In any research computing (RC) environment, the role of the user support group is a combination of education, consulting, and outreach. As this role is the most public-facing of a research computing group's team, it is important to ensure an excellent level of support is provided to users. The Research Computing group at the University of Colorado (CURC) Boulder maintains a large-scale computing cluster with several hundred active users, among other services, and has done so since 2011. The user support group at CURC provides a variety of services intended to support these users. This paper describes those services the CURC group provides, as well as explores the various ways that these services have been improved in 2018-19. The impact on users is also assessed.

Advances in the volume, diversity, and complexity of research data and associated workflows requires enhanced capabilities to access, secure, reuse, process, analyze, understand, curate, share, and preserve data. To address this need at the regional level the University of Colorado, Colorado State University, and the University of Utah formed a "Cyberteam" in 2017 to provide cyberinfrastructure (CI) support to researchers at institutions in the Rocky Mountain Advanced Computing Consortium (RMACC) encompassing states across the Intermountain West. The Cyberteam is comprised of CI professionals across the three institutions who collaborate closely, sharing expertise and resources. Since its establishment, the Cyberteam has worked to broaden accessibility and options for computing, storage, and data publishing for RMACC researchers; enhance training on data- and workflow-oriented topics; improve engagement with researchers using CI; and better understand user needs and challenges. One key accomplishment has been the development of a series of focus group and survey instruments to achieve better understanding the CI needs and challenges of researchers across a diverse spectrum of disciplines. This paper provides an overview of the RMACC Cyberteam's objectives, accomplishments, challenges, and future direction.

Today's data centers contain large numbers of compute nodes that require substantial power, and therefore require a large amount of cooling resources to operate at a reliable temperature. The high power consumption of the computing and cooling systems produces extraordinary electricity costs, requiring some data center operators to be constrained by a specified electricity budget. In addition, the processors within these systems contain a large number of cores with shared resources (e.g., last-level cache), heavily affecting the performance of tasks that are co-located on cores and contend for these resources. This problem is only exacerbated as processors move to the many-core realm. These issues lead to interesting performance-power tradeoffs; by considering resource management in a holistic fashion, the performance of the computing system can be maximized while satisfying power and temperature constraints. In this work, the performance of the system is quantified as the total reward earned from completing tasks by their individual deadlines. By designing three resource allocation techniques, we perform a rigorous analysis on thermal, power, and co-location aware resource management using two different facility configurations, three different workload environments, and a sensitivity analysis of the power and thermal constraints. (C) 2017 Elsevier Inc. All rights reserved.

RMACC Summit is a heterogeneous supercomputer cluster with an aggregate floating point performance of 379 TFLOPS (Rmax, as currently configured) that provides about 85 million core-hours/year to researchers from institutions participating in the Rocky Mountain Advanced Computing Consortium (RMACC). The development of Summit was a collaborative effort toward specifying a system that meets the needs of researchers at multiple universities, and included implementation and testing of several new technologies. We discuss our experiences in creating and maintaining a successful ongoing collaboration between the two universities that are RMACC Summit's primary operators, and consider both the technical and support challenges of extending that collaboration to other regional users.

Operators of high-performance computing (HPC) facilities face conflicting trade-offs between the operating temperature of the facility, reliability of compute nodes, energy costs, and computing performance. Intelligent management of the HPC facility typically involves taking a proactive approach by predicting the thermal implications of allocating tasks to different cores around the facility. This offers the benefit of operating the HPC facility at a hotter CRAC temperature while avoiding hotspots. However, such an approach can be a time-consuming process that requires complicated air flow models to be calculated for every mapping decision. We propose a framework in which offline analysis is used to assist an online resource manager by predicting the thermal implications of mapping a given workload. The goal is to maximize the reward earned from completing tasks by their individual deadlines throughout the day, while adhering to a daily energy budget and temperature threshold constraints. We show that our proposed techniques can earn significantly greater reward than traditional load balancing and thermal management schemes.

Two experiments were conducted to assess the efficacy and duration of action of two dopaminergic compounds, pergolide and cabergoline, on daily prolactin secretion in geldings and on prolactin responses to a small dose of sulpiride over 10 days. In the first experiment, oral administration of 2 mg of pergolide was compared to a single injection of 2 mg of pergolide in a slow-release vehicle and a single injection of 5 mg of cabergoline in slow-release vehicle. Controls received vehicle only. All drug treatments reduced (P < .05) prolactin concentrations relative to that in controls but differed substantially in duration of action (oral pergolide approximately 6 hours or less, injected pergolide 6 to 24 hours, and injected cabergoline at least 6 days). In the second experiment, repeated small doses of sulpiride (2 mu g/kg of body weight intravenously) were used to stimulate prolactin release in mares, and the ability of seven daily injections of pergolide (2 mg each) and a single injection of cabergoline (5 mg) in slow-release vehicle to suppress this release were compared. Control mares receiving vehicle injections had robust prolactin responses to the sulpiride injections on all days of injection (days 1, 0, 1, 2, 3, 4, 6, 8, and 10 relative to treatment). Prolactin responses were muted (P < .05) by pergolide and cabergoline treatments on the first day of injection (day 0, 30 min after treatment) and were basically absent on days 1 to 8. The single injection of cabergoline continued to be suppressive through day 10, whereas mares previously treated with pergolide (through day 6) had begun to recover a prolactin response by day 10. We conclude that either daily 2-mg pergolide injections in slow-release vehicle or a single injection of 5 mg of cabergoline in slow-release vehicle is an effective way to apply dopaminergic activity to horses for approximately 7 to 10 days and may have application in the treatment of pituitary pars intermedia dysfunction in affected horses. (C) 2013 Elsevier Inc. All rights reserved.

Between February 15 and May 17, 2011, a total of 88 broodmares (10 maiden, 10 barren, and 68 foaling) maintained on pasture in southeast Texas were examined three times weekly (Tuesday, Thursday, Saturday) by transrectal palpation and ultrasonography. On Tuesday or Thursday, mares in estrus with uterine edema, a relaxed cervix, and a dominant follicle >= 34 mm in diameter were alternately assigned to treatment with the following: group (1) 2,500-unit human chorionic gonadotropin (hCG), intravenous; group (2) 1.0-mg BioRelease Histrelin (Biorelease Technologies, Lexington, KY), intramuscular; or group (3) 0.5-mg BioRelease Histrelin, intramuscular. Ovulation was confirmed by ultrasonographic examination. The percentage of mares ovulating within 2 days appeared to be similar between maiden, barren, and foaling mares, so responses for all mares were totaled for analysis. A nonsignificant trend for higher ovulation rates within 2 days was noted for both dose rates of histrelin compared with hCG treatment (31/37, 84%; 34/37, 92%; and 33/36,92% for groups 1-3, respectively) (P = .45). Ovulatory responses appeared to improve for both products as the season progressed, yet no differences were detected between response rates to histrelin or hCG for any month (P >= .50). The use of 1.0- or 0.5-mg BioRelease Histrelin was found to be at least equally effective as hCG treatment for inducing ovulation within 2 days of treatment throughout the breeding season. (C) 2012 Elsevier Inc. All rights reserved.

Since 1966, exogenous progestins have been used in equine practice for pregnancy maintenance, estrous suppression, and control of erratic sexual behavior. This study was designed to investigate the use of a new compounded controlled-release progesterone preparation (BioRelease P4 LA 300) in early and late spring transitional mares. In the first experiment, the pharmacodynamic properties of the preparation were studied in five geldings. In the second experiment, the use of a single intramuscular injection (600 mg) was tested in 68 embryo-recipient mares maintained under natural photoperiod in the Southern Hemisphere. Experiment 1 demonstrated elevated serum concentrations of progesterone (>1 ng/mL) for 7.6 ± 2.2 days. In experiment 2, there was no effect of treatment in mares that were treated on September 18, independent of their follicular status at day of treatment (10 to 15 mm; 20 to 25 mm, respectively). When mares with a follicular size of 20 to 25 mm were treated on October 14, significantly more progestin-treated mares (10/12; 83%) ovulated between 10 and 24 days after treatment than untreated controls (3/12; 25%) (P < .05). Additionally, there was a trend in mares treated on October 14 for a shorter treatment to ovulation interval (mean ± SD, 18.6 ± 8.7 days) compared with untreated controls (mean ± SD, 26.7 ± 14.7 days) (P = .07). Administration of one single injection of long-acting progesterone is a simple and effective method of controlling the first ovulation of the season in late transitional mares.

Three experiments were conducted to test the efficacy of different doses of estradiol cypionate (ECP) and domperidone for inducing ovulation in seasonally anovulatory (January; Experiments 1 and 2) and transitional period (March; Experiment 3) mares. In the first two experiments, mares in Kentucky and Louisiana were administered domperidone (3 g in biodegradable particles) alone or after pretreatment with 100 or 150 mg of ECP; another group received ECP, domperidone, and progesterone, and a fifth group received ECP and progesterone only (the latter two in Kentucky). Control mares in both states received no treatment. The proportion of mares ovulating within 35 days (for mares treated in January) was greater (P = .0002) for those receiving any combination of ECP plus domperidone relative to mares not receiving the combined treatment. Addition of progesterone to ECP plus domperidone did not enhance (P = .7) the response relative to the combination alone; domperidone by itself, or ECP plus progesterone, did not alter the response relative to controls (P > .24). Experiment 3 was conducted in Louisiana as a 2 × 3 factorial, with two doses of domperidone (1.5 or 3 g) and three doses of ECP (0, 75, or 150 mg). There was no main effect of domperidone or ECP dose; a greater proportion (P = .055) of mares receiving any combination of ECP plus domperidone ovulated in 21 days compared with those receiving no ECP. In conclusion, pretreatment with ECP before injection of domperidone 10 days later can be used to increase the proportion of mares ovulating early in the year; within the limits of the present experiments, there appears to be no difference in doses of ECP of 75, 100, or 150 mg, and no difference in doses of domperidone of 1.5 or 3 g.