Abstract: The technology generally relates to high bandwidth memory (HBM) and optical connectivity stacking. Disclosed systems and methods herein allow for 3D-stacking of HBM dies that are interconnected with an optical interface in a manner that allows for compact, high-performance computing. An optical chiplet can be configured to be placed onto a stack of HBM dies, with a cooling die that is positioned between the HBM dies and the optical chiplet. The optical chiplet may be configured to connect the HBM optics module package to one or more other components of the package via to one or more optical fibers.

Abstract: A microelectronic system may include a substrate having a first surface, one or more interposers mounted to and electrically connected to the first surface, first and second application specific integrated circuits (ASICs) each at least partially overlying and electrically connected to one of the interposers, a plurality of high-bandwidth memory elements (HBMs) each at least partially overlying and electrically connected to one of the interposers, and an active silicon bridge mounted to and electrically connected to the first surface and providing an electrical connection between the first and second ASICs, the active silicon bridge having active microelectronic devices therein. The microelectronic system may be configured such that the first and second ASICs and the active silicon bridge each have a purely digital CMOS interface therein. A plurality of bumps providing the electrical connection between the ASICs and the active silicon bridge may be configured to receive serial data therethrough.

Abstract: In an embodiment, a memory system may include at least two types of DRAM, which differ in at least one characteristic. For example, one DRAM type may be a high density DRAM, while another DRAM type may have lower density but may also have lower latency and higher bandwidth than the first DRAM type. DRAM of the first type may be on one or more first integrated circuits and DRAM of the second type may be on one or more second integrated circuits. In an embodiment, the first and second integrated circuits may be coupled together in a stack. The second integrated circuit may include a physical layer circuit to couple to other circuitry (e.g., an integrated circuit having a memory controller, such as a system on a chip (SOC)), and the physical layer circuit may be shared by the DRAM in the first integrated circuits.

Abstract: The technology generally relates to disaggregating memory from an application specific integrated circuit (“ASIC”) package. For example, a high-bandwidth memory (“HBM”) optics module package may be connected to an ASIC package via one or more optical links. The HBM optics module package may include HBM dies(s), HBM chiplet(s) and an optical chiplet. The optical chiplet may be configured to connect the HBM optics module to one or more optical fibers that form an optical link with one or more other components of the ASIC package. By including an optical chiplet in the HBM optics module package, the HBM optics module package may be disaggregated from an ASIC package.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing data on a memory controller. One of the methods comprises obtaining a first request and a second request to access respective data corresponding to the first and second requests at a first memory device of the plurality of memory devices; and initiating interleaved processing of the respective data; receiving an indication to stop processing requests to access data at the first memory device and to initiate processing requests to access data at a second memory device, determining that the respective data corresponding to the first and second requests have not yet been fully processed at the time of receiving the indication, and in response, storing, in memory accessible to the memory controller, data corresponding to the requests which have not yet been fully processed.

Abstract: In an embodiment, a memory system may include at least two types of DRAM, which differ in at least one characteristic. For example, one DRAM type may be a high density DRAM, while another DRAM type may have lower density but may also have lower latency and higher bandwidth than the first DRAM type. DRAM of the first type may be on one or more first integrated circuits and DRAM of the second type may be on one or more second integrated circuits. In an embodiment, the first and second integrated circuits may be coupled together in a stack. The second integrated circuit may include a physical layer circuit to couple to other circuitry (e.g., an integrated circuit having a memory controller, such as a system on a chip (SOC)), and the physical layer circuit may be shared by the DRAM in the first integrated circuits.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing data on a memory controller. One of the methods comprises obtaining a first request and a second request to access respective data corresponding to the first and second requests at a first memory device of the plurality of memory devices; and initiating interleaved processing of the respective data; receiving an indication to stop processing requests to access data at the first memory device and to initiate processing requests to access data at a second memory device, determining that the respective data corresponding to the first and second requests have not yet been fully processed at the time of receiving the indication, and in response, storing, in memory accessible to the memory controller, data corresponding to the requests which have not yet been fully processed.

Abstract: In an embodiment, a memory system may include at least two types of DRAM, which differ in at least one characteristic. For example, one DRAM type may be a high density DRAM, while another DRAM type may have lower density but may also have lower latency and higher bandwidth than the first DRAM type. DRAM of the first type may be on one or more first integrated circuits and DRAM of the second type may be on one or more second integrated circuits. In an embodiment, the first and second integrated circuits may be coupled together in a stack. The second integrated circuit may include a physical layer circuit to couple to other circuitry (e.g. an integrated circuit having a memory controller, such as a system on a chip (SOC)), and the physical layer circuit may be shared by the DRAM in the first integrated circuits.

Abstract: An apparatus includes a memory circuit and a memory controller circuit. The memory controller circuit may include a write request queue. The memory controller circuit may be configured to receive a memory request to access the memory circuit and determine if the memory request includes a read request or a write request. A received read request may be scheduled for execution, while a received write request may be stored in the write request queue. The memory controller circuit may reorder scheduled memory requests based on achieving a specified memory access efficiency and based on a number of write requests stored in the write request queue.

Abstract: Techniques relating to arbitration in a memory controller are disclosed. In some embodiments, the memory controller is configured to transition between read turns and writes turn according to a turn schedule. In some embodiments, the memory controller also receives reports from circuitry requesting memory transactions and determines a current latency tolerance value based on the reports. In some embodiments, the memory controller is configured to switch from a write turn to a read turn prior to a scheduled switch based on the current latency tolerance meeting a threshold value.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing data on a memory controller. One of the methods comprises obtaining a first request and a second request to access respective data corresponding to the first and second requests at a first memory device of the plurality of memory devices; and initiating interleaved processing of the respective data; receiving an indication to stop processing requests to access data at the first memory device and to initiate processing requests to access data at a second memory device, determining that the respective data corresponding to the first and second requests have not yet been fully processed at the time of receiving the indication, and in response, storing, in memory accessible to the memory controller, data corresponding to the requests which have not yet been fully processed.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing data on a memory controller. One of the methods comprises obtaining a first request and a second request to access respective data corresponding to the first and second requests at a first memory device of the plurality of memory devices; and initiating interleaved processing of the respective data; receiving an indication to stop processing requests to access data at the first memory device and to initiate processing requests to access data at a second memory device, determining that the respective data corresponding to the first and second requests have not yet been fully processed at the time of receiving the indication, and in response, storing, in memory accessible to the memory controller, data corresponding to the requests which have not yet been fully processed.

Abstract: An embodiment of an apparatus includes a plurality of processing circuits, a plurality of memory circuits, and a memory controller circuit coupled to each memory circuit via a respective communication channel. A particular processing circuit may generate a data stream that includes a plurality of data blocks. The memory controller circuit may receive the plurality of data blocks from the particular processing circuit. The memory controller circuit may distribute the plurality of data blocks among the plurality of memory circuits based on respective utilizations of the plurality of communication channels.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing data on a memory controller. One of the methods comprises obtaining a first request and a second request to access respective data corresponding to the first and second requests at a first memory device of the plurality of memory devices; and initiating interleaved processing of the respective data; receiving an indication to stop processing requests to access data at the first memory device and to initiate processing requests to access data at a second memory device, determining that the respective data corresponding to the first and second requests have not yet been fully processed at the time of receiving the indication, and in response, storing, in memory accessible to the memory controller, data corresponding to the requests which have not yet been fully processed.

Abstract: In an embodiment, a memory system may include at least two types of DRAM, which differ in at least one characteristic. For example, one DRAM type may be a high density DRAM, while another DRAM type may have lower density but may also have lower latency and higher bandwidth than the first DRAM type. DRAM of the first type may be on one or more first integrated circuits and DRAM of the second type may be on one or more second integrated circuits. In an embodiment, the first and second integrated circuits may be coupled together in a stack. The second integrated circuit may include a physical layer circuit to couple to other circuitry (e.g. an integrated circuit having a memory controller, such as a system on a chip (SOC)), and the physical layer circuit may be shared by the DRAM in the first integrated circuits.

Abstract: In one embodiment, a system includes a memory that includes a live section and a spares section. The live section may be mapped to the address space of the system, and may be accessed in response to memory operations. Once an entry in the live section has been detected as failed, an entry in the spares section may be allocated to replace the failed entry. During subsequent accesses to the failed entry, the allocated entry may be used instead. In an embodiment, the failed entry may be coded with an indication of the allocated entry, to redirect to the allocated entry. In one implementation, for example, the failed entry may be coded with N copies of a pointer to the allocated entry, each copy protected by corresponding ECC.

Abstract: In an embodiment, a memory system may include at least two types of DRAM, which differ in at least one characteristic. For example, one DRAM type may be a high density DRAM, while another DRAM type may have lower density but may also have lower latency and higher bandwidth than the first DRAM type. DRAM of the first type may be on one or more first integrated circuits and DRAM of the second type may be on one or more second integrated circuits. In an embodiment, the first and second integrated circuits may be coupled together in a stack. The second integrated circuit may include a physical layer circuit to couple to other circuitry (e.g. an integrated circuit having a memory controller, such as a system on a chip (SOC)), and the physical layer circuit may be shared by the DRAM in the first integrated circuits.

Abstract: A memory controller circuit coupled to a memory circuit that includes multiple banks may receive multiple access requests including a particular access request to a particular bank of the plurality of banks. The particular access request is associated with a particular virtual channel of a plurality of virtual channels. The memory controller circuit may select a given access requests of the multiple access requests based on an arbitration category value associated with a virtual channel of the given access request and modify the arbitration category value in response to selecting the given access request.

Abstract: A method and apparatus for optimizing calibrations of a memory subsystem is disclosed. A memory controller of a memory subsystem includes a memory interface suitable for coupling to a DRAM having a plurality of banks. The memory controller includes a state machine the state machine may initiate calibration of circuitry within the memory controller. Responsive to initiating the calibration, the state machine also causes a refresh command to be transmitted to the DRAM. The calibration is then performed concurrent with the refresh of the DRAM. Subsequent to transmitting the refresh command, the state machine causes the memory interface to be placed into a low power state.

Abstract: A memory controller circuit coupled to multiple memory circuits may receive requests to access particular locations within the multiple memory circuits. A request may be assigned a particular quality-of-service level. During operation, the memory controller circuit may reallocate the quality-of-service level of a particular request to a new quality-of-service level based on accumulated bandwidth credits associated with the new quality-of-service level.