More Like this

1. Exploring How the NICE Cybersecurity Workforce Framework Aligns Cybersecurity Jobs with Potential Candidates

   Estes, Adriane C.; Kim, Dan J.; Yang, T. Andrew (August 2018, The 14th International Conference on Frontiers in Education: Computer Science and Computer Engineering (FECS'18). Las Vegas, USA.)

   We discuss the NICE Cybersecurity Workforce Framework (NCWF), and its role in aligning cybersecurity jobs with candidates. As a workforce development tool, the NCWF can contribute to better retention, reduced new hire training, and cybersecurity education development. The effectiveness of the NCWF, however, requires discretion from hiring managers, academics, and job seekers. Through skills mapping and calibration, the NCWF helps to identify and resolve skill deficiencies; as a framework of core competencies for cybersecurity jobs, the NCWF helps employers to write job descriptions understood by applicants. We first review the NCWF, and then explain how it may enable mapping between jobs and qualifications. We also discuss the effects of job mapping on organizations and candidates, and its long-term benefits. 

   more » « less
2. PAL: A Variability-Aware Policy for Scheduling ML Workloads in GPU Clusters

   Jain, Rutwik; Tran, Brandon; Chen, Keting; Sinclair, Matthew D; Venkataraman, Shivaram (November 2024, ACM)

   Large-scale computing systems are increasingly using accelerators such as GPUs to enable peta- and exa-scale levels of compute to meet the needs of Machine Learning (ML) and scientific computing applications. Given the widespread and growing use of ML, including in some scientific applications, optimizing these clusters for ML workloads is particularly important. However, recent work has demonstrated that accelerators in these clusters can suffer from performance variability and this variability can lead to resource under-utilization and load imbalance. In this work we focus on how clusters schedulers, which are used to share accelerator-rich clusters across many concurrent ML jobs, can embrace performance variability to mitigate its effects. Our key insight to address this challenge is to characterize which applications are more likely to suffer from performance variability and take that into account while placing jobs on the cluster. We design a novel cluster scheduler, PAL, which uses performance variability measurements and application-specific profiles to improve job performance and resource utilization. PAL also balances performance variability with locality to ensure jobs are spread across as few nodes as possible. Overall, PAL significantly improves GPU-rich cluster scheduling: across traces for six ML workload applications spanning image, language, and vision models with a variety of variability profiles, PAL improves geomean job completion time by 42%, cluster utilization by 28%, and makespan by 47% over existing state-of-the-art schedulers. 

   more » « less
3. Innovation, Network Capabilities, and Sustainable Development of Regional Economies in China

   https://doi.org/10.3390/su11174770  

   Li, Dandan; Wei, Yehua Dennis; Miao, Changhong; Wu, Yangyi; Xiao, Weiye (September 2019, Sustainability)

   This paper studies the relationships between network capabilities and innovation development in the context of two types of innovation networks: scientific knowledge networks (SKN) and technological knowledge networks (TKN). Focusing on two types of network capabilities, namely acquisition capability and control capability, the paper uses spatial regime models to compare the impacts of multiple factors on different spatial regimes. The main conclusions are the following. First, as regards SKN, the political-administrative hierarchy has shaped the spatial evolution of acquisition capacity, forming a pattern consisting of three dominant cities (Beijing, Shanghai, Nanjing), three subsidiary cities (Guangzhou, Hangzhou, Wuhan), and multiple lesser centers (Tianjin, Chengdu, Xi’an). Moreover, high control capability cities are mainly clustered in the coastal areas, specifically, one monocentric city (Beijing) and two polycentric metropolises (Shanghai, Wuhan). Second, for TKN, cities with high acquisition and control also are mainly found in coastal areas, with Shanghai and Beijing dominating network capabilities. The model’s analysis confirms the positive effect of network capabilities on innovation development, especially in scientific knowledge networks, and the driver for regional innovation development appears to have shifted from global pipeline (globalization) to local buzz (localized talents). This paper concludes with suggestions regarding network capabilities’ potential to reduce regional inequality and achieve sustainable development of regional economies. 

   more » « less
4. Sojourn Time Minimization of Successful Jobs

   https://doi.org/10.1145/3561074.3561083  

   Yao, Yuan; Paolieri, Marco; Golubchik, Leana (August 2022, ACM SIGMETRICS Performance Evaluation Review)

   Due to a growing interest in deep learning applications [5], compute-intensive and long-running (hours to days) training jobs have become a significant component of datacenter workloads. A large fraction of these jobs is often exploratory, with the goal of determining the best model structure (e.g., the number of layers and channels in a convolutional neural network), hyperparameters (e.g., the learning rate), and data augmentation strategies for the target application. Notably, training jobs are often terminated early if their learning metrics (e.g., training and validation accuracy) are not converging, with only a few completing successfully. For this motivating application, we consider the problem of scheduling a set of jobs that can be terminated at predetermined checkpoints with known probabilities estimated from historical data. We prove that, in order to minimize the time to complete the first K successful jobs on a single server, optimal scheduling does not require preemption (even when preemption overhead is negligible) and provide an optimal policy; advantages of this policy are quantified through simulation. Related Work. While job scheduling has been investigated extensively in many scenarios (see [6] and [2] for a survey of recent result), most policies require that the cost of waiting times of each job be known at scheduling time; in contrast, in our setting the scheduler does not know which job will be the K-th successful job, and sojourn times of subsequent jobs do not contribute to the target metric. For example, [4, 3] minimize makespan (i.e., the time to complete all jobs) for known execution times and waiting time costs; similarly, Gittins index [1] and SR rank [7] minimize expected sojourn time of all jobs, i.e., both successfully completed jobs and jobs terminated early. Unfortunately, scheduling policies not distinguishing between these two types of jobs may favor jobs where the next stage is short and leads to early termination with high probability, which is an undesirable outcome in our applications of interest. 

   more » « less
5. Non‐competes, business dynamism, and concentration: Evidence from a Florida case study

   https://doi.org/10.1111/jems.12349  

   Kang, Hyo; Fleming, Lee (April 2020, Journal of Economics & Management Strategy)

   Abstract Most research on non‐competes has focused on employees; here we study how non‐competes affect firm location choice, growth, and consequent regional concentration, using Florida's 1996 legislative change that eased restrictions on their enforcement. Difference‐in‐differences models show that following the change, establishments of large firms were more likely to enter Florida; they also created a greater proportion of jobs and increased their share of employment in the state. Entrepreneurs or establishments of small firms, in contrast, were less likely to enter Florida following the law change; they also created a smaller proportion of new jobs and decreased their share of employment. Consistent with these location and job creation dynamics, regional business concentration increased following the law change in Florida. Nationwide cross‐sections demonstrate consistent correlations between state‐level non‐compete enforcement and the location, employment, and concentration dynamics illustrated in Florida. 

   more » « less