00:10
Hello and welcome to business leaders. It's time to prioritize sustainability. My name is John Brown. I'm a senior analyst for sustainability with the Enterprise Strategy group and I'm gonna talk a little bit today about what we're seeing in um sustainability. So pure storage and enterprise strategy group did some research on sustainability.
00:30
We learned quite a bit about how organizations are addressing sustainability in a practical way. But before we go too deeply into that, let me set a little background of where we are, how we got here and what the challenges are today first. And I'd like to remind folks of the importance of this point. Computing that is using computers to enhance
00:50
human capabilities is super beneficial part of the fabric of life at this point. Let me give you some perspective on what we couldn't do without them. First off, data analysis and modeling, leading to breakthroughs in fields like climate modeling, genomics, drug discovery, medical imaging efficiency, automation, increasing productivity and reducing costs in industries like manufacturing
01:14
and transportation and even in it where automation and digital transformation rule, ecommerce, 15.9% of retail according to the Saint Louis fed is ecommerce, the entire financial system, social media, online education, remote work movies and video games. All of these things are based on computing. We're not gonna give this up. We are getting more efficient.
01:43
Just like in cars, technology can help reduce emissions without destroying the benefit of transporting us. Removing environmental impact is analogous as removing costs and cost reduction in a product. It can be done. The hot new thing is artificial intelligence. Perhaps you've heard some hype about this. Perhaps you might even have a plan to implement in a recent ESG tech target study called
02:08
Generative A I. And it operations some 97% of it professionals said that their organization either has A I in production or has a pilot running or plans to have a pilot running within the next year. I've been in the business of observing it trends for more than 20 years. I've never seen anything with this level of near term adoption planned.
02:30
When cloud computing came out, there were skeptics and cloud followed the adoption curve that we're used to A I appears to be something different. What I'm saying is that if you have plans around artificial intelligence, you're in very good company. Artificial Intelligence promises to help us with everything that computers have
02:49
traditionally been helpful with. Automation is an obvious use case. But we expect a lot more to happen because of the technology. The current Achilles heel to A I is that it potentially uses a lot of energy in 2022 U Mas, the University of Massachusetts and one of my alma maters found that using a single training run of an LLM can emit
03:14
626,000 pounds of CO2. The equivalent of the lifetime emissions of five American cars. That is a lot of carbon emitting carbon is the opposite of being sustainable. So you can see that A I and the 97% of us who have plans to adopt it are facing a challenge. How do I reconcile my desire for A I with my desire for being sustainable?
03:42
So let's outline the problem we face and put ourselves back into reality. First point is that sustainability is a thing CS RD the corporate sustainability reporting directive. Um you know, is one if you have more than 500 employees and you're in a, this one applies to you, the U NFS S, the United Nations Forum on sustainability
04:06
standards, the GR I, the Global Reporting initiative, SAS B, the sustainability accounting Standards boards uh which is going through a name change to ISSB um you know, lots of regulations out there. If we look at adoption statistics, 71% of us place sustainability as a very high or above average corporate priority.
04:29
78% of organizations surveyed are subject to government sustainability mandates. 74% of organizations surveyed have a sustainability officer or an ESG executive in place and 74% also have published goals or commitments to becoming carbon neutral. If we look at what the vendors are doing to help uh meet those goals,
04:53
Microsoft, their aim is to remove more carbon dioxide from the atmosphere than it emits. By the year 2030 this would effectively reverse the company's historical emissions. And by 2050 Microsoft plans to remove all the carbon it has emitted since its founding in 1975. That is a big goal.
05:14
Mc JP Morgan Chase has committed to achieving carbon neutrality by 2030 aligning its financing portfolio with net zero emissions by 2050 Exxon. They aim to achieve net zero greenhouse gas emissions from their operated assets by 2050. When the largest non state owned oil producer commits to net zero, you've got some pretty clear evidence that sustainability is indeed mainstream.
05:44
In most cases, unless you have a big fleet of jets, it's your data center and your manufacturing that can be most impactful and helping your organization meet its carbon and energy objectives. Don't take my word for it. We asked about this in the study, we asked which business departments are or will be the most active in driving sustainability initiatives.
06:06
56% said it would be it, which was higher than those that said it was the procurement team or manufacturing. And note here if I ask manufacturing companies this, they rightfully say that manufacturing is number one, but everybody outside of manufacturing said it is the top driver here. Second point data centers use a lot of energy and artificial intelligence is coming.
06:34
And so there are some new problems ahead. We already talked about how many cars worth of admissions and A I training run is in 2023 in data from the carbon collective data centers consumed 7.4 gigawatts of power. One gigawatt of power is enough to power 876,000 homes for a year. The Hoover Dam, a big project produces two gigawatts of relatively clean power.
07:02
The largest coal fired power plants can produce four gigawatts of dirty power and solar. As of 2023 produced 5.6% of the country's electricity which was a big jump. Data centers consume 10 to 50 times more energy per amount of floor space than a typical commercial building. So if we're already using a huge amount of electricity and water for cooling and the
07:31
demand looks like it's going to increase, what are we to do? It's time for leaders. You watching this video to lead, start thinking about this. Now, ask yourself some questions, how can we integrate sustainability into our core business strategy rather than treating it as a separate initiative?
07:54
What is our supply chain sustainability profile and how can we improve it? How can we incorporate circular economy principles into our product design and manufacturing and consumption processes? What are the potential cost savings and revenue opportunities associated with sustainability initiatives. And on this point, our study has some very good
08:19
news. We found that over 50% of the organizations that implemented sustainability initiatives in it saw lower operational costs, lower capital expenditures, increased profitability than they had initially expected. Upside surprises from doing the right thing. That is actually noteworthy work with your utility suppliers. The power companies are incredibly important in
08:45
the sustainability equation, but they're also important for pure capacity reasons. If we know that A I is going to consume a lot of power 700 to 1000 watts for a single GP U. And that doesn't even include the cooling that is required. Will your power company have enough power to meet your needs? What are we gonna do if they don't look within, think about data center efficiency,
09:09
data centers have this concept of pue power usage efficiency. Pue equals total facility power divided by it, equipment, power and the result is how much power is being used to do the things you want to do versus things you need to do like overhead and cool. You want to have the lowest possible number, it's a ratio and you're looking for your pue to be as low as possible in a perfect world.
09:35
100% of the power going in would go to compute and none of it would go to cooling voltage modification, lighting fans and that sort of stuff. And in some cases total facility power is limited. So when that happens, you need to work on using the watts, you have better, getting more work done with fewer watts per workload.
09:58
In compute the largest productive use of power goes to CP US and GP US if you're using API, so you want efficient chips and processors to stretch the power across more workloads. You also want utilization of existing equipment to be as high as possible as this is generally the most efficient way of computing. I spoke to one university that is doing A I research and they actually monitor their GP US to make sure that any GP U that's powered up is
10:28
being used at at least 80% of its capacity and they even have alarms to alert them when the GP US are powered up but not being used at or near capacity so that they can consolidate workloads and all of this is for sustainability reasons. They don't want any wasted watts. Now, networking, this is a little less of an obvious area but networking is important and physical networking equipment definitely produces heat.
10:56
Let me give you a hint. They've got processors in them too and then we get to storage. I saved the best for last. But data storage is worth looking at 75% agree with the statement we have moved data or relocated a workload to reduce emissions and 83% say we have an initiative to reduce energy consumption by our storage infrastructure that involves
11:22
classifying and moving data to lower carbon and energy footprint locations, saving energy and being energy efficient in storage is like getting free money and free electricity. Imagine being able to continue to store all of the data you want but freeing up energy in your data center so that it can be used for arguably sexier purposes.
11:45
Not that storage isn't sexy mind you but you know, and it can be. But if we're being honest, A I might have an edge when it comes to pure sizzled power storage can be the cornerstone of your energy efficiency and sustainability commitment journey. And now might be the right time to think about your data storage strategy. We know that A I relies on having access to
12:10
fast storage. We know that storage is a major consumer of power in the data center. Maybe we should look for storage that excels in power efficiency and excels in speed and performance. This isn't the trade off that you might be imagining you can have both speed, power efficiency and scalability are possible if you select the right storage.
12:33
So I'm gonna conclude this presentation right there on storage, encouraging you the leader to ask yourself the strategic sustainability questions. You know, how do we integrate sustainability in our every day? And to think about what you can do in data centers that your organization operates to make sure that the highest percentage of energy we are using is going into productive work instead
12:57
of to inefficiency, waste and overhead storage is a great place to start this thinking. I hope you found this chat informative and motivating. If you're really motivated right now, I'd suggest you take a look at what pure storage is doing when it comes to storage that is designed to be sustainable and high performance from the ground up.
13:18
Go to pure storage's website to learn more about pure storage and the efficiency innovations that pure is bringing to market. Thank you and have a good day.
