Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 23 for “"SIMT"”.
-
Towards scalar synchronization in SIMT architectures
… which led to the research on synchronization in SIMT hardware. It then proposes instruction set and hardware changes that simplify the implementation of mutual exclusion when porting multiple-instruction, multiple data (MIMD) programs with synchronization to accelerators employing …
-
Architectural support for inter-thread synchronization in SIMT architectures
Single-Instruction Multiple-Threads (SIMT) architectures have seen widespread interest in accelerating data parallel applications. In the SIMT model, small groups of scalar threads operate in lockstep. Within each group, current SIMT implementations serialize the execution of threads that follow …
-
SIMT: A holistic framework for embedding simulation into the health care systems.
… a new framework, known as SIMulation Thinking (SIMT), to overcome the identified challenges and barriers. SIMT includes five key components: infrastructure, management, culture change, methodology and modelling. Whilst the SIMT framework presents the important elements that need to be considered …
-
Code generation and adaptive control divergence management for light weight SIMT processors
… (HARP) - a single instruction multiple thread (SIMT) architecture for integration into DRAM systems, particularly 3D memory stacks as a distinct processing layer to exploit the enormous internal memory bandwidth. However, this potential can only be realized with an optimizing compilation …
-
Evaluating the Cost-Benefit Tradeoffs of SIMT Control Mechanisms in Resource-Constrained GPUs
… tradeoffs associated with increasingly complex SIMT control mechanisms in a resource-constrained GPU core. Four design implementations are evaluated: a Base Tiny GPU without warp scheduling capability; a Warp Scheduler using round-robin warp selection; a Branch Divergence implementation …
-
A Hardware-Software Integrated Solution for Improved Single-Instruction Multi-Thread Processor Efficiency
… branching efficiency. Unlike current SIMT hardware branching architectures, this hardware-software solution allows programmers the ability to fine tune branching behavior for their application or allow the compiler to implement a generic software solution. To support a wide range of …
-
Performance-efficient mechanisms for managing irregularity in throughput processors
… adopt the single-instruction multiple-thread (SIMT) programming model that balances programmability and hardware efficiency. With SIMT, the programmer writes application code to be executed by scalar threads and each thread is supported with conditional branch and fine-grained load/store …
-
Uncertainty-Aware Adaptation for Simultaneous Machine Translation under Domain Shift
Simultaneous Machine Translation (SiMT) uses advancements in Natural Language Processing (NLP) to provide seamless, real-time communication between languages, while balancing translation quality and latency. However, these systems are highly sensitive to domain shift, where discrepancies between …
-
Vertical Integration in Open-Source Hardware-Software Co-Design for Accelerator Architecture Research
… with their Single Instruction Multiple Threads (SIMT) execution model, heavily leverage data-parallel multi-threading to maximize throughput at a relatively low energy cost. Over the past decade, GPGPU architecture research has primarily focused on cycle-level simulations that model the hardware …
-
Memory optimizations for hardware accelerated ray tracing on GPUs
… thread scheduling, showing that despite their SIMT efficiency benefits, they provide limited performance benefits for ray tracing workloads. The second work introduces a traversal algorithm leveraging treelets (subtrees of BVH tree), where threads fully traverse a treelet before moving to the …
-
On replay and hazards in graphics processing units
… employing a single-instruction, multiple thread (SIMT) execution model. Hazards are events that occur in the execution of a program which limit performance. We explore design tradeoffs in hazard handling. We find that in common architectures hazards that stall the pipeline cause other unrelated …
-
Domain-Specific Design on FPGA and NPU for scientific computing and On-Device LLM Inference
… relying on multithreading, SIMD, and SIMT execution, but increasingly encounter scalability limits imposed by the power and memory walls. In contrast, Field-Programmable Gate Arrays (FPGAs) and Neural Processing Units (NPUs) offer a high-efficiency alternative through dataflow-oriented …
-
Static Cost Estimation for Data Layout Selection on GPUs
… work, and perhaps even the design of new SIMT architectures. This thesis presents a comprehensive cost analysis for data layout selection on GPUs. We build our cost function based on knowledge of the GPU memory hierarchy, and develop an algorithm which enables researchers to perform …
-
Sparse LU Factorization for Large Circuit Matrices on Heterogenous Parallel Computing Platforms
… our implementation on multi-core, SIMD and SIMT architectures namely Intel Xeon CPU, Intel Xeon Phi coprocessor and an Nvidia GPU respectively, each using different programming models suited for the respective platform to determine the architecture best suited for parallelizing direct sparse …
-
Characterization and Exploitation of GPU Memory Systems
… all types of computation. The massively parallel SIMT architecture of the GPU can still be constraining in terms of achievable performance. GPU-based systems will typically only be able to achieve between 40%-60% of their peak performance. One of the major problems affecting this effeciency is the …
-
Batched Linear Algebra Problems on GPU Accelerators
… are all Single-Instruction Multiple-Thread (SIMT) architectures. Our solutions and methods are based on NVIDIA GPUs and can be extended to other accelerators, such as the Intel Xeon Phi and AMD GPUs based on OpenCL.</p>
-
From the mass-concentration relation of haloes to GPUs and into the web
… unbedingt notwendig in den Anwendungsprogrammen SIMT (Single Instruction Multiple Thread) und/oder SIMD (BSingle Instruction Multiple Data) Instruktionen zu nutzen, um die Rechenleistung der aktuellen und kommenden Generationen von Supercomputern so gut wie möglich auszunutzen zu können. Darüber …
-
Modeling and Analysis of Large-Scale On-Chip Interconnects
… parallel single-instruction multiple-thread (SIMT) based graphics processing unit (GPU) platforms to tackle power grid analysis with very promising performance. Our GPU based network analyzer is capable of solving tens of millions of power grid nodes in just a few seconds. Additionally, with …
-
Accelerating Data Parallel Applications via Hardware and Software Techniques
… with their Single-Instruction Multiple-Thread (SIMT) execution model offer massive parallelism that is combined with a relative ease of programming. The large amount and diversity of resources on the GPU is intended to ensure applications with different characteristics can achieve high …
-
Applications on emerging paradigms in parallel computing
… with thread-level shared memory and SIMT parallelism, to the higher levels of distributed memory parallelism as in supercomputers and clusters, and scaling them to large distributed systems as</p> <p>server farms and clouds. All together these form a large hierarchy of parallelism. …
Page 1 of 2