Noelle Walsh’s Post

The benefits of AI in energy & sustainability far outweigh the costs - Microsoft is working to streamline the logistical footprint of AI while unlocking its potential to foster new science, advance new energy technology, and optimize energy system operations. AI is an essential tool in the race to address climate change.

🌍 The energy industry is evolving rapidly in its efforts to reduce carbon emissions, and Microsoft's powerful solutions like Microsoft Cloud for Sustainability, Microsoft Sustainability Manager, and Microsoft Fabric are playing a crucial role in this transformation.   These technologies enable companies to capture critical sustainability data, generate detailed reports, and take informed action based on the insights. From streamlining data management to improving reporting accuracy, Microsoft technology is helping organizations drive impact reduction and efficiently manage their sustainability efforts.   Check out this blog from Microsoft to learn how scalability, standardization, and security contribute to #sustainability: https://lnkd.in/drqNWf8m   #MicrosoftCloud #ESG #MicrosoftFabric #MicrosoftSustainabilityManager  

"Coal-generated energy in the US is set to stick around longer than previously expected, to fulfill increasing demand from datacenters... The impact of AI apparently isn't recent. Microsoft's carbon emissions increased almost 30 percent from 2020 to the middle of 2023, according to the firm's latest sustainability report." #datacenters #powergeneration #greenagenda #AI https://lnkd.in/dwzri9QG

In line with our commitment to responsible AI and our ambitious sustainability commitments, we’re determined to tackle this challenge so the world can harness the full benefits of AI. There are three areas where we’re deeply invested and increasing our focus. The first is optimizing datacenter energy and water efficiency. The second is advancing low-carbon materials, creating global markets to help advance sustainability across industries. And the third is improving the energy efficiency of AI and cloud services, empowering our customers and partners with tools for collective progress.

Microsoft’s #CarbonFootprint jumped by nearly 30% in 2023, with the construction of new data centres to meet growing cloud and #AI demands causing emissions increases across Scope 1, 2 and 3. The company's capital goods emissions (part of Scope 3) went up by almost 2 million tonnes, but even Scope 2 market-based emission rose as Microsoft's #RenewableEnergy procurement was insufficient to meet its skyrocketing power needs. (Since 2020, Microsoft’s energy consumption has more than doubled.) “Our challenges are in part unique to our position as a leading cloud supplier that is expanding its data centres," said #ChiefSustainabilityOfficer Melanie Nakagawa. Read the full analysis of #Microsoft's latest #SustainabilityReport in CSO Futures https://lnkd.in/ephbJnwB #CloudEmissions #GreenIT #EnergyTransition

At Microsoft, we recognize that AI is a powerful catalyst for advancing sustainability progress globally. As we navigate the promise of AI as an accelerant for sustainability solutions with the current realities of its resource intensity, we're taking diligent steps to solve the challenges: 1) Optimizing datacenter energy and water efficiency 2) Advancing low carbon materials 3) Improving the energy efficiency of AI and cloud services   Learn more about how we're taking action in Microsoft Chief Sustainability Officer Melanie Nakagawa's new blog. #MicrosoftSustainability #AIforGood #ClimateAction #ResponsibleTech

At Microsoft, we're pursuing our carbon negative goal through carbon reduction, carbon-free electricity and carbon removal. For carbon reduction, power efficiency and energy efficiency are fundamental to sustainability progress.   Learn how we’re advancing the sustainability of AI by bringing promising energy efficiency research out of the lab and into commercial operations. Visit aka.ms/DrivingAIEfficiency   #MSFTAdvocate #MSSustainability

Sustainability in tech is more achievable than ever—what’s your next step? Let’s make an impact together! 🌍 Reach out to explore how Deployflow can help reduce your cloud carbon footprint.

**The Power of AI (in Watts rather than impact)** Microsoft have released the second of their posts on sustainability by design for energy efficiency in AI. https://lnkd.in/evhJa4kT. This topic continues to rightly occupy minds, both in terms of overall datacentre impact on climate, and specifically regarding the impact of generative AI on that total impact. There was a recent article claiming that the major cloud providers emit 662% more than they claim. https://lnkd.in/eCNutUYE. The details of that report get pretty murky. It's really interesting how different the cloud providers are under that microscope. A lot of this is down to the purchase of "renewable energy certificates" (Recs) to offset local power consumption needs. I do feel that Microsoft's sustainability reports have been reasonably transparent about this gap, and it is clearly something they aim to close. I would certainly rather Recs were purchased as an offset than *not* purchasing them. In any case, I also feel that some of the sustainability innovation in the OAI and Microsoft stacks over the last year has been lost in the shuffle. On the one hand, you have the model efficiency improvements themselves, which are enormous, you have more capable smaller models, and there is a better set of practices for targeting use of larger models. Beyond that, we have these in-datacentre improvements. The first of these posts was more focused on macro improvements for workload optimisation and liquid cooling. Certainly if we can squeeze more our of each datacentre, we need fewer of them, and the raw materials environmental costs of building datacentres are one of the biggest problems Microsoft is dealing with in this space. It is highly emission intensive to build more. The second of these posts focuses more on the silicon. This is really interesting, although a lot of it isn't here yet. Microsoft are working with the rest of the industry to standardise new 4-bit and 6-bit data types for training and inference efficiency. Then there are other clocking and phasing improvements that can improve throughput by up to 2.35x for the same power/cost. All told, there is still a ton of headroom for innovation here. I'm not despondent that generative AI energy demands are such a spike right now. I feel that we will get this under control, as there are so many dimensions to optimisation efforts, and we've already seen so much improvement.

🌱 Microsoft's latest renewable energy leap is monumental! 🌟 They've signed *the* most substantial contract to date, skyrocketing their green energy capabilities. Through a partnership with Brookfield, they're set to add a whopping 10.5 gigawatts of clean power across the US and Europe! That's not just good news; it's a sustainable revolution! 🌞 Committing over $10 billion, this move propels them into a greener future while powering their expanding data centers. Think AI, cloud computing, and next-gen tech—all humming along on clean, renewable energy. This is a game changer, friends! Not just for Microsoft but for all of us keen on a sustainable future. With this commitment, they're also fostering an energy-secure, low-carbon economy. Bravo! 🚀 ✨ What could this mean for renewable energy trends worldwide? Will others follow this bold path? Let's chat below! 💬 #RenewableEnergy #Sustainability #Innovation