A computing device includes processing units and a shared processor cache. Each cache line is associated with a different plurality of accounting bits, including a unit bit associated with each processing unit. An operation by a particular processing unit on a particular cache line is identified. If the operation is a read consumed by that processing unit, and when a unit bit for that processing unit in the accounting bits associated with the particular cache line is not set, at least the value portion of the particular cache line is stored or referenced in the trace, and the particular unit bit is set. If the operation is a write, each unit bit in the accounting bits associated with the cache line that are associated with any processing unit other than the particular processing unit is cleared, and the particular unit bit associated with the particular processing unit is set.

A method for compressing data in a local cache of a web server is described. A local cache compression engine accesses values in the local cache and determines a cardinality of the values of the local cache. The local cache compression engine determines a compression rate of a compression algorithm based on the cardinality of the values of the local cache. The compression algorithm is applied to the cache based on the compression rate to generate a compressed local cache.

A transaction is detected. The transaction has a begin-transaction indication and an end-transaction indication. If it is determined that the begin-transaction indication is not a no-speculation indication, then the transaction is processed.

Disclosed in the present disclosure is a data caching method and apparatus, the data caching method includes the following steps: receiving a data request message sent by a user terminal; if detecting that a cache apparatus does not include the target access data requested by the data request message, then sending the target access data in a storage apparatus to the user terminal; extracting parameter information of the target access data in the storage apparatus, and determining whether the parameter information matches a preset parameter condition; and, if the parameter information matches the preset parameter condition, then transmitting the target access data to the cache apparatus.

A computer-implemented method is presented for caching content. The method includes storing content received from a data stream, employing a cache array and a retrieval data structure to process the content, determining, by the processor, if the content is popular content which appears frequently, a position in the cache array holding a strong reference to the content, determining, by the processor, if the content is unpopular content which appears infrequently, an arbitrary position in the cache array holding a strong reference to the unpopular content, the retrieval data structure holding a weak reference to the content, and issuing cached content to a cache data storage device accessed by a user interface of a computing device communicating with the data processing system.

Described herein is a system and method for performing a clustered coherent cloud read cache without coherency messaging. At a cloud reader agent having a read cache, a request for a particular portion of data is received from a client. The request includes an identifier of the node making the request, a current time on the node making the request, a last stubbed time of the particular portion of data, and/or an identifier of the particular portion of data. When the particular portion of data is stored in the read cache of the cloud reader agent, the cloud reader agent can determine whether the cached particular portion of data is valid based, at least in part, upon the received identifier of the node making the request, the received last stubbed time of the particular portion of data, and/or, a stored time associated with the cached particular portion of data.

A content delivery network may provide content items to requesting devices using a popularity-based distribution hierarchy. A central analysis system may determine popularity data for a content item stored in a first caching device. The central analysis system may determine that a change in the popularity data is beyond a threshold value. The central analysis system may then transmit an instruction to move the content item from the first caching device to a second caching device in a different tier of caching devices than the first caching device. The central analysis system may update a content index to indicate that the content item has been moved to the second caching device. A user device may be redirected to request the content item directly from the second caching device.

In an embodiment, a processor includes a plurality of cores to independently execute instructions, a shared cache coupled to the cores and including a plurality of lines to store data, and a power controller including a low power control logic to calculate a flush latency to flush the shared cache based on a state of the plurality of lines. Other embodiments are described and claimed.

Technologies for predictive caching include a computing device to receive sensor data generated by one or more sensors of the computing device and determine a device context of the computing device based on the sensor data. Based on the device context, the computing device determines a file to cache that has similar characteristics to another file recently accessed by a user of the computing device. The computing device includes a file cache with a first partition to store files identified to have similar characteristics to files recently accessed by a user and a second partition to store files identified based on access patterns of the user. The computing device stores the determined file to the first partition.

Combined operational steps and device characteristics help preserve data against integrity threats. Data is divided into critical data and non-critical data, based on criteria such as customer requirements, workload criticality, or virtual machine criticality. Data may be generated in a compute node for storage in a storage node, for example. Critical data is stored in a battery-backed memory aperture at physical addresses where it will be flushed ahead of the non-critical data due to a flush order imposed by or on the battery-backed memory, e.g., a bottom-up NVDIMM flush order. Redundant copies of the data (especially non-critical data) may also be kept in case it does not get flushed in time. Battery-backed memory apertures are sized and located according to their battery's characteristics, and may be relocated or resized as conditions change. Flush defragging is performed to optimize use of the aperture, especially within the portion that holds critical data.