Dynatrace

SEPTEMBER 13, 2023

This is a set of best practices and guidelines that help you design and operate reliable, secure, efficient, cost-effective, and sustainable systems in the cloud. The framework comprises six pillars: Operational Excellence, Security, Reliability, Performance Efficiency, Cost Optimization, and Sustainability.

AWS 238

AWS Efficiency Azure Cloud 238

Dynatrace

JANUARY 31, 2024

But energy consumption isn’t limited to training models—their usage contributes significantly more. Model observability provides visibility into resource consumption and operation costs, aiding in optimization and ensuring the most efficient use of available resources.

Cache 197

Cache Azure Infrastructure Monitoring 197

ACM Sigarch

MAY 31, 2023

Heterogeneous and Composable Memory (HCM) offers a feasible solution for terabyte- or petabyte-scale systems, addressing the performance and efficiency demands of emerging big-data applications. even lowered the latency by introducing a multi-headed device that collapses switches and memory controllers. The recently announced CXL3.0

Latency 52

Latency Hardware Cache Architecture 52

VoltDB

APRIL 25, 2024

While Wi-Fi theoretically can achieve 5G-like speeds, it falls short in providing the consistent performance and reliability that 5G offers, including low latency, higher speeds, and increased bandwidth. This scarcity of available spectrum can lead to competition and high costs for organizations seeking to acquire licenses.

Network 52

Network IoT Wireless Infrastructure 52

The Morning Paper

JANUARY 30, 2020

Edge servers are the middle ground – more compute power than a mobile device, but with latency of just a few ms. The current system assumes an application specific regression model is available on the servers which can predict processing time given the current parameters of the job (e.g. in the cloud).

Mobile 104

Mobile Cloud Latency Games 104

The Morning Paper

OCTOBER 11, 2020

The CubeSat form-factor limits what can be packed into the device and how much power is available. Without higher-risk deployable solar arrays, a cubesat relies on surface-mounted solar panels to harvest energy. This results in peak available power of about 7.1W. So uplink data volume is on the order of kilobytes per pass.

Systems 125

Systems Latency Architecture Energy 125

The Morning Paper

JUNE 13, 2019

As we saw with the SOAP paper last time out, even with a fixed model variant and hardware there are a lot of different ways to map a training workload over the available hardware. The following figure highlights how just one of these variables, batch size, impacts throughput and latency on ResNet50. The paper is silent on this issue.

Hardware 58

Hardware Latency Serverless Energy 58