We were pleased to welcome Minister @StephenMorganMP who visited our medical researchers and LifeLab facility to see how we are helping to improve children’s long-term health prospects.
Last week we enjoyed hosting colleagues from universities in the Netherlands @tweetsunl to discuss the opportunities and challenges of using generative AI in education.
Data centers in colleges and universities are crucial for supporting the extensive technological infrastructure required for modern education and research. These centers house critical servers and storage systems that manage vast amounts of data, ensuring reliable access to academic resources, administrative applications, and communication networks. They enable the secure storage and processing of sensitive information, including student records, faculty research, and institutional data.
Moreover, data centers facilitate advanced research by providing the computational power needed for data-intensive studies in fields like bioinformatics, climate science, and artificial intelligence. They support virtual learning environments and online course management systems, essential for the increasingly prevalent hybrid and online education models. Efficient data centers also contribute to campus sustainability goals by optimizing energy use through modern, eco-friendly technologies.
Additionally, robust data center infrastructure enhances the university’s ability to attract top-tier faculty and students by demonstrating a commitment to cutting-edge technology and resources. They also play a vital role in disaster recovery and business continuity, ensuring that educational and administrative functions can resume quickly after disruptions. Overall, data centers are integral to the academic mission, operational efficiency, and strategic growth of colleges and universities.
We have followed development of the technical standards that govern the success of these “installations” since 1993; sometimes nudging technical committees — NFPA, IEEE, ASHRAE, BICSI and UL. The topic is vast and runs fast so today we will review, and perhaps respond to, the public consultations that are posted on a near-daily basis. Use the login credentials at the upper right of our home page.
The National Institute of Standards and Technology seeks information to assist in carrying out several of its responsibilities under the Executive order on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence issued on October 30, 2023. Among other things, the E.O. directs NIST to undertake an initiative for evaluating and auditing capabilities relating to Artificial Intelligence (AI) technologies and to develop a variety of guidelines, including for conducting AI red-teaming tests to enable deployment of safe, secure, and trustworthy systems.
Attention Is All You Need | Authors: Ashish Vaswani et al. (2017). This groundbreaking paper introduced the Transformer architecture, replacing recurrent layers with self-attention mechanisms to enable parallelizable, efficient sequence modeling. It laid the foundational blueprint for all subsequent LLMs, revolutionizing natural language processing by capturing long-range dependencies without sequential processing.
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding | Authors: Jacob Devlin et al. (2018). BERT pioneered bidirectional pre-training via masked language modeling, allowing models to understand context from both directions. As an encoder-only Transformer, it achieved state-of-the-art results on 11 NLP tasks and established the pre-training/fine-tuning paradigm that underpins bidirectional LLMs like those in search and classification.
Training Compute-Optimal Large Language Models | Authors: Jordan Hoffmann et al. (2022). Known as the Chinchilla paper, it revealed that optimal LLM performance requires balanced scaling of model size and data volume (e.g., 70B parameters trained on 1.4T tokens outperform larger models with less data). This shifted research toward data-efficient training, influencing efficient LLM development.
Did you know? If you’ve seen clocks advertised to consumers as “atomic clocks,” those are actually listening to NIST radio stations’ time signals so they can count the seconds accurately. pic.twitter.com/hTTO0smikl
— National Institute of Standards and Technology (@NIST) January 31, 2024
Introduction. [Abstract]. The rapid growth of data centers, with their enormous energy and water demands, necessitates targeted policy interventions to mitigate environmental impacts and protect local communities. To address these issues, states with existing data center tax breaks should adopt sustainable growth policies for data centers, mandating energy audits, strict performance standards, and renewable energy integration, while also requiring transparency in energy usage reporting. “Renewable energy additionality” clauses should ensure data centers contribute to new renewable capacity rather than relying on existing resources. If these measures prove insufficient, states should consider repealing tax breaks to slow unsustainable data center growth. States without tax breaks should avoid such incentives altogether while simultaneously implementing mandatory reporting requirements to hold data centers accountable for their environmental impact. Broader measures should include protecting local tax revenues for schools, regulating utility rate hikes to prevent cost-shifting to consumers, and aligning data center energy demands with state climate goals to avoid prolonging reliance on fossil fuels.
PDF Page 570: Outlets for educational occupancies (STDMi comment)
PDF Page 52: Demand factors for schools (Definition of schools/colleges should correlate with ICC and ASHRAE occupancies — our historical claim and proposals).
PDF Page 539: “Meeting rooms” should recognize school occupancies according to ICC.
The University of Michigan has supported the voice of the United States education facility industry since 1993 — the second longest tenure of any voice in the United States. That voice has survived several organizational changes but remains intact and will continue its Safer-Simpler-Lower Cost-Longer Lasting advocacy on Code Panel 3 in the 2029 Edition.
Today, during our customary “Open Door” teleconference we will examine the technical concepts under the purview of Code Panel 1; among them:
Article 206 Signaling Circuits
Article 300 General Requirements for Wiring Methods and Materials
Article 590 Temporary Installations
Chapter 7 Specific Conditions for Information Technology
Chapter 9 Conductor Properties Tables
Public Input on the 2029 Edition will be received until April 9, 2026.
Page 522/523: 305.2 Group E, day care facilities for five or fewer children.
Page 624: Group E Security
Page 1440: Storm Shelters
§
Today at the usual hour we review a selection of global building codes and standards that guide best practice for safety, accessibility, and functionality for day care facilities; with special interest in the possibilities for co-locating square footage into the (typically) lavish unused space in higher education facilities.
Use the login credentials at the upper right of our home page.
Today we review the codes and standards that apply to the instructional and research facilities that support nursing science and practice. There is no single organization with a best practice catalog as there are in other disciplines we follow. Best practice is inspired by the inherited wisdom of practitioners, faculty and students who work alongside other members of healthcare provider teams.
Use the login credentials at the upper right of our home page.
Safety and sustainability for any facility begins with an understanding of who shall occupy the built environment and how. University settings, with mixed-use phenomenon arising spontaneously and temporarily, often present challenges. Educational communities are a convergent settings for families; day care facilities among them. First principles regarding occupancy classifications for day care facilities appear in Section 308 of the International Building Code, Institutional Group I; linked below:
The ICC Institutional Group I-4 classification includes buildings and structures occupied by more than five persons of any age who received custodial care for fewer than 24 hours per day by persons other than parents or guardian, relatives by blood, marriage or adoption, and in a place other than the home of the person cared far. This group includes both adult and child day care.
We maintain focus on child day care. Many educational communities operate child day care enterprises for both academic study and/or as auxiliary (university employee benefit) enterprises.
Princeton University Child Care Center
Each of the International Code Council code development groups fetch back to a shared understanding of the nature of the facility; character of its occupants and prospective usage patterns.
The 2024 revision of the International Building code is in production now. Ahead of the formal, market release of the Group A tranche of titles you can sample the safety concepts in play during this revision with an examination of the documents linked below:
Search on the terms “day care” and “daycare” in the link at the top of this page to get a sample of the prevailing concepts; use of such facilities as storm shelters, for example.
“The Country School” | Winslow Homer
We encourage our safety and sustainability colleagues to participate directly in the ICC Code Development process. We slice horizontally through the disciplinary silos (“incumbent verticals”) created by hundreds of consensus product developers every week and we can say, upon considerable authority that the ICC consensus product development environment is one of the best in the world. Privately developed standards (for use by public agencies) is a far better way to discover and promulgate leading practice than originating technical specifics from legislative bodies. CLICK HERE to get started. Contact Kimberly Paarlberg (kpaarlberg@iccsafe.org) for more information.
There are competitor consensus products in this space — Chapter 18 Day-Care Occupancies in NFPA 5000 Building Construction and Safety Code, for example; a title we maintain the standing agenda of our Model Building Code teleconferences. It is developed from a different pool of expertise under a different due process regime. See our CALENDAR for the next online meeting; open to everyone.
New update alert! The 2022 update to the Trademark Assignment Dataset is now available online. Find 1.29 million trademark assignments, involving 2.28 million unique trademark properties issued by the USPTO between March 1952 and January 2023: https://t.co/njrDAbSpwBpic.twitter.com/GkAXrHoQ9T
.jpg)
