Earticle

The exciting advancement related to the “modeling of digital human” in terms of a computational phantom for radiation dose calculations has to do with the latest hype related to deep learning. The advent of deep learning or artificial intelligence (AI) technology involving convolutional neural networks has brought an unprecedented level of innovation to the field of organ segmentation. In addition, graphics processing units (GPUs) are utilized as boosters for both real-time Monte Carlo simulations and AI-based image segmentation applications. These advancements provide the feasibility of creating three-dimensional (3D) geometric details of the human anatomy from tomographic imaging and performing Monte Carlo radiation transport simulations using increasingly fast and inexpensive computers. This review first introduces the history of three types of computational human phantoms: stylized medical internal radiation dosimetry (MIRD) phantoms, voxelized tomographic phantoms, and boundary representation (BREP) deformable phantoms. Then, the development of a person-specific phantom is demonstrated by introducing AI-based organ autosegmentation technology. Next, a new development in GPU-based Monte Carlo radiation dose calculations is introduced. Examples of applying computational phantoms and a new Monte Carlo code named ARCHER (Accelerated Radiation- transport Computations in Heterogeneous EnviRonments) to problems in radiation protection, imaging, and radiotherapy are presented from research projects performed by students at the Rensselaer Polytechnic Institute (RPI) and University of Science and Technology of China (USTC). Finally, this review discusses challenges and future research opportunities. We found that, owing to the latest computer hardware and AI technology, computational human body models are moving closer to real human anatomy structures for accurate radiation dose calculations.

GPU들은 다수의 병렬 연산 처리장치를 가지고 있기 때문에 높은 병렬성을 보여준다. 뛰어난 연산 자원을 활용하기 위해 서 최신 GPU들을 범용 응용프로그램에 활용하는 GPGPU 기술이 개발되었다. 하지만 GPU가 높은 연산 처리 능력을 가지 고 있음에도 불구하고, 메모리 자원의 제한으로 인해 최대 성능을 보여주지 못한다. 그러므로 GPU의 장점을 충분히 활용 하기 위해서는 뛰어난 성능의 메모리 구조가 필요하다. 본 논문에서는 뛰어난 성능의 메모리 구조를 개발하기 위한 연구의 당위성을 보이기 위해서, 이상적인 메모리 구조를 가진 GPU가 일반적인 메모리 구조를 가진 GPU에 비해 월등히 나은 성 능을 보이는 것을 정량적으로 측정하여 보여주고자 한다. 실험 결과에 따르면, 이상적인 메모리 구조를 가진 GPU는 일반 적인 메모리 구조를 가진 GPU에 비해 높은 성능을 보인다. 뿐만 아니라, 실험 결과는 응용프로그램에 따라 성능 향상 폭 이 상이한 것도 보여주고 있다.

Recent Graphics Processing Units(GPUs) have many computational resources to provide high parallelism. Moreover, GPUs can execute general purpose applications to exploit high computational power. However, a GPU does not accomplish peak performance due to limited memory system even though a GPU has high throughput. Therefore, memory architecture providing high performance should be supported. In this paper, we show that the performance of the GPU with ideal memory system is much higher than the performance of the GPU with conventional memory system which indicates the effects of limited memory resources in order to prove that studying memory system on the GPU is very important. According to our simulation results, the GPU with ideal memory system shows better performance than the GPU with conventional memory system. Moreover, experimental results also show the performance depending on benchmarks. Therefore, characteristics of applications should be considered for the research focusing on the performance of memory system on the GPU.

천리안위성은 2010년 6월 27일에 발사되어 성공적으로 In Orbit Test (IOT)를 수행하고 있다. 천리안 위성을 지상에서 컨트롤 하기 위하여 ETRI 에서는 위성관제시스템을 개발하였으며, 현재 KARI에서 위성관제시스템을 운영중이다. 위성관제시스템의 일부인 임무계획 시스템은 기상/해양 이미지 촬영에 관한 임무요청, 위성체 기동 요청, 각동 이벤트 등을 종합하여 충돌 없는 임무스케줄을 만들어내게 되는데 이에 복잡한 스케줄링 기법이 요구된다. 천리안 위성의 임무 스케줄링 기법은 CPU 연산을 기본으로 하고 있으나, 이 논문에서는 Graphics Processing Units(GPU) 를 통한 임무 스케줄링 기법의 적용에 따르는 고려사항을 설명한다. 그리고 CPU 기반의 임무 스케줄링 기법과 GPU 기반의 임무 스케줄링 기법의 장단점을 분석한다.

Load forecasting has always been essential to the operation and planning of power systems in deregulated electricity markets. Various methods have been proposed for load forecasting, and the neural network is one of the most widely accepted and used techniques. However, to obtain more accurate results, more information is needed as input variables, resulting in huge computational costs in the learning process. In this paper, to reduce training time in multi-layer perceptron-based short-term load forecasting, a graphics processing unit (GPU)-based computing method is introduced. The proposed approach is tested using the Korea electricity market historical demand data set. Results show that GPU-based computing greatly reduces computational costs.

This paper presents a study on a high-performance design for a block cipher algorithm implemented on modern many-core graphics processing units (GPUs). The recent emergence of VLSI technology makes it feasible to fabricate multiple processing cores on a single chip and enables general-purpose computation on a GPU (GPGPU). The GPU strategy offers significant performance improvements for all-purpose computation and can be used to support a broad variety of applications, including cryptography. We have proposed an efficient implementation of the encryption/decryption operations of a block cipher algorithm, SEED, on off-the-shelf NVIDIA many-core graphics processors. In a thorough experiment, we achieved high performance that is capable of supporting a high network speed of up to 9.5 Gbps on an NVIDIA GTX285 system (which has 240 processing cores). Our implementation provides up to 4.75 times higher performance in terms of encoding and decoding throughput as compared to the Intel 8-core system.

Currently mobile services are on the verge of full commercialization ahead of 5G mobile communication (5G). The first goal could be to preempt the 5G market through realistic media services utilizing VR (Virtual Reality) and AR (Augmented Reality) technologies that users can most easily experience. Basically this movement is based on the advanced development of smart devices and high quality graphics processing computing power of mobile application processors. Accordingly, the importance of mobile GPUs is emerging and the most concern issue becomes a model for predicting the power and performance for smooth operation of high quality mobile contents. In many cases, the performance of mobile GPUs has been introduced in terms of power consumption of mobile GPUs using dynamic voltage and frequency scaling and throttling functions for power consumption and heat management. This paper introduces several studies of mobile GPU performance prediction model with user-friendly methods not like conventional power centric performance prediction models.

소비전력 증가와 같은 문제점들로 인하여, 마이크로프로세서만으로는 컴퓨팅 시스템의 성능을 향상시키기 점점 어려워지고 있다. 이와 같은 상황에서, 대용량 병렬 연산에 특화된 그래픽 처리 장치를 활용하여 중앙 처리 장치가 담당하던 범용 작업을 수행하게 하는 범용 그래픽 처리 장치 기술이 컴퓨터 시스템의 성능을 개선시킬 수 있는 방안으로 주목을 받고 있다. 하지만, 그래픽스 관련 응용프로그램과 범용 응용프로그램의 특징은 매우 상이하기 때문에, 그래픽 처리 장치가 범용 응용프로그램을 수행하는 경우에는 많은 제약 사항으로 인하여 자신의 뛰어난 연산 자원을 활용하지 못하는 실정이다. 일반적으로 그래픽스 관련 응용프로그램에 비해 범용 응용프로그램은 메모리를 매우 많이 요청하기 때문에 범용 그래픽 처리 장치 기술을 효율적으로 활용하기 위해서는 메모리 설계가 매우 중요하다. 특히, 긴 접근 시간을 요구하는 외부 메모리 요청은 성능에 큰 오버헤드이다. 그러므로 외부 메모리로의 접근 횟수를 줄일 수 있는 다중 레벨 캐쉬 구조를 효율적으로 활용할 수 있다면, GPU의 성능은 크게 향상 될 것이 분명하다. 본 논문에서는 다중 레벨 캐쉬 구조에 따른 그래픽 처리 장치의 성능을 다양한 벤치마크 프로그램을 통하여 정량적으로 분석하고자 한다.

Even though the performance of microprocessor is improved continuously, the performance improvement of computing system becomes hard to increase, in order to some drawbacks including increased power consumption. To solve the problem, general-purpose computing on graphics processing units(GPGPUs), which execute general-purpose applications by using specialized parallel-processing device representing graphics processing units(GPUs), have been focused. However, the characteristics of applications related with graphics is substantially different from the characteristics of general-purpose applications. Therefore, GPUs cannot exploit the outstanding computational resources sufficiently due to various constraints, when they execute general-purpose applications. When designing GPUs for GPGPU, memory system is important to effectively exploit the GPUs since typically general-purpose applications requires more memory accesses than graphics applications. Especially, external memory access requiring long latency impose a big overhead on the performance of GPUs. Therefore, the GPU performance must be improved if hierarchical memory architecture which can reduce the number of external memory access is applied. For this reason, we will investigate the analysis of GPU performance according to hierarchical cache architectures in executing various benchmarks.

효율적인 시간 영역 3차원 파동 전파 모델링을 위해 그래픽 프로세서를 사용하였다. 그래픽 프로세서는 대규모 병렬 연산을 위한 프로세서로, 그래픽 프로세서를 효율적으로 이용하기 위해서는 계산 과정과 메모리 복사 과정을 최적화할 필요가 있다. 본 연구에서는 메모리 관리에 초점을 맞추고 메모리 관리 방법에 따라 그래픽 프로세서를 이용한 프로그램의 성능이 어떻게 달라지는지 확인하였다. 또한 유한 차분법 차수와 속도 모델의 크기를 변화시켜가며 메모리 복사가 프로그램 성능에 미치는 영향을 시험하였다. 그 결과 3차원 파동장 전체를 복사하는 프로그램에서 메모리 관리가 유한 차분법 계산보다 큰 비중을 차지함을 알 수 있었다.

We used graphics processing units for an efficient time-domain 3D wave propagation modeling. Since graphics processing units are designed for massively parallel processes, we need to optimize the calculation and memory management to fully exploit graphics processing units. We focused on the memory management and examined the performance of programs with respect to the memory management methods. We also tested the effects of memory transfer on the performance of the program by varying the order of finite difference equation and the size of velocity models. The results show that the memory transfer takes a larger portion of the running time than that of the finite difference calculation in programs transferring whole 3D wavefield.

In this paper, we propose a parallel approximate string matching algorithm with k-mismatches for multiple fixed-length patterns (PMASM) in DNA sequences. PMASM is developed from parallel single pattern approximate string matching algorithms to effectively calculate the Hamming distances for multiple patterns with a fixed-length. In the preprocessing phase of PMASM, all target patterns are binary encoded and stored into a look-up memory. With each input character from the input string, the Hamming distances between a substring and all patterns can be updated at the same time based on the binary encoding information in the look-up memory. Moreover, PMASM adopts graphics processing units (GPUs) to process the data computations in parallel. This paper presents three kinds of PMASM implementation methods in GPUs: thread PMASM, block-thread PMASM, and shared-mem PMASM methods. The shared-mem PMASM method gives an example to effectively make use of the GPU parallel capacity. Moreover, it also exploits special features of the CUDA (Compute Unified Device Architecture) memory structure to optimize the performance. In the experiments with DNA sequences, the proposed PMASM on GPU is 385, 77, and 64 times faster than the traditional naive algorithm, the shift-add algorithm and the single thread PMASM implementation on CPU. With the same NVIDIA GPU model, the performance of the proposed approach is enhanced up to 44% and 21%, compared with the naive, and the shift-add algorithms.