摘要: We explore using hashing to pack sparse data into a compact table while retaining efficient random access. Specifically, we design perfect multidimensional hash function -- one that is precomputed on static have no collisions. Because our makes single reference small offset table, queries always involve exactly two memory accesses and are thus ideally suited for parallel SIMD evaluation graphics hardware. Whereas prior work strives pseudorandom mappings, instead the preserve spatial coherence thereby improve runtime locality of reference. demonstrate numerous applications including vector images, texture sprites, alpha channel compression, 3D-parameterized textures, 3D painting, simulation, collision detection.