Phaidra

As we approach the end of the year, operations leaders everywhere finalize plans for improvement activities in the coming year. As compute demand continues to sky-rocket, it's more important than ever for data center teams to consider what tasks they could implement that will help with the (likely) tumultuous year ahead. Hopefully this can help anyone interested to think through what small or large improvements they could incorporate to help future-proof their data center. https://lnkd.in/eWKsZ8rb

Ben Affleck pretty much nails it here… “…what #AI is going to do is going to disintermediate the more laborious, less creative, and more costly aspects of filmmaking that will allow costs to come down…” This is the best and most appropriate application for AI - tasks that are complex, mundane and repetitive. Phaidra’s customers are simply offloading the level 1 “optimized operational control” thinking from their operations teams to a Virtual Plant Operator. This frees the team up for all the level 2 and beyond cognitive work, like strategic planning, customer/vendor onboarding and smart-hand maintenance all around the site. https://lnkd.in/eHQgxTTe

Thank you for the mention Virta Ventures - Cooling data centers in an energy efficient way remains a hot topic ;) Even with the next generation of data centers being planned and designed now, there is so much opportunity to dramatically improve the operational capabilities of existing sites without over-taxing the already spread-too-thin operations teams. Let #AI handle the mundane, repetitive and complex tasks like micro-managing a cooling system to be as efficient as possible. https://lnkd.in/g9w2wYnn

November's edition covers all the most relevant news in software-defined physical industries. + Palantir's heretical scaling playbook + The fight for a winning ‘world model’, AI which can understand the physical world + The startups solving the energy crisis for Big Tech + Waymo deploys end-to-end learning and raises $6BN (link in comments) cc Archetype AI Physical Intelligence zystem.io Waymo World Labs Skild AI Nabeel S. Qureshi Sarah Constantin Phaidra

"AI works best when it complements human expertise. By automating routine tasks, AI frees up human workers to focus on higher-level strategic operations." - Ramzi CHARIF, VP of Technical Operations - EMEA at VIRTUS Data Centres Ramzi has a great take on whether or not AI would reduce jobs in the data center operations space. AI control services are a tool to enhance the capabilities of an operations team, not replace the people on them. "By enhancing human capabilities, preserving institutional knowledge, and optimizing operations, AI has the potential to transform data centers from behind-the-scenes infrastructure into agile, future-ready hubs of innovation." https://lnkd.in/eTCztJHX

Start your week off listening to one of our co-founders and CTO, Vedavyas Panneershelvam (Veda) discuss our mission and how we are deploying AI powered control systems to achieve it!

This was a good read on the challenges that startups (including my own company, Phaidra) face in the data center industry: https://lnkd.in/geV5p_td The central challenge is that the data center industry is not an industry where you can "move fast and break things." It is the very definition of mission-critical. The barrier to adopting new technologies is rightfully high given the stakes. At the same time, the startups that do manage to cross the chasm will have an incredibly strong moat around their business — and plenty of expansion opportunities in a rapidly-growing industry. I'm bullish. The Phaidra team has spent the last 5+ years perfecting autonomous AI control technology in anticipation of this moment (and even longer before that at Google and DeepMind). We're now entering the scaling phase with many of the world's largest and most innovative data center companies. As the old adage goes: the best way to predict the future is to build it. There's no doubt in my mind that self-learning control systems is the future of the data center industry — and that of the industrial sector more broadly. The key question is: who will be the winners when the entire industry shifts en masse towards intelligent, self-learning facilities? History doesn't just favor the bold — it also favors the prepared. P.S. Shoutout to our awesome investor Sophie Bakalar of Collaborative Fund who's quoted in the article! They're exceptional folks to work with.

As the #AI boom continues to drive up demand for compute power, the leading #datacenter hyperscale companies have an enormous capability to influence the entire industry's current and future operational practices. Learn a bit more about the wonderful sustainability goals that have been set and the operational challenges in the face of achieving them in our latest blog. https://lnkd.in/eUfWycjm

“The growth is really outpacing the current infrastructure that we have. We need newer, better, faster ways to achieve the promise we have heard about AI.” - Sophie Bakalar Excellent quick read from TechCrunch on the emerging technologies (like Phaidra) that will help the #datacenter industry keep up with the demand boom driven by #AI applications. https://lnkd.in/eGj9D88y

The benefits promised by #artificialintelligence are grand, but it's certainly reasonable to believe whether or not the promises are even possible to achieve. Hope for achieving hyper-efficiency through #AI control is high, but trust in the systems that may deliver isn't growing, at least not yet. It's important that #operations leaders in the #datacenter space get to see an AI system in action with constraints set for extra caution. Uptime Institute, as usual, puts together an excellent survey of industry leaders on their sustainability practices and the promises (and fears) around new technologies that could help achieve the goals set - read more in the Executive Summary of the Global Data Center Survey for 2024 (link in the comments)