponds to an identifier of a particular device, then the particular device can attend to the message, and other devices without the same identifier can disregard the message.

Abstract: Devices and techniques for thread execution control in a barrel processor are described herein. An apparatus includes a barrel processor, which includes local memory including a hazard data structure; and thread scheduling circuitry; wherein the barrel processor is configured to perform operations through use of the thread scheduling circuitry, the operations including: identifying an instruction to place into a pipeline for the barrel processor, the instruction corresponding to a thread; reading a hazard indication entry from a hazard data structure, the hazard indication entry corresponding to the thread, and wherein the hazard indication entry is set by a preceding instruction in the thread; and in response to reading the hazard indication entry, rescheduling the thread to a later time based on the hazard identification.

Abstract: A system comprises an interposer including interconnect and multiple chiplets arranged on the interposer. Each chiplet includes multiple chiplet input-output (I/O) channels interconnected to I/O channels of other chiplets by the interposer; a chiplet I/O interface for the chiplet I/O channels that includes multiple interface layers; and initialization logic circuitry configured to advance initialization of the chiplet interface sequentially through the interface layers starting with a lowest interface layer.

Abstract: A system comprises an interposer including multiple conductive interconnects; multiple chiplets arranged on the interposer and interconnected by the interposer; each chiplet including a die-to-die physical layer interface including one or more pads to engage the interconnect of the interposer; and wherein at least one chiplet includes multiple input-output channels organized into at least one column and arranged in an order at a periphery of the chiplet forming a die-to-die physical layer interface to engage the interconnects of the interposer, wherein the order of the channels of the column is programmable.

Abstract: Systems and methods which provide a modular processor framework and instruction set architecture designed to efficiently execute applications whose memory access patterns are irregular or non-unit stride are disclosed. A hybrid multithreading framework (HMTF) of embodiments provides a framework for constructing tightly coupled, chip-multithreading (CMT) processors that contain specific features well-suited to hiding latency to main memory and executing highly concurrent applications. The HMTF of embodiments includes an instruction set designed specifically to exploit the high degree of parallelism and concurrency control mechanisms present in the HMTF hardware modules. The instruction format implemented by a HMTF of embodiments is designed to give the architecture, the runtime libraries, and/or the application ultimate control over how and when concurrency between thread cache units is initiated.

Abstract: Systems and methods which provide a modular processor framework and instruction set architecture designed to efficiently execute applications whose memory access patterns are irregular or non-unit stride as disclosed. A hybrid multithreading framework (HMTF) of embodiments provides a framework for constructing tightly coupled, chip-multithreading (CMT) processors that contain specific features well-suited to hiding latency to main memory and executing highly concurrent applications. The HMTF of embodiments includes an instruction set designed specifically to exploit the high degree of parallelism and concurrency control mechanisms present in the HMTF hardware modules. The instruction format implemented by a HMTF of embodiments is designed to give the architecture, the runtime libraries, and/or the application ultimate control over how and when concurrency between thread cache units is initiated.

Abstract: Instead of alternatively utilizing only one fabric or the other fabric of a redundant pair, both fabrics simultaneously transmit duplicate information, such that each packet forwarding module (PFM) receives the output of both fabrics simultaneously. In real time, an internal optics module (IOM) analyzes each information chunk coming out of a working zero switch fabric; simultaneously examines a parallel output of a working one duplicate switch fabric; and compares on a chunk-by-chunk basis the validity of each and every chunk from both switch fabrics. The IOM does this by examining forward error correction (FEC) check symbols encapsulated into each chunk. FEC check symbols allow correcting a predetermined number of bit errors within a chunk. If the chunk cannot be corrected, then the IOM provides indication to all PFMs downstream that the chunk is defective. Under such conditions, the PFMs select a chunk from the non-defective switch fabric.

Abstract: A chunk format for a large-scale, high data throughput router includes a preamble that allows each individual chunk to have clock and data recovery performed before the chunk data is retrieved. The format includes a chunk header that contains information specific to the entire chunk. A chunk according to the present format can contain multiple packet segments, with each segment having its own packet header for packet-specific information. The format provides for a scrambler seed which allows scrambling the data to achieve a favorable zero and one balance as well as minimal run lengths. There can be a random choice of available scrambler seeds for any particular chunk to avoid malicious forcing of zero and one patterns or run lengths of bit zeroes and ones. There are a chunk cyclical redundancy check (CRC) as well as forward error correction (FEC) bytes to detect and/or correct any errors and also to insure a high degree of data and control integrity.

Abstract: A chunk format for a large-scale, high data throughput router includes a preamble that allows each individual chunk to have clock and data recovery performed before the chunk data is retrieved. The format includes a chunk header that contains information specific to the entire chunk. A chunk according to the present format can contain multiple packet segments, with each segment having its own packet header for packet-specific information. The format provides for a scrambler seed which allows scrambling the data to achieve a favorable zero and one balance as well as minimal run lengths. There are forward error correction (FEC) bytes as well as a chunk cyclical redundancy check (CRC) to detect and/or correct any errors and also to insure a high degree of data and control integrity. Advantageously, a framing symbol inserted into the chunk format itself allows the receiving circuitry to identify or locate a particular chunk format.