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 14 of 14 for “"tail latency"”.

  1. Achieving high CPU efficiency and low tail latency in datacenters

    … On the one hand, applications demand low latency-on the order of microseconds-in order to respond quickly to user requests. On the other hand, datacenter operators require high CPU efficiency in order to reduce operating costs. Unfortunately, today's systems do a poor job of providing low …

    mit Repository record for Achieving high CPU efficiency and low tail latency in datacenters (opens in a new tab)

  2. New techniques to lower the tail latency in stream processing systems

    … of effort has been made to reduce the average latency of stream processing systems, how to shorten their tail latency has received little attention. This thesis presents a series of novel techniques for reducing the tail latency in stream processing systems like Apache Storm. Concretely, we …

    uiuc Repository record for New techniques to lower the tail latency in stream processing systems (opens in a new tab)

  3. Mitigating Compute Congestion for Low Latency Datacenter RPCs

    Latency-sensitive applications in recent datacenter workloads, such as interactive machine learning inference, high-frequency algorithm trading, cloud gaming, and interactive AR/VR applications impose stringent latency requirements. These applications heavily rely on low-latency RPCs as an …

    mit Repository record for Mitigating Compute Congestion for Low Latency Datacenter RPCs (opens in a new tab)

  4. Optimizing interactive analytics engines for heterogeneous clusters

    … Compared to Getafix, Getafix-H improves the tail latency by 18% and reduces memory usage by up to 27% (2-3X improvement over Scarlett). In presence of stragglers, Getafix-H improves tail latency by 55% and reduces memory usage by upto 20% compared to Getafix. Getafix-H enables sysadmins to …

    uiuc Repository record for Optimizing interactive analytics engines for heterogeneous clusters (opens in a new tab)

  5. Towards SLO-aware Resource Scheduling for Serverless Inference Workloads

    … critical factors such as response time and tail latency. Additionally, Python's Global Interpreter Lock (GIL) poses challenges for parallel computing in high-request traffic scenarios. This thesis addresses the need for efficient and cost-effective Machine Learning (ML) inference …

    vt Repository record for Towards SLO-aware Resource Scheduling for Serverless Inference Workloads (opens in a new tab)

  6. UniNet: accelerating the container network data plane in IaaS clouds

    … SmartNICs for enhanced performance and reduced latency, and (3) instituting an isolated control plane that separates VM- and container-level rule insertions, making it tenant-accessible. UniNet boosts CNI throughput by 7.08x on average, cuts tail latency by 41.6%, and reduces CPU usage by up to …

    uiuc Repository record for UniNet: accelerating the container network data plane in IaaS clouds (opens in a new tab)

  7. Centralized performance control for datacenter networks

    … several properties, including low median and tail latency, high utilization (throughput), and congestion (loss) avoidance. Current datacenter networks inherit the principles that went into the design of the Internet, where packet transmission and path selection decisions are distributed among …

    mit Repository record for Centralized performance control for datacenter networks (opens in a new tab)

  8. Implementing accelerated key-value store: From SSDs to datacenter servers

    … and hardware techniques that provides bounded tail latency and design flexibility. PinK prototype reduces the read and 99th percentile latency by 22% and improves read throughput by 44% compared to LightStore prototype. The PinK prototype showed 42-73% better latency and 37% better throughput …

    mit Repository record for Implementing accelerated key-value store: From SSDs to datacenter servers (opens in a new tab)

  9. Congestion Control in Highly Variable Networks

    … advocate designing separate feedback mechanisms tailored specifically to the nuances of each network environment. Understanding how conditions are varying in each environment can help us unravel what kind of information about the network conditions can improve adaption to such variations. …

    mit Repository record for Congestion Control in Highly Variable Networks (opens in a new tab)

  10. A hardware and software architecture for efficient datacenters

    … low efficiency stems from two sources. First, latency-critical applications, which form the backbone of user-facing, interactive services, need guaranteed low response times, often a few tens of milliseconds or less. By contrast, current systems are architected to maximize long-term, average …

    mit Repository record for A hardware and software architecture for efficient datacenters (opens in a new tab)

  11. Squeezing the most benefit from network parallelism in datacenters

    … the network into symmetric components. Using a detailed switch hardware model, we simulate DRILL and show it outperforms recent edge-based load balancers particularly in the tail latency under heavy load, e.g., under 80% load, it reduces the 99.99th percentile of flow completion times of Presto …

    uiuc Repository record for Squeezing the most benefit from network parallelism in datacenters (opens in a new tab)

  12. Latency Tradeoffs in Distributed Storage Access

    … a number of new challenges, including the low latency handling of such data and ensuring that the network providing access to the data does not become the bottleneck. The traditional relational model is not well suited for efficiently storing and retrieving unstructured and semi-structured …

    temple Repository record for Latency Tradeoffs in Distributed Storage Access (opens in a new tab)

  13. Improving the end-to-end latency of datacenter applications using coordination across application components

    … lead to significant delays in their end-to-end latency. However, the organizations running these applications have strict requirements on this latency as it directly affects their revenue and operational costs. Addressing this problem, the goal of this dissertation is to develop scheduling and …

    uiuc Repository record for Improving the end-to-end latency of datacenter applications using coordination across application components (opens in a new tab)