 Okay, we're back here live in Las Vegas, this is IBM Edge. I'm John Furrier, the founder, so I'm joined by my co-host. Dave Vellante here from wikibond.org. Steve Wotowicz is here. He is the vice president of storage and network management software at IBM. Woj, welcome back to theCUBE. Thank you, thank you very much. You did a fantastic job on that last name. Yeah, so, I'm the president. What I take away from Amboosh's keynote in this other year analyst event was the conversation is different. The conversation is saying, hey, I am going to propose a business solution to you to do something, Salesforce related, retail, whatever vertical. There's business value that is going to drive the client's business. If it's not, then it's not a solution. Look, at the end of the day, I think that the point is... That's not, hey, I'm going to load up on some drives. That's kind of something. Right, IT is enabling business drivers, right? As you said, you know, there's a lot going on right? As you said, you know, the conversation is changing. No longer is it about, you know, how many IOPS. Now it's about, what is your SLA, right? Of that service you want to provide, knowing that the fundamental IT underlying infrastructure, you know, the bit heads, we have to figure out what the, you know, improve the IOPS, improve the heat, improve the efficiency, improve the density of the storage, you know, be able to scale out, scale up, all those type of things. So that's the question I have for you, that Dave and I wanted to probe, which is, okay, that's a complicated discussion. It is. Because now, it's within the IBM family, you go in traversing different groups, you know, what do you call them, you know, and then you got third-party multi-vendor, now you're talking openness, open source, I know you guys have open source DNA in the company, so how do you talk about that optimization? To me, optimization is there's never an end goal. It's always a matter of progressively getting better. Right? Because 10 years ago, would anyone have thought about, you know, the characteristics that we're seeing today with Flash? Probably 10 years from now, there may be something else. I mean, as you saw today, where we were, you know, we're pushing atoms around. Right? I suspect there'll be orders of magnitude above and beyond what we're seeing today with Flash, even, you know, we're seeing tremendously forward. Workloads are always needs to be improved in the long road and goal towards optimization. Workload optimization. And every workload could be different. I mean, if you want to go back and look at, you know, the archives of, you know, the digs of, you know, Egypt from 50 years ago, you can go find them, but it's not going to be instantaneous. Right? But being able to detect fraud in a moment's notice, knowing that, you know, a shoplifter is in your store based on facial recognition, knowing the propensities of where they go in the store, you kind of want that right now. So David Flore and I were doing a little back of the napkin the other day on how many copies, on average, actually get created. Yep. And the numbers were pretty astounding. When you think about somebody creates a file and then they email it to somebody and then they look at it from their laptop or their device or their iPad and they send it on. I mean, it's really, it gets exponential, right? I guess it's good if you're selling storage, but it's not good if you're buying storage. Yeah, we're trying to manage it, right. But then as well, you've got copies in the cloud. That's right. I mean, it's kind of a mess. And so from forgetting even about what do I keep? Right. You know, you don't even know what you have. Right. So my specific question is, how do you help solve that problem? So, as you said, you know, look at any industry analysts or survey that's done, typically somewhere between the range of 15 to 20 copies. And we had way more than that, by the way. Well, yeah. And I suspect that's true for some copies of some things, some pieces of data. What we're recommending to clients is to try and figure out how do you implement many of the technologies that exist today. So, you know, I hate to say that IT departments and teams are going to become irrelevant if they don't catch up, but implementing many of the technologies that exist today, whether it's deduplication. So your example exactly, right? Reply to all with attachments. You don't need to keep 5,000 copies of an attachment of a PowerPoint presentation saying, you know, welcome. Right. Right. Get rid of all of them. Then provisioning, deduplication, replication, compression, all these things can help you shrink the size of your copy pools. Right. The things that you make copies of your primary storage. One of the things that we're looking at very intensely is today, when you have a specific workload, you typically take the data, make a copy, and move the data to the application. In the future, we're going to try and reduce the number of copies you need and move the application to the data. Right. So... It's a function, baby. Well, exactly. Right. And when you talk about cloud, right? And the reason people are going to the cloud is they don't want to keep up with the infrastructure. They don't want to manage infrastructure that's growing so fast. They can't... They can't procure the capital expense, so they vote for OPX. Right. So, trying to reduce the number of copies in. The type of copies, the number of copies, and what the copies are used for is a really, really important part of how to reduce. So, do you want high availability and failover? Do you want near instant recovery for, you know, disaster recovery purposes? Do you want to offload it for long-term preservation in an archive to a tape? Or do you want to move it to another spinning disk for development and tests? Run your, you know, pre-production workloads on? Do analytics? I mean, I suspect everybody's talked about analytics and, you know, the analytic, you know, data-driven management. Yeah. Talk more about that. Yeah. Without analytics, the exponential curve is just not going to stop.