Tesla K20c Research Articles

Field‐programmable gate array acceleration of the Tersoff potential in LAMMPS

AbstractMolecular dynamics simulation is a common method to help humans understand the microscopic world. The traditional general‐purpose high‐performance computing platforms are hindered by low computational and power efficiency, constraining the practical application of large‐scale and long‐time many‐body molecular dynamics simulations. In order to address these problems, a novel molecular dynamics accelerator for the Tersoff potential is designed based on field‐programmable gate array (FPGA) platforms, which enables the acceleration of LAMMPS using FPGAs. Firstly, an on‐the‐fly method is proposed to build neighbor lists and reduce storage usage. Besides, multilevel parallelizations are implemented to enable the accelerator to be flexibly deployed on FPGAs of different scales and achieve good performance. Finally, mathematical models of the accelerator are built, and a method for using the models to determine the optimal‐performance parameters is proposed. Experimental results show that, when tested on the Xilinx Alveo U200, the proposed accelerator achieves a performance of 9.51 ns/day for the Tersoff simulation in a 55,296‐atom system, which is a 2.00 increase in performance when compared to Intel I7‐8700K and 1.70 to NVIDIA Tesla K40c under the same test case. In addition, in terms of computational efficiency and power efficiency, the proposed accelerator achieves improvements of 2.00 and 7.19 compared to Intel I7‐8700K, and 4.33 and 2.11 compared to NVIDIA Titan Xp, respectively.

3D least-squares reverse time migration in VTI media based on pseudoacoustic wave equation and multi-GPU parallel acceleration

As the high-density seismic acquisition is progressively becoming an essential mode of seismic exploration, developing high-precision imaging methods for 3D datasets has become an urgent industry requirement. Unfortunately, the two-dimensional least-squares reverse time migration (LSRTM) can hardly deal with the influence of the third dimension because the structure of reflectors is constantly changing in nearly all directions. Furthermore, anisotropy is very common in underground media, which makes the LSRTM based on isotropic assumption fail to locate the reflectors accurately and seriously limits the identification of geological stratification. Considering the above two problems, this paper derives the least-squares reverse time migration in three-dimensional vertical transversely isotropic media (3D-VTI-LSRTM) and develops a workflow which can realize 3D-LSRTM in VTI media based on this theory. Our workflow uses multi-GPU acceleration and heterogeneous CPU/GPU parallelism. Compared with the general parallel CPU workflow or single GPU workflow, it is feasible that the GPU application significantly improves computing efficiency in practical applications. In addition, this paper extends VTI-LSRTM from 2D to 3D, which solves the problem that the 2D algorithm cannot eliminate the lateral direction reflection artifacts. For the self-made 3D-VTI depression model and the 3D-VTI-SEG/EAGE salt model, we use four NVIDIA Tesla k40c graphics cards for testing. The results of the tests are correct and significantly improved compared to the 3D-VTI-RTM method.

Parallel Optimal Design of SSR Response Signal Processing Algorithms Based on GPU

In order to solve the key problems of secondary surveillance radar signal debugging on special devices such as DSP and FPGA, a parallel processing scheme of secondary surveillance radar response signal is proposed based on CPU–GPU architecture. It can effectively reduce the difficulty of code development and improve the portability of program. The parallel optimization design of each processing model of response signal is made through the characteristics of shared memory in CPU–GPU architecture to improve the processing speed of the algorithm. The proposed scheme is tested and analyzed on different graphics cards by the secondary surveillance radar Mode-5 response signal in this paper. The experimental results showed that it takes 8390.52960 us to implement the signal processing algorithm based on NVIDIA Tesla K40c graphics card, which can save 51.98% of time than NVIDIA Quadro K4200 graphics card, and makes it possible to do the real-time processing of the secondary surveillance radar signal.

Algorithms to Speed up Contour Tracing in Real Time Image Processing Systems

Contour tracing is an important pre-processing step in many image processing applications such as feature recognition, biomedical imaging, security and surveillance. As single processor architectures reach their performance limits, parallel processing architectures offer energy efficient and high performance solutions for real time applications. Parallel processing architectures, are thus used for several real time image processing applications. Among the several interconnection schemes available, Cayley graph based interconnections offer easy routing and symmetric implementation capabilities. For parallel processing systems with a Cayley graph based interconnection scheme, torus, we developed three accelerated algorithms corresponding to three existing families of contour tracing algorithms. We simulated these algorithms on a parallel processing framework to quantify the normalized speed-up possible in any torus connected parallel processing system. We also compared our best performing algorithm with the existing parallel processing implementations for Nvidia GPUs. We observed a speed-up of up to 468 times using our algorithms on a parallel processing architecture in comparison to the corresponding algorithm on a single processor architecture. We evaluated a speed up of 194 (and 47) compared to the existing parallel processing contour tracing implementation on Tesla K40c (and Quadro RTX 5000 GPU hardware respectively). We observe that for torus connected parallel processing architectures used for image processing, our algorithms can be used to speed up contour tracing, without any hardware modification.

Hardware Implementation of Hierarchical Temporal Memory Algorithm

Hierarchical temporal memory (HTM) is an un-supervised machine learning algorithm that can learn both spatial and temporal information of input. It has been successfully applied to multiple areas. In this paper, we propose a multi-level hierarchical ASIC implementation of HTM, referred to as processor core, to support both spatial and temporal pooling. To improve the unbalanced workload in HTM, the proposed design provides different mapping methods for the spatial and temporal pooling, respectively. In the proposed design, we implement a distributed memory system by assigning one dedicated memory bank to each level of hierarchy to improve the memory bandwidth utilization efficiency. Finally, the hot-spot operations are optimized using a series of customized units. Regarding scalability, we propose a ring-based network consisting of multiple processor cores to support a larger HTM network. To evaluate the performance of our proposed design, we map an HTM network that includes 2,048 columns and 65,536 cells on both the proposed design and NVIDIA Tesla K40c GPU using the KTH database as input. The latency and power of the proposed design is 6.04 ms and 4.1 W using GP 65 nm technology. Compared to the equivalent GPU implementation, the latency and power is improved 12.45× and 57.32×, respectively.

Microwave Correlation Forward-Looking Super-Resolution Imaging Based on Compressed Sensing

Forward-looking correlated imaging plays an increasingly important role in modern radar imaging systems. It overcomes disadvantages of traditional side or squint synthetic aperture radar (SAR) which is dependent on specific relative motion between the radar and target scene. A new microwave forward-looking correlated 3-D imaging method based on random radiation field combined with sparse reconstruction is proposed in this article. Firstly, phased array radar (PAR) is adopted to form different and random antenna patterns. Then, combined with the compressed sensing (CS) theory, the target image can be recovered with very few samples which can break through Rayleigh resolution limitation. Furthermore, the proposed method can achieve resolution at least 5.5 times higher than real aperture imaging. To raise computation efficiency of sparse reconstruction, an improved quasi-Newton iteration method based on graphics processing unit (GPU) platform is developed. Meanwhile, a GPU-based (NVIDIA Tesla K40c) accelerated computing method can significantly reduce the processing time compared with the time given by a personal computer (PC). Both simulation and field experiment verify the validity of the proposed method.

Multipatch-GLCM for Texture Feature Extraction on Classification of the Colon Histopathology Images using Deep Neural Network with GPU Acceleration

Cancer is one of the leading causes of death in the world. It is the main reason why research in this field becomes challenging. Not only for the pathologist but also from the view of a computer scientist. Hematoxylin and Eosin (H&E) images are the most common modalities used by the pathologist for cancer detection. The status of cancer with histopathology images can be classified based on the shape, morphology, intensity, and texture of the image. The use of full high-resolution histopathology images will take a longer time for the extraction of all information due to the huge amount of data. This study proposed advance texture extraction by multi-patch images pixel method with sliding windows that minimize loss of information in each pixel patch. We use texture feature Gray Level Co-Occurrence Matrix (GLCM) with a mean-shift filter as the data pre-processing of the images. The mean-shift filter is a low-pass filter technique that considers the surrounding pixels of the images. The proposed GLCM method is then trained using Deep Neural Networks (DNN) and compared to other classification techniques for benchmarking. For training, we use two hardware: NVIDIA GPU GTX-980 and TESLA K40c. According to the study, Deep Neural Network outperforms other classifiers with the highest accuracy and deviation standard 96.72±0.48 for four cross-validations. The additional information is that training using Theano framework is faster than Tensorflow for both in GTX-980 and Tesla K40c.

WooKong: A Ubiquitous Accelerator for Recommendation Algorithms with Custom Instruction Sets on FPGA

Recommendation algorithms, such as Neighborhood-based Collaborative- Filtering (CF), have been widely applied in various emerging machine learning applications. However, under the circumstance of the explosive big data, it poses significant challenges to CF recommendation algorithms as it is becoming quite time and energy-consuming. It has to be optimized and accelerated by powerful engines to process on large data scale. To solve these problems, in this article, we propose WooKong, a ubiquitous accelerator architecture for the collaborative-filtering recommendation on FPGA. It is able to accommodate three types of CF recommendation algorithms, including User-based CF, Item-based CF, and SlopeOne recommendations algorithms, with five different similarity analysis metrics including Jaccard, Cosine, CosineIR, euclidean, and Pearson. To maintain flexibility for these different CF algorithms and metrics, we adopt custom instruction sets to manipulate the learning and prediction accelerators. We implement a hardware prototype on a real Xilinx Zynq FPGA development board. Experimental results show that the proposed learning and prediction accelerators can achieve 8.0X speedup and 1.7X speedup compared with an Intel i7 processor respectively. The accelerator has the energy benefits of up to 137.4X compared with an NVIDIA Tesla K40C GPU, with the affordable hardware cost.

Real-Time Lung Tumor Tracking Using a CUDA Enabled Nonrigid Registration Algorithm for MRI.

Objective: This study intends to develop an accurate, real-time tumor tracking algorithm for the automated radiation therapy for cancer treatment using Graphics Processing Unit (GPU) computing. Although a previous moving mesh based tumor tracking approach has been shown to be successful in delineating the tumor regions from a sequence of magnetic resonance image, the algorithm is computationally intensive and its computation time on standard Central Processing Unit (CPU) processors is too slow to be used clinically especially for automated radiation therapy system. Method: A re-implementation of the algorithm on a low-cost parallel GPU-based computing platform is utilized to accelerate this computation at a speed that is amicable to clinical usages. Several components in the registration algorithm such as the computation of similarity metric are inherently parallel which fits well with the GPU parallel processing capabilities. Solving a partial differential equation numerically to generate the mesh deformation is one of the computationally intensive components which has been accelerated by utilizing a much faster shared memory on the GPU. Results: Implemented on an NVIDIA Tesla K40c GPU, the proposed approach yielded a computational acceleration improvement of over 5 times its implementation on a CPU. The proposed approach yielded an average Dice score of 0.87 evaluated over 600 images acquired from six patients. Conclusion: This study demonstrated that the GPU computing approach can be used to accelerate tumor tracking for automated radiation therapy for mobile lung tumors. Clinical Impact: Accurately tracking mobile tumor boundaries in real-time is important to automate radiation therapy and the proposed study offers an excellent option for fast tumor region tracking for cancer treatment.

Modeling and Decoupling the GPU Power Consumption for Cross-Domain DVFS

Dynamic voltage and frequency scaling (DVFS) is a popular technique to improve the energy-efficiency of high-performance computing systems. It allows placing the devices into lower performance states when the computational demands are lower, opening the possibility for significant power/energy savings. This work presents a GPU power consumption model, used to predict the GPU power consumption of any application at different frequency levels. To obtain this model, an estimation algorithm is proposed, relying on careful benchmarking of the GPU architecture. The model can estimate the contribution of twelve different GPU components (FP32-ADD/MUL/FMA, FP64-ADD/MUL/FMA, INT, SF, CF units, shared memory, L2-cache, and DRAM) to the GPU power consumption. Different model use cases are evaluated (fixed-frequency, DVFS, and scaling-factors), which can obtain both the total or the per-component GPU power consumption. A technique to export models to a distinct GPU from the one it was estimated on is also proposed. These approaches were extensively validated on five different GPUs from the three most recent microarchitectures with a set of 42 standard benchmarks, achieving very accurate predictions. In particular, the scaling-factor power model achieves an average prediction error of 3.5 percent (Titan Xp), 4.6 percent (GTX Titan X), 3.1 percent (GTX 980) and 2.4 percent (Tesla K40c).

Multi-relaxation time lattice Boltzmann simulations of oscillatory instability in lid-driven flows of 2D semi-elliptical cavity

In this study, the multi-relaxation-time lattice Boltzmann method is applied to investigate the oscillatory instability of lid-driven flows in two-dimensional semi-elliptical cavities with different vertical-to-horizontal aspect ratios K in the range of 1.0–3.0. The program implemented in this study is parallelized using compute unified device architecture (CUDA), a parallel computing platform, and computations are carried out on NVIDIA Tesla K40c GPU. To carry out precise calculations, the CUDA algorithm is extensively investigated, and its parallel efficiency indicates that the maximum speedup is 47.6 times faster. Furthermore, the steady-oscillatory Reynolds numbers are predicted by implementing the CUDA-based programs. The amplitude coefficient is defined to quantify the time-dependent oscillation of the velocity magnitude at the monitoring point. The simulation results indicate that the transition Reynolds numbers correlate negatively with the aspect ratio of the semi-elliptical cavity and are smaller than those of the rectangular cavity at the same aspect ratio. In addition, the detailed vortex structures of the semi-elliptical cavity within a single period are also investigated when the Reynolds number is larger than the steady-oscillatory value to determine the effects of periodic oscillation of the velocity magnitude.

GPU-accelerated hierarchical Bayesian estimation of luminosity functions using flux-limited observations with photometric noise

We describe a computational framework for hierarchical Bayesian inference with simple (typically single-plate) parametric graphical models that uses graphics processing units (GPUs) to accelerate computations, enabling deployment on very large datasets. Its C++ implementation, CUDAHM (CUDA for Hierarchical Models) exploits conditional independence between instances of a plate, facilitating massively parallel exploration of the replication parameter space using the single instruction, multiple data architecture of GPUs. It provides support for constructing Metropolis-within-Gibbs samplers that iterate between GPU-accelerated robust adaptive Metropolis sampling of plate-level parameters conditional on upper-level parameters, and Metropolis-Hastings sampling of upper-level parameters on the host processor conditional on the GPU results. CUDAHM is motivated by demographic problems in astronomy, where density estimation and linear and nonlinear regression problems must be addressed for populations of thousands to millions of objects whose features are measured with possibly complex uncertainties. We describe a thinned latent point process framework for modeling such demographic data. We demonstrate accurate GPU-accelerated parametric conditional density deconvolution for simulated populations of up to 300,000 objects in ~1 hour using a single NVIDIA Tesla K40c GPU. Supplementary material provides details about the CUDAHM API and the demonstration problem.

High-performance implementation of a Runge–Kutta finite-difference scheme for the Higgs boson equation in the de Sitter spacetime

High performance computations are presented for the Higgs boson equation in the de Sitter Spacetime using explicit fourth order Runge–Kutta scheme on the temporal discretization and fourth order finite difference discretization in space. In addition to the full, (3+1)−dimensional equation we also examine the (1+1)−dimensional radial solutions. The numerical code for the (3+1)−dimensional equation is programmed in CUDA Fortran and is performed on NVIDIA Tesla K40c GPU Accelerators. The radial form of the equation is simulated in MATLAB. The numerical results demonstrate the existing theoretical result that under certain conditions bubbles form in the scalar field. We also demonstrate the known blow-up phenomena for the solutions of the related semilinear Klein–Gordon equation with imaginary mass. Our numerical studies suggest several previously not known properties of the solution for the Higgs boson equation in the de Sitter spacetime for which theoretical proofs do not exist yet: 1. smooth solution exists for all time if the initial conditions are compactly supported and smooth; 2. under some conditions no bubbles form; 3. solutions converge to step functions related to unforced, damped Duffing equations.

MSGD: A Novel Matrix Factorization Approach for Large-Scale Collaborative Filtering Recommender Systems on GPUs

Real-time accurate recommendation of large-scale recommender systems is a challenging task. Matrix factorization (MF), as one of the most accurate and scalable techniques to predict missing ratings, has become popular in the collaborative filtering (CF) community. Currently, stochastic gradient descent (SGD) is one of the most famous approaches for MF. However, it is non-trivial to parallelize SGD for large-scale CF MF problems due to the dependence on the user and item pair, which can cause parallelization over-writing. To remove the dependence on the user and item pair, we propose a multi-stream SGD (MSGD) approach, for which the update process is theoretically convergent. On that basis, we propose a Compute Unified Device Architecture (CUDA) parallelization MSGD (CUMSGD) approach. CUMSGD can obtain high parallelism and scalability on Graphic Processing Units (GPUs). On Tesla K20m and K40c GPUs, the experimental results show that CUMSGD outperforms prior works that accelerated MF on shared memory systems, e.g., DSGD, FPSGD, Hogwild!, and CCD++. For large-scale CF problems, we propose multiple GPUs (multi-GPU) CUMSGD (MCUMSGD). The experimental results show that MCUMSGD can improve MSGD performance further. With a K20m GPU card, CUMSGD can be 5-10 times as fast compared with the state-of-the-art approaches on shared memory platform.

GPU Parallel Implementation of Spatially Adaptive Hyperspectral Image Classification

Image classification is a very important tool for remotely sensed hyperspectral image processing. Techniques able to exploit the rich spectral information contained in the data, as well as its spatial-contextual information, have shown success in recent years. Due to the high dimensionality of hyperspectral data, spectral-spatial classification techniques are quite demanding from a computational viewpoint. In this paper, we present a computationally efficient parallel implementation for a spectral-spatial classification method based on spatially adaptive Markov random fields (MRFs). The method learns the spectral information from a sparse multinomial logistic regression classifier, and the spatial information is characterized by modeling the potential function associated with a weighted MRF as a spatially adaptive vector total variation function. The parallel implementation has been carried out using commodity graphics processing units (GPUs) and the NVIDIA's Compute Unified Device Architecture. It optimizes the work allocation and input/output transfers between the central processing unit and the GPU, taking full advantages of the computational power of GPUs as well as the high bandwidth and low latency of shared memory. As a result, the algorithm exploits the massively parallel nature of GPUs to achieve significant acceleration factors (higher than 70x) with regards to the serial and multicore versions of the same classifier on an NVIDIA Tesla K20C platform.

Evaluation of the computational efficacy in GPU-accelerated simulations of spiking neurons

To understand the mechanism of information processing by a biological neural network, computer simulation of a large-scale spiking neural network is an important method. However, because of a high computation cost of the simulation of a large-scale spiking neural network, the simulation requires high performance computing implemented by a supercomputer or a computer cluster. Recently, hardware for parallel computing such as a multi-core CPU and a graphics card with a graphics processing unit (GPU) is built in a gaming computer and a workstation. Thus, parallel computing using this hardware is becoming widespread, allowing us to obtain powerful computing power for simulation of a large-scale spiking neural network. However, it is not clear how much increased performance the parallel computing method using a new GPU yields in the simulation of a large-scale spiking neural network. In this study, we compared computation time between the computing methods with CPUs and GPUs in a simulation of neuronal models. We developed computer programs of neuronal simulations for the computing systems that consist of a gaming graphics card with new architecture (the NVIDIA GTX 1080) and an accelerator board using a GPU (the NVIDIA Tesla K20C). Our results show that the computing systems can perform a simulation of a large number of neurons faster than CPU-based systems. Furthermore, we investigated the accuracy of a simulation using single precision floating point. We show that the simulation results of single precision were accurate enough compared with those of double precision, but chaotic neuronal response calculated by a GPU using single precision is prominently different from that calculated by a CPU using double precision. Furthermore, the difference in chaotic dynamics appeared even if we used double precision of a GPU. In conclusion, the GPU-based computing system exhibits a higher computing performance than the CPU-based system, even if the GPU system includes data transfer from a graphics card to host memory.

Hydrodynamic modeling of flash flood in mountain watersheds based on high-performance GPU computing

Numerical accuracy and computational efficiency are the two key factors for flash flood simulation. In this paper, a two-dimensional fully hydrodynamic model is presented for the simulation of flash floods in mountain watersheds. A robust finite volume scheme is adopted to accurately simulate the overland flow with wet/dry fronts on highly irregular topography. A graphics processing unit-based parallel method using OpenACC is adopted to realize high-performance computing and then improve the computational efficiency. Since the finite volume scheme is explicit which involves many computationally intensive loop structures without data dependence, the parallel flash flood model can be easily realized by using OpenACC directives in an incremental developing way based on the serial model codes, except that data structure and transportation should be optimized for parallel algorithm. Model accuracy is validated by benchmark cases with exact solutions and experimental data. To further analyze the performance of the model, we considered a real flash flooding-prone area in China using a NVIDIA Tesla K20c card and three grid division schemes with different resolution. Results show that the proposed model can fast simulate the rainfall−runoff process related to the rapid mountain watersheds response, and a higher speedup ratio can be achieved for finer grids resolution. The proposed model can be used for real-time prediction of large-scale flash flood on high-resolution grids and thus has bright application prospects.

Moving-Target Position Estimation Using GPU-Based Particle Filter for IoT Sensing Applications

A particle filter (PF) has been introduced for effective position estimation of moving targets for non-Gaussian and nonlinear systems. The time difference of arrival (TDOA) method using acoustic sensor array has normally been used to for estimation by concealing the location of a moving target, especially underwater. In this paper, we propose a GPU -based acceleration of target position estimation using a PF and propose an efficient system and software architecture. The proposed graphic processing unit (GPU)-based algorithm has more advantages in applying PF signal processing to a target system, which consists of large-scale Internet of Things (IoT)-driven sensors because of the parallelization which is scalable. For the TDOA measurement from the acoustic sensor array, we use the generalized cross correlation phase transform (GCC-PHAT) method to obtain the correlation coefficient of the signal using Fast Fourier Transform (FFT), and we try to accelerate the calculations of GCC-PHAT based TDOA measurements using FFT with GPU compute unified device architecture (CUDA). The proposed approach utilizes a parallelization method in the target position estimation algorithm using GPU-based PF processing. In addition, it could efficiently estimate sudden movement change of the target using GPU-based parallel computing which also can be used for multiple target tracking. It also provides scalability in extending the detection algorithm according to the increase of the number of sensors. Therefore, the proposed architecture can be applied in IoT sensing applications with a large number of sensors. The target estimation algorithm was verified using MATLAB and implemented using GPU CUDA. We implemented the proposed signal processing acceleration system using target GPU to analyze in terms of execution time. The execution time of the algorithm is reduced by 55% from to the CPU standalone operation in target embedded board, NVIDIA Jetson TX1. Also, to apply large-scaled IoT sensing applications, we use NVIDIA Tesla K40c as target GPU. The execution time of the proposed multi-state-space model-based algorithm is similar to the one-state-space model algorithm because of GPU-based parallel computing. Experimental results show that the proposed architecture is a feasible solution in terms of high-performance and area-efficient architecture.

On the Evaluation of Different High-Performance Computing Platforms for Hyperspectral Imaging: An OpenCL-Based Approach

Hyperspectral imaging systems are a powerful tool for obtaining surface information in many different spectral channels that can be used in many different applications. Nevertheless, the huge amount of information provided by hyperspectral images also has a downside, since it has to be processed and analyzed. For such purpose, parallel hardware devices, such as field-programmable gate arrays (FPGAs) and graphic processing units (GPUs), are typically used, especially for hyperspectral imaging applications under real-time constraints. However, developing hardware applications typically requires expertise in the specific targeted device, as well as in the tools and methodologies that can be used to perform the implementation of the desired algorithms in that device. In this scenario, the Open Computing Language (OpenCL) emerges as a very interesting solution in which a single high-level language can be used to efficiently develop applications in multiple and different hardware devices. In this work, the parallel Fast Algorithm for Linearly Unmixing Hyperspectral Images (pFUN) has been implemented in two different NVIDIA GPUs, the GeForce GTX 980 and the Tesla K40c, using OpenCL. The obtained results are compared with the results provided by the previously developed NVIDIA CUDA implementation of the pFUN algorithm for the same GPU devices for comparing the efficiency of OpenCL against a more specific synthesis design language for the targeted hardware devices, such as CUDA is for NVIDIA GPUs. Moreover, the FUN algorithm has also been implemented into a Bitware Stratix V Altera FPGA, using OpenCL, for comparing the results that can be obtained using OpenCL when targeting different devices and architectures. The obtained results demonstrate the suitability of the followed methodology in the sense that it allows the achievement of efficient FPGA and GPU implementations able to cope with the stringent requirements imposed by hyperspectral imaging systems.

GPU Algorithm for the Scaled Opposite-Spin (SOS) MP2 Energy Evaluation

<p>The most computationally intensive part of the SOS-MP2 algorithm for the calculation of the correlation energy [1], as executed in Q-Chem, is implemented for use in a graphical processing unit (GPU). Our approach adds new routines to the library initially developed by Aspuru-Guzik and co-workers [2], aiming at maximization of bandwidth and performance, by taking advantage of the asynchronous CPU-GPU communication capability of modern GPUs. These changes permit an almost six-fold acceleration in the correlation energy calculation of linear alkanes. This was achieved employing a NVIDIA Tesla K40C (Kepler) GPU and the Compute Unified Device Architecture (CUDA).</p>