Related papers: Taurus: Lightweight Parallel Logging for In-Memory…
Shared virtual memory (SVM) is key in heterogeneous systems on chip (SoCs), which combine a general-purpose host processor with a many-core accelerator, both for programmability and to avoid data duplication. However, SVM can bring a…
Efficient large-scale inference of transformer-based large language models (LLMs) remains a fundamental systems challenge, frequently requiring multi-GPU parallelism to meet stringent latency and throughput targets. Conventional tensor…
While multi-GPU (MGPU) systems are extremely popular for compute-intensive workloads, several inefficiencies in the memory hierarchy and data movement result in a waste of GPU resources and difficulties in programming MGPU systems. First,…
Efficient execution of deep learning workloads on dataflow architectures is crucial for overcoming memory bottlenecks and maximizing performance. While streaming intermediate results between computation kernels can significantly improve…
Recent years have seen massive time-series data generated in many areas. This different scenario brings new challenges, particularly in terms of data ingestion, where existing technologies struggle to handle such massive time-series data,…
Compared to replication-based storage systems, erasure-coded storage incurs significantly higher overhead during data updates. To address this issue, various parity logging methods have been pro- posed. Nevertheless, due to the long update…
Modern deployment of large language models (LLMs) frequently involves both inference serving and continuous retraining to stay aligned with evolving data and user feedback. Common practices separate these workloads onto distinct servers in…
Emerging Non-Volatile Memories (NVMs) bring a unique challenge to the security community, namely persistent security. As NVM-based memories are expected to restore their data after recovery, the security metadata must be recovered as well.…
Recent advances in LLMs have outpaced the computational and memory capacities of edge platforms that primarily employ CPUs, thereby challenging efficient and scalable deployment. While ternary quantization enables significant resource…
There is increasing interest in using multicore processors to accelerate stream processing. For example, indexing sliding window content to enhance the performance of streaming queries is greatly improved by utilizing the computational…
The exponential growth of data-intensive machine learning workloads has exposed significant limitations in conventional GPU-accelerated systems, especially when processing datasets exceeding GPU DRAM capacity. We propose MQMS, an augmented…
Graph partitioning has long been seen as a viable approach to address Graph DBMS scalability. A partitioning, however, may introduce extra query processing latency unless it is sensitive to a specific query workload, and optimised to…
Parallel dataflow systems have become a standard technology for large-scale data analytics. Complex data analysis programs in areas such as machine learning and graph analytics often involve control flow, i.e., iterations and branching.…
The relational DBMS (RDBMS) has been widely used since it supports various high-level functionalities such as SQL, schemas, indexes, and transactions that do not exist in the O/S file system. But, a recent advent of big data technology…
The rising demand for generative large language models (LLMs) poses challenges for thermal and power management in cloud datacenters. Traditional techniques often are inadequate for LLM inference due to the fine-grained, millisecond-scale…
The increasing adoption of blockchain technology has led to a growing demand for higher transaction throughput. Traditional blockchain platforms, such as Ethereum, execute transactions sequentially within each block, limiting scalability.…
Agentic systems solve complex tasks by coordinating multiple agents that iteratively reason, invoke tools, and exchange intermediate results. To improve robustness and solution quality, recent approaches deploy multiple agent teams running…
State-of-the-art distributed in-memory datastores (FaRM, FaSST, DrTM) provide strongly-consistent distributed transactions with high performance and availability. Transactions in those systems are fully general; they can atomically…
Despite the enormous success of Hamiltonian Monte Carlo and related Markov Chain Monte Carlo (MCMC) methods, sampling often still represents the computational bottleneck in scientific applications. Availability of parallel resources can…
Transactional data structure libraries (TDSL) combine the ease-of-programming of transactions with the high performance and scalability of custom-tailored concurrent data structures. They can be very efficient thanks to their ability to…