Graphics Memory Research Articles

Low-overhead dynamic sharing of graphics memory space in GPU virtualization environments

The proliferation of GPU intensive workloads has created a new challenge for low-overhead and efficient GPU virtualization solutions over GPU clouds. gVirt is a full GPU virtualization solution for Intel’s integrated GPUs that share system’s on-board memory for graphics memory. In order to solve the inherent scalability limitation on the number of simultaneous virtual machines (VM) in gVirt, gScale proposed a dynamic sharing scheme for global graphics memory among VMs by copying the entries in a private graphics translation table (GTT) to a physical GTT along with a GPU context switch. However, copying entries between private GTT and physical GTT often causes significant overhead, which becomes worse when the global graphics memory space shared by each VM is overlapped. This paper identifies that the copy overhead caused by GPU context switch is one of the major bottlenecks in performance improvement and proposes a low-overhead dynamic memory management scheme called DymGPU. DymGPU provides two memory allocation algorithms such as size-based and utilization-based algorithms. While the size-based algorithm allocates memory space based on the memory size required by each VM, the utilization-based algorithm considers GPU utilization of each VM to allocate memory space. DymGPU is also dynamic in the sense that the global graphics memory space used by each VM is rearranged at runtime by periodically checking idle VMs and GPU utilization of each runnable VM. We have implemented our proposed approach in gVirt and confirmed that the proposed scheme reduces GPU context switch time by up to 53% and improved the overall performance of various GPU applications by up to 39%.

Changes of dopamine D2 receptor in patients with insomnia and its clinical significance

Objective To explore the changes of dopamine D2 receptor in dopamine pathway in insomnia patients and discuss its clinical significance. Methods From January 2016 to December 2016, 15 patients with insomnia (1 male, 14 females, age: (44.3±8.6) years) and 15 gender-/age- matched-healthy volunteers (control group; 3 males, 12 females, age: (40.5±9.0) years) were included to undergo resting brain 11C-Raclopride PET/CT imaging. The D2 receptor binding potential (BPND) of the dopamine pathway was calculated by molecular imaging and kinetic analysis toolbox (MIAKAT) software. The BPND, Hamilton depression scale (HAMD), transient and graphics memory scale results were compared with two-sample t test and Mann-Whitney u test between the two groups. Pearson correlation analysis was used to evaluate the correlation between BPND (nucleus accumbens, caudate nucleus, putamen) and Pittsburgh sleep quality index (PSQI), HAMD, course of disease, transient memory and graphical memory scale scores in the patient group. Results The BPND in bilateral putamen, nucleus accumbens and left caudate nucleus of patients was lower than that of controls(left putamen: z=-2.717, right putamen: z=-2.883, both P 0.05). The scores of HAMD in the patient group were higher than those in control group (t=10.273, P<0.01), while the scores of instantaneous memory (t=-4.888, P<0.01) and graphical memory scale (t=-2.624, P<0.05) were lower. There were significant negative correlations between the BPND of bilateral nucleus accumbens, caudate nucleus and putamen and the course of insomnia in the patient group (r range: -0.761 to -0.682, all P<0.01). Conclusion Patients with insomnia have abnormal neurotransmitter system of dopamine D2 and it may play a role in the pathogenesis of insomnia. Key words: Sleep initiation and maintenance disorders; Receptors, dopamine; Raclopride; Carbon radioisotopes; Positron-emission tomography; Tomography, X-ray computed; Magnetic resonance imaging

Scalable Wavelet-Based Coding of Irregular Meshes With Interactive Region-of-Interest Support

This paper proposes a novel functionality in wavelet-based irregular mesh coding, which is interactive region-of-interest (ROI) support. The proposed approach enables the user to define the arbitrary ROIs at the decoder side and to prioritize and decode these regions at arbitrarily high-granularity levels. In this context, a novel adaptive wavelet transform for irregular meshes is proposed, which enables: 1) varying the resolution across the surface at arbitrarily fine-granularity levels and 2) dynamic tiling, which adapts the tile sizes to the local sampling densities at each resolution level. The proposed tiling approach enables a rate-distortion-optimal distribution of rate across spatial regions. When limiting the highest resolution ROI to the visible regions, the fine granularity of the proposed adaptive wavelet transform reduces the required amount of graphics memory by up to 50%. Furthermore, the required graphics memory for an arbitrary small ROI becomes negligible compared to rendering without ROI support, independent of any tiling decisions. Random access is provided by a novel dynamic tiling approach, which proves to be particularly beneficial for large models of over 106 ~ 107 vertices. The experiments show that the dynamic tiling introduces a limited lossless rate penalty compared to an equivalent codec without ROI support. Additionally, rate savings up to 85% are observed while decoding ROIs of tens of thousands of vertices.

Resource efficient finite element computing on multicore architectures

In this paper, we consider the construction of efficient finite element algorithms on three-dimensional unstructured meshes that take into account the complex parallel synchronization processes, the memory distribution problems and data storage. A layer-by-layer partitioning of the meshes into subdomains without branching internal boundaries is proposed to simplify the access to independent data and parallel computing at different stages of the finite element problem solving on unstructured meshes in multiply connected domains. The predictive capacity of the time efficiency and resource intensity for the proposed algorithmic solutions is analyzed. The analysis of the resource efficiency of the algorithms is given for the element-by-element scheme for forming and solving the system of linear algebraic equations of the finite element method. It is shown that the low arithmetic intensity of the algorithms considered results in the fact that their performance is limited by the bandwidth of the memory subsystem rather than by the processors'performance. The graphic memory has a larger bandwidth than the random-access memory. This allows a significant increase in the performance of the algorithm on GPU.

Ballooning Graphics Memory Space in Full GPU Virtualization Environments

Advances in virtualization technology have enabled multiple virtual machines (VMs) to share resources in a physical machine (PM). With the widespread use of graphics-intensive applications, such as two-dimensional (2D) or 3D rendering, many graphics processing unit (GPU) virtualization solutions have been proposed to provide high-performance GPU services in a virtualized environment. Although elasticity is one of the major benefits in this environment, the allocation of GPU memory is still static in the sense that after the GPU memory is allocated to a VM, it is not possible to change the memory size at runtime. This causes underutilization of GPU memory or performance degradation of a GPU application due to the lack of GPU memory when an application requires a large amount of GPU memory. In this paper, we propose a GPU memory ballooning solution calledgBalloonthat dynamically adjusts the GPU memory size at runtime according to the GPU memory requirement of each VM and the GPU memory sharing overhead. ThegBalloonextends the GPU memory size of a VM by detecting performance degradation due to the lack of GPU memory. ThegBalloonalso reduces the GPU memory size when the overcommitted or underutilized GPU memory of a VM creates additional overhead for the GPU context switch or the CPU load due to GPU memory sharing among the VMs. We implemented thegBalloonby modifying thegVirt, a full GPU virtualization solution for Intel’s integrated GPUs. Benchmarking results show that thegBalloondynamically adjusts the GPU memory size at runtime, which improves the performance by up to 8% against thegVirtwith 384 MB of high global graphics memory and 32% against thegVirtwith 1024 MB of high global graphics memory.

Layer-based visualization and biomedical information exploration of multi-channel large histological data

Background and objectiveModern microscopes can acquire multi-channel large histological data from tissues of human beings or animals, which contain rich biomedical information for disease diagnosis and biological feature analysis. However, due to the large size, fuzzy tissue structure, and complicated multiple elements integrated in the image color space, it is still a challenge for current software systems to effectively calculate histological data, show the inner tissue structures and unveil hidden biomedical information. Therefore, we developed new algorithms and a software platform to address this issue. MethodsThis paper presents a multi-channel biomedical data computing and visualization system that can efficiently process large 3D histological images acquired from high-resolution microscopes. A novelty of our system is that it can dynamically display a volume of interest and extract tissue information using a layer-based data navigation scheme. During the data exploring process, the actual resolution of the loaded data can be dynamically determined and updated, and data rendering is synchronized in four display windows at each data layer, where 2D textures are extracted from the imaging volume and mapped onto the displayed clipping planes in 3D space. ResultsTo test the efficiency and scalability of this system, we performed extensive evaluations using several different hardware systems and large histological color datasets acquired from a CryoViz 3D digital system. The experimental results demonstrated that our system can deliver interactive data navigation speed and display detailed imaging information in real time, which is beyond the capability of commonly available biomedical data exploration software platforms. ConclusionTaking advantage of both CPU (central processing unit) main memory and GPU (graphics processing unit) graphics memory, the presented software platform can efficiently compute, process and visualize very large biomedical data and enhance data information. The performance of this system can satisfactorily address the challenges of navigating and interrogating volumetric multi-spectral large histological image at multiple resolution levels.

Desarrollo histórico de la guitarra en Honduras desde el punto de vista académico, 1900-2016

El objetivo de la presente investigación fue analizar el desarrollo histórico de la guitarra en Honduras dentro de las ciudades de Tegucigalpa y San Pedro Sula, lugares donde existe mayor referencia en el ámbito académico y artístico-musical. Además de entrevistar a exponentes del arte específicamente de la guitarra, también se incluyeron directores de centros de enseñanza musical y personalidades del acontecer nacional que poseen referencias históricas de la sinergia artística, social, cultural, económica y política del país. El profesor Donaldo Umanzor de 90 años (Tegucigalpa) y Don Francisco Carranza de 72 años (San Pedro Sula) fueron informantes claves por edad, trayectoria, viabilidad, factibilidad y ubicación geográfica. La investigación comprueba el vacío teórico que existe sobre la temática, se ha descrito una cronología de los precursores del estudio de la guitarra en el país, biografía, obra, aportes en la enseñanza, metodologías, partituras y memorias gráficas. Se evidenció el desarrollo de la guitarra en el contexto académico en estos últimos diez años, considerando las instituciones de enseñanza formal y no formal, específicamente la Escuela Nacional de Música, el Plan Propedéutico y la Licenciatura de Música con orientación en Guitarra de la UNAH, instituciones que han tenido más demanda y por ende, más eficiencia terminal. La Antología de la Guitarra en Honduras es resultado del presente estudio, siendo la primera vez que se ordena de manera sistemática un repertorio propio de la guitarra cronológicamente, considerando la edad de los autores.

50 million atoms scale molecular dynamics modelling on a single consumer graphics card

This paper presents a dynamic cell-list method for realizing large scale molecular dynamics (MD) simulations with more than 50 million atoms on a single consumer graphics card. It adapts the cell-list algorithm by introducing an efficient two-step atom location scheme and a dynamic memory allocation scheme such that only those cells containing atoms consume device memory. In addition, a large amount of memory is saved since it does not use the neighbour list. The computational efficiency is improved by reducing the memory loading times and maximizing coalesced memory access as compared to methods utilizing neighbour lists, since memory bandwidth is becoming the bottle-neck of the latest GPUs. As a result, MD simulations with more than 50 million atoms utilizing advanced three-body interaction potential are made possible on a consumer graphics card with just 11 GB of graphics memory. The proposed framework is designed to run totally on the graphics card, with all the data stored in the graphics memory to avoid the time-consuming data transfer between host and device. It achieves 2.5 times the speed and 20 times the atom number of the latest Lammps GPU package on the NVIDIA GTX 1080Ti GPU. The proposed framework is expected to help adapt existing MD packages for supporting large scale MD simulations on personal desktops and thereby extend MD to a wider range of researchers and engineers.

Signal Integrity Design and Analysis of Silicon Interposer for GPU-Memory Channels in High-Bandwidth Memory Interface

In this paper, for the first time, we designed and analyzed channels between a graphic processing unit and memory in a silicon interposer for a 3-D stacked high bandwidth memory (HBM). We thoroughly analyzed and verified the electrical characteristics of the silicon interposer considering various design parameters, such as the channel width and space, redistribution layer via, and under bump metallurgy pads. In particular, we also considered the meshed ground planes used for the proposed transmission lines, which are microstrip and strip lines. Signal integrity (SI) of the proposed channels in the silicon interposer was successfully analyzed and verified using a full 3-D electromagnetic solver and circuit simulations. Based on the extracted lumped circuit resistance, inductance, conductance and capacitance parameters, we thoroughly analyzed the channel characteristics and identified the parameters that dominantly affect SI in relation to each frequency range. From the analyzed insertion loss and far end crosstalk, we verified SI of the silicon interposer by eye-diagram simulations in terms of eye-height voltage and timing jitter in the time domain. In the worst case, the eye-height voltage and timing jitter of the proposed microstrip lines are 0.911 V and 36.8 ps, respectively, with 72 mV of signal coupling. The eye-height voltage and timing jitter of the proposed strip line are 0.887 V and 42.1 ps with 34 mV of single couplings. We show that the proposed channels of the silicon interposer can successfully transfer data at a 2-Gb/s data rate. Finally, we propose concepts and solutions for the next-generation HBM interface with higher data rates up to 8 Gb/s.

Scalable GPU Virtualization with Dynamic Sharing of Graphics Memory Space

With increasing GPU-intensive workloads deployed on cloud, cloud service providers are seeking for practical and efficient GPU virtualization solutions. However, the cutting-edge GPU virtualization techniques such as gVirt still suffer from the restriction of scalability, which constrains the number of guest virtual GPU instances. This paper presents gScale, a scalable and practical open source GPU virtualization solution based on gVirt. gScale presents a sharing mechanism which combines partition and sharing together to break the hardware limitation of global graphics memory space. Particularly, we propose two approaches for gScale: (1) the private shadow graphics translation table (GTT) , which enables global graphics memory space sharing among virtual GPUs, (2) ladder mapping and fence memory space pool, which allows CPU access host physical memory space (serving the graphics memory) to bypass global graphics memory space. Furthermore, to mitigate the performance degradation caused by switching private shadow GTT when the number of vGPUs scales up, four other mechanisms are proposed: (1) slot sharing, which improves the performance of vGPU by dividing the high global graphics memory into multiple slots, (2) fine-grained slotting, which provides a flexible virtual graphics memory configuration, (3) predictive GTT copy mechanism, which reduces the performance loss by switching private shadow GTT before context switch, (4) predictive-copy aware scheduling, which maximizes the improvement of predictive GTT copy mechanism in cloud environment. Evaluation shows that gScale scales up to 15 guest virtual GPU instances in Linux or 12 guest virtual GPU instances in Windows, which is 5x and 4x, respectively, that of gVirt. At the same time, gScale incurs a slight but acceptable runtime overhead when hosting multiple virtual GPU instances.

GMig

This paper introduces gMig, an open-source and practical GPU live migration solution for full virtualization. By taking advantage of the dirty pattern of GPU workloads, gMig presents the One-Shot Pre-Copy combined with the hashing based Software Dirty Page technique to achieve efficient GPU live migration. Particularly, we propose three approaches for gMig: 1) Dynamic Graphics Address Remapping, which parses and manipulates GPU commands to adjust the address mapping to adapt to a different environment after migration, 2) Software Dirty Page, which utilizes a hashing based approach to detect page modification, overcomes the commodity GPU's hardware limitation, and speeds up the migration by only sending the dirtied pages, 3) One-Shot Pre-Copy, which greatly reduces the rounds of pre-copy of graphics memory. Our evaluation shows that gMig achieves GPU live migration with an average downtime of 302 ms on Windows and 119 ms on Linux. With the help of Software Dirty Page, the number of GPU pages transferred during the downtime is effectively reduced by 80.0%.

Cost-effective printing of 3D objects with self-supporting property

The fused deposition modeling (FDM) printer is a simple, affordable and widely used device in the 3D printing society. However, the high price of printing materials is one of major restrictive factors for its further application. Based on the self-supporting property of printing materials, we present an optimization method to reduce the total material consumption of 3D printed objects themselves and their support structures for FDM printers in this paper. We first develop an orientation optimization scheme to reduce the outer support volume of a printed model. The volume is evaluated according to the depths of 3D model fragments obtained by the depth peeling technique in an optimization process. We then build a self-supporting frame with a set of scale-adaptive parallelepiped grids to replace the solid interior of the printed model for further reducing the material consumption. In our orientation optimization scheme, the overhanging area detecting function can detect the self-supporting regions of a 3D model in terms of the depths stored in the graphical processing unit memory. The self-supporting frame with grid structures inside printed models does not need to add additional support structures during the printing process. Experimental results indicate that our method is faster and consumes less printing materials than the state-of-the-art algorithms.

Barcoding technologies for facial biometrics: current status and new solutions

Subject of Research.Application of barcoding technologies in the tasks of facial biometrics is posed and discussed. We analyze the achievements and estimate the shortcomings of existing solutions and examples of barcode creation according to the face images and the features extracted by them. Method. The ways of the problem implementation are determined and new solutions are presented based on linear (Code 128) and two-dimensional (QR) bar codes, as well as their color variants. The composition and volume of data are considered being used in the facial biometry and related applications: medicine, criminalistics and forensic-medical examination. Among these data there are face images, as well as sets of anthropometric points and additional information to them, information about the phenotype of FI and gender, and, finally, documentary information. Main Results. We have shownthe results of these data "recording and transferring" within the framework of various barcode layouts, as well as the results of their reading and ways of hiding from reading. The proposed color barcodes are defined as "BIO Code 128" and "BIO QR-code". While graphical display and computer memory record, they can be viewed as colored raster images that carry information about the face in each layer. At this, documentary information can be read directly from such color images by standard barcode scanners, and the rest of the information (face image itself, its anthropometric, accompanying parameters) is read and restored after their decomposition into layers R, green and blue.Practical Relevance. The layout variants of the "BIO Code 128" and "BIO QR-code" barcodes and the programs for their generation (written in the MATLAB package environment) can be used in the further studies of the barcoding problem in the tasks of the facial biometrics and its applications.

Leveraging Power-Performance Relationship of Energy-Efficient Modern DRAM Devices

Computer servers are equipped with an increasing number of memory modules each with more capacity, making main-memory systems now the second most energy-consuming component trailing only processors in big-memory servers. These big-memory servers and their main-memory systems should offer high energy efficiency. In pursuit of energy-efficient main-memory systems, prior work exploited mobile low power double data rate (LPDDR) devices' advantages (lower power than DDR devices) while attempting to surmount their limitations (longer latency, lower bandwidth, or both). However, we show that such main-memory architectures (based on the latest LPDDR4 devices) are no longer effective and even hurt overall energy efficiency of servers by 49% on memory-intensive workloads compared with ones based on DDR4 devices. The reason is that the power consumption of modern DDR4 devices has substantially reduced by adopting the strength of mobile and graphics memory whereas LPDDR4 has focused more on increasing data transfer rates while sacrificing energy efficiency; the power consumption of DDR4 devices can significantly vary across manufacturers in this analysis. Moreover, exploring new energy-saving features of DDR4 devices in depth, we show that activating these features often hurts overall energy efficiency of servers because of their performance penalties. Subsequently, we propose a simple but effective scheme that adaptively exploits DRAM power-down modes and hence improves the system energy-delay product by 4.0%.

Atlas visual de la memoria. Una forma de visualizar y representar el conflicto en Colombia

The present text investigates the definition of theoretical and technical parameters from information science and studies on photography as a document and memory device, contributing to the development of the Digital Memories Repository, which attempts to recover, save and To construct alternative forms of visibilizing and representing the documentary information that is produced in the social institutions and organizations of human rights, as a contribution to the preservation of the graphic and conceptual memory of the conflict in Colombia.

Building the graphics memory of the stiffness matrix of the beam

Building the graphics memory of the stiffness matrix of the beam

Acerca del concepto de memoria gráfica

En Brasil, desde principios del siglo XXI, la expresión memoria gráfica ha sido usada, cada vez con más frecuencia, para denominar una línea de estudios que busca comprender la importancia y el valor de los artefactos visuales, en particular, los impresos efímeros, para la creación de un sentido de identidad local a través del diseño. Tales artefactos han sido invisibilizados por mucho tiempo por las corrientes dominantes sobre lo que sería el diseño y, en consecuencia, de lo que debe hacer parte de la historia del diseño. Otras historias del diseño que incluyan tales artefactos aún deben ser descubiertas y narradas.Este artículo propone una reflexión acerca de los contextos en los que ha sido utilizada la memoria gráfica en el campo de la investigación de los artefactos visuales en Brasil, partiendo de una discusión sobre las conexiones posibles con conceptos más conocidos, como los de cultura de la impresión, cultura visual y cultura material, dando ejemplos de investigaciones realizadas, y llegando a una propuesta de cómo la idea de memoria gráfica puede ser adoptada como una estrategia para abordar el descubrimiento o la construcción de historias de diseño gráfico y visual en países no hegemónicos.

O-CNN

We present O-CNN , an Octree-based Convolutional Neural Network (CNN) for 3D shape analysis. Built upon the octree representation of 3D shapes, our method takes the average normal vectors of a 3D model sampled in the finest leaf octants as input and performs 3D CNN operations on the octants occupied by the 3D shape surface. We design a novel octree data structure to efficiently store the octant information and CNN features into the graphics memory and execute the entire O-CNN training and evaluation on the GPU. O-CNN supports various CNN structures and works for 3D shapes in different representations. By restraining the computations on the octants occupied by 3D surfaces, the memory and computational costs of the O-CNN grow quadratically as the depth of the octree increases, which makes the 3D CNN feasible for high-resolution 3D models. We compare the performance of the O-CNN with other existing 3D CNN solutions and demonstrate the efficiency and efficacy of O-CNN in three shape analysis tasks, including object classification, shape retrieval, and shape segmentation.

Cross-Layer Optimization for Multilevel Cell STT-RAM Caches

Spin-transfer torque random access memory (STT-RAM), as an emerging nonvolatile memory technology, provides very dense array structure and extremely low leakage power consumption. It demonstrates a great potential in replacing conventional static random access memory technology to develop the next-generation on-chip cache memory of microprocessors and graphics processing units. The multilevel cell (MLC) design of STT-RAM that stores two or more bits in one cell potentially has higher storage capacity and faster system performance, attracting significant attention. In this paper, we first quantitatively evaluated the data storage density of the MLC STT-RAM. Our results revealed limited density improvement because of the large size of access transistor induced by high write current amplitude requirement and asymmetry of switching behavior. Moreover, the read and write accesses of existing MLC STT-RAM cache designs require two-step operation. The system level evaluation shows that the long access latency could amortize the performance speed brought by larger cache size, and even degrade the system performance for some applications. To unleash the potential of MLC STT-RAM cache, we proposed a new design through a cross-layer co-optimization. The memory cell structure integrated the reversed stacking of magnetic junction tunneling for a more balanced device and design tradeoff. In architecture development, we presented an adaptive mode switching mechanism: based on application’s memory access behavior, the MLC STT-RAM cache can dynamically change between low latency single-level cell mode and high capacity MLC mode. Furthermore, we divided cache lines into fast and slow regions and investigated new data migration policies to allocate frequently access data to fast regions. Simulation results show that the proposed techniques can improve the system performance by 10.2% and reduce the energy consumption on cache by 9.5% compared with conventional MLC STT-RAM cache design.

Effect of graphic simplification and graphic metaphor on the memory and identification of travel map

With the rapid development of the tourism industry, the design of tourist guide signs has become a crucial concern. This study investigated the effects of 3 graphic simplification techniques and 3 graphic metaphor techniques on graphic memory and identification. The results indicated that simplification and the interaction between simplification and metaphor influence graphic memory; moreover, simplification and metaphor were found to significantly affect identification. These findings suggest that iconic simplification with appearance-resembling metaphors, and metaphor techniques with text simplification result in superior memory. Simplification techniques with appearance-resembling metaphors and metaphor techniques with text simplification also yielded superior identification. This paper thus provides a reference model for map designers who are producing digital and travel maps.