An apparatus comprises memory access circuitry to access a memory system; a plurality of memory mapped registers, including at least an insert register and a producer pointer register; and control circuitry to perform an insert operation in response to receipt of an insert request from a requester device sharing access to the memory system. The insert request specifies an address mapped to the insert register and an indication of a payload. The insert operation includes controlling the memory access circuitry to write the payload to a location in the memory system selected based on a producer pointer value stored in the producer pointer register, and updating the producer pointer register to increment the producer pointer value.

Described apparatuses and methods track access metadata pertaining to activity within respective address ranges. The access metadata can be used to inform prefetch operations within the respective address ranges. The prefetch operations may involve deriving access patterns from access metadata covering the respective ranges. Suitable address range sizes for accurate pattern detection, however, can vary significantly from region to region of the address space based on, inter alia, workloads produced by programs utilizing the regions. Advantageously, the described apparatuses and methods can adapt the address ranges covered by the access metadata for improved prefetch performance. A data structure may be used to manage the address ranges in which access metadata are tracked. The address ranges can be adapted to improve prefetch performance through low-overhead operations implemented within the data structure. The data structure can encode hierarchical relationships that ensure the resulting address ranges are distinct.

A compression attached memory module (CAMM) includes a memory device, a first array of compression contact pads and a second array of compression contact pads. The first array of compression contact pads receives a first memory channel and a second memory channel. The first memory channel is coupled to the first memory device. The second array of compression contact pads is coupled to a subset of the first compression contact pads associated with the second memory channel.

An information handling system includes a printed circuit board, a z-axis compression connector, and a compression attached memory module (CAMM). The compression connector coupled receives a first memory channel and a second memory channel from the PCB. The CAMM receives the first memory channel and the second memory channel from the first compression connector. The CAMM is configured to provide memory transaction on only the first memory channel.

Examples include techniques associated with mapping system memory physical addresses to proximity domains. Examples include mapping system memory physical addresses for a memory coupled with a multi-die system to proximity domains that include cores of a multi-core processor and the associated level 3 (L3) cache for use by each core included in a respective proximity domain. The mapping is to facilitate cache line ownership of a cache line in an L3 cache by an input/output device or agent located on a separate die from the multi-core processor.

A refresh tracking circuit and associated methods are disclosed herein. The tracking circuit may be configured to track a primary count value and a secondary count value based on addresses associated with received commands. The primary and secondary count values may be configured to control corresponding refresh operations respectively associated with a primary address and a secondary address.

According to one aspect, a method for determining, for a memory allocation, placements in a memory area of data blocks generated by a neural network, comprises a development of an initial sequence of placements of blocks, each placement being selected from several possible placements, the initial sequence being defined as a candidate sequence, a development of at least one modified sequence of placements from a replacement of a given placement of the initial sequence by a memorized unselected placement, and, if the planned size of the memory area obtained by this modified sequence is less than that of the memory area of the candidate sequence, then this modified sequence becomes the candidate sequence, the placements of the blocks for the allocation being those of the placement sequence defined as a candidate sequence once each modified sequence has been developed.

The present memory restoration system enables a collection of computing systems to prepare inactive rewritable memory for reserve and future replacement of other memory while the other memory is active and available for access by a user of the computing system. The preparation of the reserved memory part is performed off-line in a manner that is isolated from the current user of the active memory part. Preparation of memory includes erasure of data, reconfiguration, etc. The memory restoration system allows for simple exchange of the reserved memory part, once the active memory part is returned. The previously active memory may be concurrently recycled for future reuse in this same manner to become a reserved memory. This enables the computing collection infrastructure to “swap” to what was previously the inactive memory part when a user vacates a server, speeding up the server wipe process.

In an embodiment, a system may support programmable hashing of address bits at a plurality of levels of granularity to map memory addresses to memory controllers and ultimately at least to memory devices. The hashing may be programmed to distribute pages of memory across the memory controllers, and consecutive blocks of the page may be mapped to physically distant memory controllers. In an embodiment, address bits may be dropped from each level of granularity, forming a compacted pipe address to save power within the memory controller. In an embodiment, a memory folding scheme may be employed to reduce the number of active memory devices and/or memory controllers in the system when the full complement of memory is not needed.

A method includes synthetizing a hardware description language (HDL) code into a netlist comprising a first a second and a third components. The method further includes allocating addresses to each component of the netlist. Each allocated address includes assigned addresses and unassigned addresses. An internal address space for a chip is formed based on the allocated addresses. The internal address space includes assigned addresses followed by unassigned addresses for the first component concatenated to the assigned addresses followed by unassigned addresses for the second component concatenated to the assigned addresses followed by unassigned addresses for the third component. An external address space for components outside of the chip is generated that includes only the assigned addresses of the first component concatenated to the assigned addresses of the second component concatenated to the assigned addresses of the third component. Internal addresses are translated to external addresses and vice versa.