Abstract: Techniques to recover data from an indirected non-volatile memory system after unexpected power failure, as, e.g., NAND memory in electronic devices are disclosed.

Abstract: A method and system to allow power fail-safe write-back or write-through caching of data in a persistent storage device into one or more cache lines of a caching device. No metadata associated with any of the cache lines is written atomically into the caching device when the data in the storage device is cached. As such, specialized cache hardware to allow atomic writing of metadata during the caching of data is not required.

Abstract: A method and system to perform caching based at least on one or more file-level heuristics. The caching of a storage medium in a caching device is performed by a cache policy engine. The cache policy engine receives file-level information of input/output access of data of the storage medium and caches or evicts the data of the storage medium in the caching device based on the received file-level information. By utilizing information about the files and file operations associated with the disk sectors or logical block addresses of the storage medium, the cache policy engine can make a better decision on the data selection of the storage medium to be cached in or evicted from the caching device in one embodiment of the invention. Higher cache hit rates can be achieved and the performance of the system utilizing the cache policy engine is improved.

Abstract: A technique includes performing a plurality of write operations to store data in different physical memory locations. Each of the physical memory locations are associated with a logical address that is shared in common among the physical addresses. The technique includes storing sequence information in the physical memory locations to indicate which one of the write operations occurred last.

Abstract: A method is provided. The method includes receiving data and classifying received data in one of several tiers of data. The method also includes storing each tier of data on a different non-volatile memory device.

Abstract: Embodiments of methods to securely bind a disk cache encryption key to a cache device are generally described herein. Other embodiments may be described and claimed.

Abstract: A non-volatile memory, such as a NAND memory, may be encrypted by reading source blocks, writing to destination blocks, and then erasing the source blocks. As part of the encryption sequence, a power fail recovery procedure, using sequence numbers, is used to reestablish a logical-to-physical translation table for the destination blocks.

Abstract: Techniques to manage various errors in memory such as, e.g., NAND memory in electronic devices are disclosed. In some embodiments, erase, read, and program error handling errors are managed.

Abstract: In one embodiment, the present invention includes a method for maintaining a sequence of writes into a disk cache, where the writes correspond to disk write requests stored in the disk cache, and ordering cache writes from the disk cache to a disk drive according to the sequence of writes. In this way, write ordering from an operating system to a disk subsystem is maintained. Other embodiments are described and claimed.

Abstract: A volatile or nonvolatile cache memory can cache mass storage device read data and write data. The cache memory may become inaccessible, and I/O operations may go directly to the mass storage device, bypassing the cache memory. A log of write operations may be maintained to update the cache memory when it becomes available.

Abstract: Incrementing sequence numbers in the metadata of non-volatile memory is used in the event of a resume from power fail to determine which data in the memory is current and valid, and which data is not. To reduce the amount of metadata space consumed by these sequence numbers, the numbers are permitted to be small enough to wrap around when the maximum value is reached. Two different techniques are disclosed to keep this wrap around condition from causing ambiguity in the relative values of the sequence numbers.

Abstract: Techniques to recover data from an indirected non-volatile memory system after unexpected power failure, as, e.g., NAND memory in electronic devices are disclosed.

Abstract: A method is provided for reducing the number of writes in a non-volatile memory (122). The method involves writing data in the non-volatile memory and determining a set of data from the data in the non-volatile memory to be written to a removable memory (126) that is operatively coupled to the non-volatile memory (e.g., a NAND memory). The method also involves writing the set of data to the removable memory (e.g., a hard disk) from the non-volatile memory. The method further involves writing a delineation marker (e.g., a sequence number) to the non-volatile memory specifying that the set of data has been written to the removable memory. Notably, the metadata of the data in the non-volatile memory comprises at least one marker set as a specific marker type (e.g., a valid marker and a dirty marker).

Abstract: Write operations store data in different physical memory locations. Each of the physical memory locations are associated with a logical address that is shared in common among the physical addresses. Sequence information stored in the physical memory location indicates which one of the write operations occurred last. The available erased memory location can be split into a list of erased memory locations available to be used and a list of erased memory locations not available to be used. Then, on a failure, only the list of erased memory locations available to be used needs to be analyzed to reconstruct the consumption states of memory locations.

Abstract: Disclosed is a method for reducing number of writes in a write-back non-volatile cache memory. The method comprises: writing a plurality of data in the cache memory, wherein cache lines meta data for each of the plurality of data is marked as dirty; determining a set of data of the plurality of the data in the cache memory to be flushed to a hard disk, wherein the hard disk is operatively coupled to the cache memory; flushing the set of data of the plurality of data to the hard disk from the cache memory; and writing a clean-marker to the cache memory specifying which of the plurality of the data has been flushed to the disk.

Abstract: A volatile or nonvolatile cache memory can cache mass storage device read data and write data. The cache memory may become inaccessible, and I/O operations may go directly to the mass storage device, bypassing the cache memory. A log of write operations may be maintained to update the cache memory when it becomes available.

Abstract: In one embodiment, the present invention includes a method for writing data to a disk if inserting the data into a cache, such as a disk cache associated with the disk, would cause a threshold of dirty data in the cache to be met or exceeded. Further, in certain embodiments, the cache may store data according to a first cache policy and a second cache policy. A determination of whether to store data according to the first or second policies may be dependent upon an amount of dirty data in the cache, in certain embodiments. In certain embodiments, the cache may include at least one portion reserved for clean data.

Abstract: Updating a spatial partitioning data structure during run-time in an efficient manner includes several pre-processing steps. Pre-processing includes generating a first spatial partitioning data structure for a model at a first resolution, generating a second spatial partitioning data structure for the model at a second resolution, analyzing the first and second spatial partitioning data structures to identify differences between spatial partitioning of the model at the first and second resolutions, and storing the differences in a spatial partitioning update data structure. This pre-processing may be repeated for one or more pairs of successive resolutions of the model. Subsequently, during run-time, the model's resolution may be changed from the first resolution to the second resolution. In response, a spatial partitioning data structure corresponding to the first resolution may be updated using the spatial partitioning update data structure to reflect the change in resolution of the model.

Abstract: Updating a spatial partitioning data structure during run-time in an efficient manner includes several pre-processing steps. Pre-processing includes generating a first spatial partitioning data structure for a model at a first resolution, generating a second spatial partitioning data structure for the model at a second resolution, analyzing the first and second spatial partitioning data structures to identify differences between spatial partitioning of the model at the first and second resolutions, and storing the differences in a spatial partitioning update data structure. This pre-processing may be repeated for one or more pairs of successive resolutions of the model. Subsequently, during run-time, the model's resolution may be changed from the first resolution to the second resolution. In response, a spatial partitioning data structure corresponding to the first resolution may be updated using the spatial partitioning update data structure to reflect the change in resolution of the model.

Abstract: A method and apparatus for generating stereoscopic displays in a computer system. Each frame in a sequence of frames includes a left image and a right image, and each image includes a plurality of pixels. Depth information for objects depicted in the display is stored in a z buffer. Either the left image or the right image is computed as an approximation of the other using the depth information stored in the z buffer. The approximated image is alternated between the left and the right image on a frame-by-frame basis, so that the left and right image are each approximated every other frame. Pixels which are not filled in the approximated image are assigned values based on the corresponding pixels in the same (non-approximated) image from the preceding frame.