Hierarchical Delta Research Articles

GCom: Fine-grained Compressors in Graphics Memory of Mobile GPU

Nowadays, GPUs significantly boost rendering performance. However, the high memory requirements limit their use, especially on low-end mobile platforms. Compression techniques have been widely adopted to reduce memory consumption but face two primary issues when applied to mobile GPUs: 1) low repetition ratio caused by small raw data sizes and concurrency, and 2) low locality caused by unpredictable rendering behaviors. These two limitations result in a low compression ratio when compressors are applied to low-end mobile devices. This paper introduces gCom , a fine-grained rendering compressor accelerated by GPUs. To improve the compression ratio, gCom incorporates the following innovations: 1) Unlike other compression techniques that use frames or tiles as basic processing units, gCom is the first to employ a fine-grained processing unit (i.e., the color channel), enhancing repetition amplification without increasing raw data. 2) gCom introduces two key features, hierarchical delta , and channel decorrelator , which maximize the locality of adjacent channels and reduce raw data size. 3) To maintain the original GPU throughput, gCom revolutionizes the Golomb-Rice algorithm and proposes a new compression approach, the Parallel-Oriented Golomb-Rice (POGR) algorithm, enabling parallel execution of both decompression and compression processes. The entire design of gCom utilizes only idle resources and existing commands on mobile GPUs, thus keeping purchasing costs low. To date, gCom has improved the channel locality by nearly 50%. The best compression achievement received by gCom has reached around 20%.

Redusharptor: A Tool to Simplify Developer-Written C# Unit Tests

Modern software systems are complex and locating, isolating and fixing a fault even with a failing test is tedious and time-consuming. Simplifying failing test(s) can significantly reduce the developer effort by reducing the irrelevant program entities that developers need to observe. Delta Debugging (DD) algorithm automatically reduces the failing tests. Hierarchical Delta Debugging (HDD) algorithm improves DD for hierarchical tests like source code and HTML files. Many modern implementations of these algorithms work on a generic tree-like structure and fail to consider complex structures, intricacies, and interdependence of program elements of a particular programming language. We propose a tool ReduSharptor to simplify C# tests that uses language-specific features and interdependence of C# program elements using Roslyn compiler APIs. We evaluate ReduSharptor on a set of 30 failing C# tests to demonstrate its applicability and accuracy.

The effect of hoisting on variants of Hierarchical Delta Debugging

AbstractMinimizing failing test cases is an important preprocessing step on the path of debugging. If much of a test case that triggered a bug does not contribute to the actual failure, then the time required to fix the bug can increase considerably. However, test case reduction itself can be a time‐consuming task, especially if done manually. Therefore, automated minimization techniques have been proposed, the minimizing Delta Debugging and the Hierarchical Delta Debugging (HDD) algorithms being the most well known. In this paper, we investigated the input format of HDD, searching for structures that the algorithm cannot reduce. Motivated by the findings, we have created an algorithmic framework that enabled the use of transformations other than pruning. Furthermore, with the Transformation‐based Minimization framework, we propose to extend HDD and its coarse and recursive variants with a reduction method that does not prune subtrees but replaces them with compatible subtrees further down the hierarchy, called hoisting. We have evaluated various combinations of pruning and hoisting on multiple test suites and found that hoisting can help to further reduce the size of test cases by 27% on average and by 80% as best case compared with the baseline algorithm.

Gapless-REMA100: A gapless 100-m reference elevation model of Antarctica with voids filled by multi-source DEMs

The digital elevation model (DEM) of Antarctica is a fundamental dataset for a wide range of glaciological applications, from fieldwork planning to ice sheet dynamics analysis. The DEM data at high spatial resolution provides a more detailed depiction of the topography. The reference elevation model of Antarctica (REMA) mosaic is a recently released high-resolution Antarctic DEM product with a high absolute vertical accuracy of 1 m or less. The REMA mosaic was generated using an optical photogrammetry technique and can reflect the surface elevation of the ice sheet. However, the REMA mosaic has a large number of data voids of various spatial sizes, which affects its wide glaciological application. Therefore, we propose the generation of a gapless 100-m reference elevation model of Antarctica (Gapless-REMA100) with the voids filled by combining multi-source DEMs. In this study, the best-fit fill DEMs were automatically selected between almost all available circum-Antarctica or regional DEMs. Targeting numerous voids in large areas of the 100-m REMA mosaic, we propose a novel hierarchical delta surface method for interpolating the created delta surface to achieve an efficient, unbiased, and continuous filled surface. High-precision laser altimetry data were used to evaluate the vertical accuracy of the REMA mosaic. To eliminate the influence of the temporal surface elevation change and penetration depth between the DEM and laser altimetry data, we developed a new local contrast evaluation method by calculating the vertical accuracy of the equal-area buffer as a control to that of the center void. The experimental results with the Ice, Cloud and Land elevation Satellite-2 (ICESat-2) laser altimetry data validate that the filled 100-m REMA mosaic of voids can reach an elevation accuracy similar to that of the original REMA dataset for the west and east Antarctic ice sheet as well as the entire continent. In the Antarctic Peninsula area, the vertical accuracy of the filled REMA mosaic by the proposed method is superior to that of the officially released version of the Polar Geospatial Center. This is the first time that a gapless, seamless, and accurate 100-m REMA mosaic has been released which has the potential to be extensively used in glaciological applications. The proposed methods can also be used in void-filling of different types of DEM products and their elevation accuracy evaluations.

Toward Fast and Reliable Potential Energy Surfaces for Metallic Pt Clusters by Hierarchical Delta Neural Networks.

Data-driven machine learning force fields (MLFs) are more and more popular in atomistic simulations and exploit machine learning methods to predict energies and forces for unknown structures based on the knowledge learned from an existing reference database. The latter usually comes from density functional theory calculations. One main drawback of MLFs is that physical laws are not incorporated in the machine learning models, and instead, MLFs are designed to be very flexible to simulate complex quantum chemistry potential energy surface (PES). In general, MLFs have poor transferability, and hence, a very large trainset is required to span all the target feature space to get a reliable MLF. This procedure becomes more troublesome when the PES is complicated, with a large number of degrees of freedom, in which building a large database is inevitable and very expensive, especially when accurate but costly exchange-correlation functionals have to be used. In this manuscript, we exploit a high-dimensional neural network potential (HDNNP) on Pt clusters of sizes from 6 to 20 as one example. Our standard level of energy calculation is DFT GGA (PBE) using a plane wave basis set. We introduce an approximate but fast level with the PBE functional and a minimal atomic orbital basis set, and then, a more accurate but expensive level, using a hybrid functional or nonlocal vdW functional and a plane wave basis set, is reliably predicted by learning the difference with HDNNP. The results show that such a differential approach (named ΔHDNNP) can deliver very accurate predictions (error <10 meV/atom) in reference to converged basis set energies as well as more accurate but expensive xc functionals. The overall speedup can be as large as 900 for a 20 atom Pt cluster. More importantly, ΔHDNNP shows much better transferability due to the intrinsic smoothness of the delta potential energy surface, and accordingly, one can use much smaller trainset data to obtain better accuracy than the conventional HDNNP. A multilayer ΔHDNNP is thus proposed to obtain very accurate predictions versus expensive nonlocal vdW functional calculations in which the required trainset is further reduced. The approach can be easily generalized to any other machine learning methods and opens a path to study the structure and dynamics of Pt clusters and nanoparticles.

Reverse Engineering Input Syntactic Structure from Program Execution and Its Applications

Program input syntactic structure is essential for a wide range of applications such as test case generation, software debugging, and network security. However, such important information is often not available (e.g., most malware programs make use of secret protocols to communicate) or not directly usable by machines (e.g., many programs specify their inputs in plain text or other random formats). Furthermore, many programs claim they accept inputs with a published format, but their implementations actually support a subset or a variant. Based on the observations that input structure is manifested by the way input symbols are used during execution and most programs take input with top-down or bottom-up grammars, we devise two dynamic analyses, one for each grammar category. Our evaluation on a set of real-world programs shows that our technique is able to precisely reverse engineer input syntactic structure from execution. We apply our technique to hierarchical delta debugging (HDD) and network protocol reverse engineering. Our technique enables the complete automation of HDD, in which programmers were originally required to provide input grammars, and improves the runtime performance of HDD. Our client study on network protocol reverse engineering also shows that our technique supersedes existing techniques.