Related papers: Architectural Impact on Performance of In-memory D…
Memory allocation, though constituting only a small portion of the executed code, can have a "butterfly effect" on overall program performance, leading to significant and far-reaching impacts. Despite accounting for just approximately 5% of…
Performance analysis has always been an afterthought during the application development process, focusing on application correctness first. The learning curve of the existing static and dynamic analysis tools are steep, which requires…
Today's big data clusters based on the MapReduce paradigm are capable of executing analysis jobs with multiple priorities, providing differential latency guarantees. Traces from production systems show that the latency advantage of…
The growing performance gap between multi-core CPUs and main memory necessitates hardware-aware software design paradigms. This study provides a comprehensive performance analysis of Data Oriented Design (DOD) versus the traditional…
A key need in different disciplines is to perform analytics over fast-paced data streams, similar in nature to the traditional OLAP analytics in relational databases i.e., with filters and aggregates. Storing unbounded streams, however, is…
This paper presents a stream-oriented architecture for structuring cluster applications. Clusters that run applications based on this architecture can scale to tenths of thousands of nodes with significantly less performance loss or…
BigBench is the new standard (TPCx-BB) for benchmarking and testing Big Data systems. The TPCx-BB specification describes several business use cases -- queries -- which require a broad combination of data extraction techniques including…
Time-evolving stream datasets exist ubiquitously in many real-world applications where their inherent hot keys often evolve over times. Nevertheless, few existing solutions can provide efficient load balance on these time-evolving datasets…
Multi-accelerator servers are increasingly being deployed in shared multi-tenant environments (such as in cloud data centers) in order to meet the demands of large-scale compute-intensive workloads. In addition, these accelerators are…
We describe a universal modeling approach for predicting single- and multicore runtime of steady-state loops on server processors. To this end we strictly differentiate between application and machine models: An application model comprises…
Sequential computation is well understood but does not scale well with current technology. Within the next decade, systems will contain large numbers of processors with potentially thousands of processors per chip. Despite this, many…
The need for modern data analytics to combine relational, procedural, and map-reduce-style functional processing is widely recognized. State-of-the-art systems like Spark have added SQL front-ends and relational query optimization, which…
This work elaborates on a High performance computing (HPC) architecture based on Simple Linux Utility for Resource Management (SLURM) [1] for deploying heterogeneous Large Language Models (LLMs) into a scalable inference engine. Dynamic…
Graphics processing units (GPUs) can improve deep neural network inference throughput via batch processing, where multiple tasks are concurrently processed. We focus on novel scenarios that the energy-constrained mobile devices offload…
Deploying DNNs on System-on-Chips (SoC) with multiple heterogeneous acceleration engines is challenging, and the majority of deployment frameworks cannot fully exploit heterogeneity. We present MATCHA, a unified DNN deployment framework…
Fully Homomorphic Encryption (FHE) is known to be extremely computationally-intensive, application-specific accelerators emerged as a powerful solution to narrow the performance gap. Nonetheless, due to the increasing complexities in FHE…
The growing disparity between CPU core counts and available memory bandwidth has intensified memory contention in servers. This particularly affects highly parallelizable applications, which must achieve efficient cache utilization to…
Large language model (LLM) decoding is a major inference bottleneck because its low arithmetic intensity makes performance highly sensitive to memory bandwidth. 3D-stacked near-memory processing (NMP) provides substantially higher local…
In some models of parallel computation, jobs are split into smaller tasks and can be executed completely asynchronously. In other situations the parallel tasks have constraints that require them to synchronize their start and possibly…
In this article, we investigate the impact of architectural parameters of array-based DNN accelerators on accelerator's energy consumption and performance in a wide variety of network topologies. For this purpose, we have developed a tool…