NAND flash memory has become the dominant storage media choice in a vast majority of application scenarios. Compared to mechanical hard disks, flash offers better access performance, energy efficiency, and shock resistance. However, the unique hardware peculiarities of this technology require dedicated facilities to manage the flash space and data. The implementation of flash management facilities has alternatively been realized either at the device or host computer level. Managing flash on the device side eases integration/compatibility and increases performance in certain scenarios. However, the limited computing resources inherent to devices and the lack of higher-level file system/application information make these solutions suboptimal in many situations. Managing flash on the host allows leveraging its abundant resources, and host-side knowledge such as data access patterns can be exploited to optimize flash management, at the cost of increased host-side complexity. The pros and cons of each approach also led to the appearance of hybrid, cross-layer solutions, enabling the collaboration of different layers of the storage stack. Recently, the pressure on modern storage systems requires that an increasing amount of flash management responsibilities is offloaded to the host, and the development of application-specific cross-layer solutions: In that context, it is crucial to review these developments. In this article, we make a comprehensive survey of the host-side management technologies of flash memory, application-/system-level flash-friendly designs, and emergent applications based on flash memory.

For many emerging data-centric computing applications, it is a key capability to efficiently store, manage, and access multidimensional data. To achieve this, many multidimensional index data structures have been proposed. However, when existing multidimensional index data structures are maintained on modern nonvolatile memories (NVMs), such as NAND flash memory, they often face challenges in effective management of multidimensional data and handling of memory medium peculiarities, such as the write-once property and the need for block reclamation of NAND flash memory. Without appropriate management, these challenges often result in serious amplification of the read/write traffic, which degrades the performance of multidimensional data structures. Motivated by the urgent needs of efficient multidimensional index data structures on modern NVMs, we propose the FIRM-tree, a time-efficient and space-economic index data structure for multidimensional point data on NAND flash memory. Unique to the prior work, the FIRM-tree holistically utilizes RAM and flash memory space, and dedicatedly leverages the page reprogrammability of modern NAND flash memory, to enhance data access performance and flash management overheads. We then verify our proposal through analytical and experimental studies, where the results are quite encouraging.

Recently, the value of data has been widely recognized, which highlights the significance of data-centric computing in diversified application scenarios. In many cases, the data are multidimensional, and the management of multidimensional data often confronts greater challenges in supporting efficient data access operations and guaranteeing the space utilization. On the other hand, while many existing index data structures have been proposed for multidimensional data management, however, their designs are not fully optimized for modern nonvolatile memories, in particular the byte-addressable persistent memories. As a result, they might undergo serious access performance degradation or fail to guarantee space utilization. This observation motivates the redesigning of index data structures for multidimensional point data on modern persistent memories, such as the phase-change memory. In this work, we present the WARM-tree , a m ultidimensional t ree for r educing the w rite a mplification effect, for multidimensional point data. In our evaluation studies, as compared to the bucket PR quadtree and R*-tree, the WARM-tree can provide any worst-case space utilization guarantees in the form of \(\frac{m-1}{m}\) ( m ∈ ℤ^+) and effectively reduces the write traffic of key insertions by up to 48.10% and 85.86%, respectively, at the price of degraded average space utilization and prolonged latency of query operations. This suggests that the WARM-tree is a potential multidimensional index structure for insert-intensive workloads.

Curvilinear reformatting of 3D magnetic resonance imaging data has been recognized by the medical community as a helpful noninvasive tool for displaying the cerebral anatomy. It consists of automatically creating, with respect to a reference surface, a series of equidistant curvilinear slices at progressively deeper cuts. In comparison with planar slices, it allows more precise localization of lesions and identification of subtle structural abnormalities. However, current curvilinear reformatting tools either rely on the time-consuming manual delineation of guiding curves on 2D slices, or require costly automatic brain segmentation procedures. In addition, they extract the skin and skull, impeding a precise topographic correlation between the location of the brain lesion and skin surface. This impairs planning of craniotomy for neurosurgery, and of the appropriate implantation of electrodes for intracranial electroencephalography in presurgical evaluation. In this work, we present a novel approach based on direct manipulation of the visualized volume data. By using a 3D painting metaphor, the reference surface can be defined incrementally, according to the principle that the user interacts with what she/he sees. As a response, an animation of the reformatting process is displayed. The focus of this paper is a new volume tagging algorithm behind user interactions. It works at an interactive frame rate on current graphics hardware.

We present a simple yet effective snapping technique for constraining the motion of the cursor of an input device to the surface of 3D models whose geometry is arbitrarily deformed by a programmable hardware fragment and vertex processor. The technique works in image space and thus snaps the cursor to the geometry actually rendered instead of the geometry originally submitted to the rendering pipeline. We also present a method to establish a correspondence between snapped geometry in image space and object space, and an efficiency improvement based on the control of frequency of frame buffer accesses. Performance tests are conducted and compared against the standard picking and snapping algorithm used by the D3DX library of the Microsoft Direct3D API. We conclude by emphasizing the feasibility of our algorithm when facing the new advances of the graphics hardware for deforming geometry on the GPU.