A cache memory apparatus including a cache memory including a bank, a partition configuration unit configured to divide the cache memory into partitions by allocating the bank to a texture among textures for rendering, and a controller configured to receive a partition ID, of texture data requested by a device that performs the rendering, determine whether the requested texture data is stored in a partition corresponding to the partition ID among the plurality of partitions, and output the requested texture data to the device based on a result of the determination.

A virtual-address cache module receives at least a portion of a virtual address and in response indicates a hit or a miss. A first cache structure stores only memory blocks with virtual addresses that are members of a set of multiple synonym virtual addresses that have all been previously received by the virtual-address cache module during the operating period, where each member of a particular set of multiple synonym virtual addresses translates to a common physical address, and a memory block with the common physical address is stored in at most a single storage location within the first cache structure. A second cache structure stores only memory blocks with virtual addresses that do not have any synonym virtual addresses that have been previously received by the virtual-address cache during the operating period.

A processor includes a first prefetcher that prefetches data in response to memory accesses and a second prefetcher that prefetches data in response to memory accesses. Each of the memory accesses has an associated memory access type (MAT) of a plurality of predetermined MATs. The processor also includes a table that holds first scores that indicate effectiveness of the first prefetcher to prefetch data with respect to the plurality of predetermined MATs and second scores that indicate effectiveness of the second prefetcher to prefetch data with respect to the plurality of predetermined MATs. The first and second prefetchers selectively defer to one another with respect to data prefetches based on their relative scores in the table and the associated MATs of the memory accesses.

A system and method for distributing data over a plurality of remote storage nodes. Data are split into segments and each segment is encoded into a number of codeword chunks. None of the codeword chunks contains any of the segments. Each codeword chunk is packaged with at least one encoding parameter and identifier, and metadata are generated for at least one file and for related segments of the at least one file. The metadata contains information to reconstruct from the segments, and information for reconstructing from corresponding packages. Further, metadata are encoded into package(s), and correspond to a respective security level and a protection against storage node failure. A plurality of packages are assigned to remote storage nodes to optimize workload distribution. Each package is transmitted to at least one respective storage node as a function iteratively accessing and retrieving the packages of metadata and file data.

Described are techniques for destaging data. Write data for write operations are stored in cache page(s). The cache may be partitioned into cache pages and write data stored thereon may be marked as write pending denoting that write data needs to be written out to physical storage. Weight values may be determined for the cache pages in accordance with factors. A first cache page may be selected having a highest such weight value where the selected first cache page includes at least some write data marked as write pending. The write data of the selected first cache page may be destaged to physical storage.

Embodiments of the invention relate to tracking processor transactional read and write sets, thereby eliminating speculative mispredictions. Both non-speculative read set and write set indications are maintained for a transaction. The indications are stored in cache. In addition, load and write queues of addresses are maintained. The load queue of addresses relates to speculative members of a read set and the write queue of addresses relates to speculating member of a write set. For a received read request, a transaction resolution process takes place, and a resolution is performed if an address match in the write queue is detected. Similarly, for a receive write request the transaction interference additionally checks the load queue and the non-speculative read set for the pending address.

Secure memory repartitioning technologies are described. Embodiments of the disclosure may include a processing device including a processor core and a memory controller coupled between the processor core and a memory device. The memory device includes a memory range including a section of convertible pages that are convertible to secure pages or non-secure pages. The processor core is to receive a non-secure access request to a page in the memory device, responsive to a determination, based on one or more secure state bits in one or more secure state bit arrays, that the page is a secure page, insert an abort page address into a translation lookaside buffer, and responsive to a determination, based on the one or more secure state bits in the one or more secure state bit arrays, that the page is a non-secure page, insert the page into the translation lookaside buffer.

Pre-decompressing a compressed form of data that has been pre-fetched into a cache to facilitate subsequent retrieval of a decompressed form of the data from the cache is presented herein. A system retrieves, from a first portion of a cache, a compression chunk comprising compressed data blocks representing a compressed form of a group of data blocks in response to a first cache hit from the first portion of the cache being incurred, decompresses the compression chunk to obtain a decompressed chunk comprising uncompressed data blocks representing an uncompressed form of the group of data blocks, and inserts the uncompressed data blocks into a second portion of the cache. Further, the system retrieves, from the second portion of the cache, an uncompressed data block of the uncompressed data blocks in response to a second cache hit from the second portion of the cache being incurred.

A streaming engine employed in a digital data processor specifies a fixed read only data stream defined by plural nested loops. An address generator produces address of data elements. A steam head register stores data elements next to be supplied to functional units for use as operands. An element duplication unit optionally duplicates data element an instruction specified number of times. A vector masking unit limits data elements received from the element duplication unit to least significant bits within an instruction specified vector length. If the vector length is less than a stream head register size, the vector masking unit stores all 0's in excess lanes of the stream head register (group duplication disabled) or stores duplicate copies of the least significant bits in excess lanes of the stream head register.

Embodiments of a multitenant content server that employs embodiments of a database architecture for use in multitenant environments that includes a global partition and a tenant partition for each tenant. These partitions can be accessed using sessions corresponding to the partitions and interfaces associated with the tenants.