NVIDIA Unveils Rubin Platform & Open AI Vision at CES | Tech Field Day News Rundown: January 7, 2026

Marvell and Excon the final data center Frontier AWS raises the stakes. We're saying goodbye to Lou Gerstner logging for legal reasons ServiceNow acquires amis. And we should probably talk about A-I-C-E-S in this week's episode of the Tech Field Day Rundown.

Hello everyone. Welcome to the first episode of the Tech Field Day rundown. It is 2026, officially, it is the 7th of January.

We're very happy that you are joining us once again, and in the immortal words of Dee Snyder and Twisted Sister, I'm not gonna take it anymore 'cause well, that's what today is. But what I am gonna take is a moment for my co-host to introduce himself. Al, it's good to see you again.

Well, thank you and welcome to 2026. The, uh, 7th of January is indeed a beautiful day here in New Zealand, where it is not currently. International program is day, although maybe it should be International Programmers Day, uh, and of course National Bobblehead Day over there where you're, uh, Tom.

Yeah, you know, the funny thing is, I, I realized that people may not have their own bobblehead. Uh, I actually have a couple of them courtesy of our friends over at SolarWinds many years ago, and it really weirded my kids out that I have my own bottle heads. But you know what doesn't weird the kids out, or the adults for that matter is this wonderful lineup of news.

We, we were keeping track of it over the holidays so that we didn't miss any of the big exciting stuff. And we're gonna jump right in because our friends over at Marvell have announced that they are going to acquire Excon Technologies for about $540 million. This, of course, will strengthen their leadership in AI data center connectivity by adding several advanced PCIE and CXL switching silicon, uh, lines and experienced engineering talent.

The deal also expands Marvel's UA Link scale up switch capabilities, which positions the company to support next generation multi rack AI systems that require ultralow latency, high bandwidth interconnections, ex con's production, and next generation switching products broaden Marls portfolio. And that's going to expand their addressable market. They're expected to contribute revenue starting in the second half of fiscal year 2027.

The transaction is slated to close early in 2026 pending those pesky approvals. Al, did Marvell make a good move here? Absolutely.

I think this is a, a really significant move for them in terms of getting hung density and in, in data center infrastructure. Um, that's fundamentally what the excon, uh, technology is about. Uh, the CXL switch technology with a really high density and for, for PCI five, uh, switching capability in here, uh, if you've been following us for a while, if you've been following tech fields, you know, we're X as being to up your, your computer and particularly your large scale, uh, compute environment, being able to have a, a whole, uh, collection of GPUs that you can dynamically attach to a whole collection of servers and dynamically attach a whole collection of storage so that you can rebalance this collection of compute that you have based on the workload that you need at the time.

Uh, this kind of technology is absolutely gonna be vital as we're building out bigger and bigger data centers for running ai. And in common with some of the stories we're hearing throughout this news rundown, there'll be a focus on cost and cost efficiency that we maybe didn't see in the rush to get on board with ai. So I think Marvell is absolutely hitting one of the places that's gonna be very important for customers over the next couple of years of being able to build those composable infrastructures for real using CXL.

And, uh, this is gonna be useful for, for in, in businesses. Of course, most of the enterprise customers don't buy directly from Marvell. They'll be buying the package solutions that get shipped out.

And again, this is where having that larger organization in Marvell rather than just Excon, uh, brings that economy of scale that will reach out to deliver these solutions to larger numbers at data centers. I think this is a great thing to see. I'd love to see it happen a little faster, getting more products, uh, shipped out earlier so that we can have, uh, large data center infrastructures for running AI on premises and kind of sitting behind quite a lot of the stories we're also getting is that increased cost in ran around apocalypse uh, pricing that we're seeing over the last few months, uh, that will play into this too.

SpaceX is planning to transform some of their starlink version three satellites into orbital data centers, maybe as early as well this year. Uh, leveraging the terabit speeds that they have, the links between those satellites, the fact that they've got a heck of a lot of those satellites up there in this mesh network. Uh, and also using, reusing their low cost launch capacity with Starship, uh, brings back that vision that we had earlier from Jeff Bezos and Eric Schmidt for space based cloud computing to disrupt conventional terrestrial infrastructure and offering global reach sustainability, Ben benefits, uh, but maybe some challenges around regulation security and that also important engineering of making stuff work.

Um, Tom, time for your soapbox. Oh, man, you got it. It's a small soap box.

It's easily liftable into orbit. Um, where do I start? Uh, Starship hasn't actually launched yet.

Uh, I mean, it's it's gone off the launch pad, but it, it seems to blow up, but you know, who knew? Uh, space is hard, right? Um, but SpaceX has mostly figured it out.

I, I think, except for all the times that it didn't. And, uh, what are we gonna do? Oh, we're gonna put some data centers in orbit.

Well, why could that be a problem? Hmm. Let's see here.

I've talked about this on a number of occasions. We know cooling, cooling is very hard in space. Um, how are you gonna work on things in space?

Do you, do you remember when we had to work on Hubble? Do you remember how much of a pain in the neck that was? And that was one telescope.

Uh, things have to be bulletproof when you send them up. Uh, how Bulletproof Think Voyager two. Yeah, that thing's still running after doing the math here.

48 years. It's not running well. I mean, it's mostly down to like two instruments.

Uh, uh, but they keep it running. Uh, and that was something we built in like 1970 ish with that technology. Uh, this is stuff that breaks every couple of years.

So w why exactly do you wanna launch things into space? Uh, I mean, there's plenty of land around here to build, uh, data centers on, right? Oh, that's right.

You have to ask permission from people to build data centers on that land. You have to repay them for that. And do you think you can just blast some starlink satellites or your, uh, hyper, uh, satellites into orbit and just use them for whatever?

Uh, you guys seem to think that there's a lot of data center stuff out there that needs to be built. Uh, I'm not necessarily a naysayer on the, uh, the level of zitron, but how about we use what we've got before we start polluting space with junk? Uh, we, we've seen this over and over again.

Like, what's gonna happen if they manage to build out space the way they want it to? Are we gonna be surrounded by some kind of shield of, uh, starlink satellites and, uh, failed, uh, spaceship launch vehicles that that didn't, you know, didn't get it into the right orbit or whatever. Like, there is a lot of reasons to doubt this.

And if you or someone in any form of like, I don't know it, or finance or whatever, you should be asking some very serious questions about why we have to put these things in space. And if the answer doesn't come down to anything other than, well, we can, like, that's the problem. Instead of us saying, wouldn't it be cool to have a data center in space if I was on the board of one of these companies?

Well, my question would be, why do we need to have a data center in space? And if you can't answer that other than, well, it's cool, then don't do it. Uh, but I, I'll have to get off my soapbox yet because it's about to reenter the atmosphere and land at Point Nemo.

And, uh, Lord knows, I don't wanna get stuck out there long ago. AWS would loudly tell customers that they keep dropping the price of cloud services. But you'll notice that those announcements quietly stopped a few years ago in their place.

We get very quiet announcements of price increases. The latest round of increases. Were about 15% on the hourly costs of everyone's favorite Nvidia, H 200 GPUs.

You know, the ones that everybody wants to use for AI training. Will this drive AI training to other locations? Or is this just the first shoe to drop where everyone is gonna be raising the prices for their AI GPU clusters across the board to kind of drive a little bit more cash out of this market?

That's the thing, isn't it? Uh, are we at the getting to the end of that land grab of we've gotta prove that this is the right place to, uh, run your ai, which when you're doing a land grab into a new space, you don't mind losing some money. 'cause once you grab the land, you can jack up the rentals.

Now, to be fair to AWS last year, they did actually reduce the, the cost of these same GPU based, uh, P five E and P five E in, uh, instance types. These are the really large, uh, EC2 instances that that'll run you $43 an hour now 49, 75 an hour after the price increase. They did drop the prices quite a lot sometimes during the, the time last year.

So it's not that they haven't been price drops, it's just that there's also quiet price increases. Uh, as I mentioned in the, the earlier story around Marvel and Excon, the cost of ram, the cost of the, uh, high bandwidth memory is going up. And so maybe this is simply a reflection that AWS's cost to build new, uh, high performance large instances is going up.

Or maybe it is that they've got all of these instances, they're in really high demand because they've done a good job of telling people this is the right place to run your ai. And now because demand is exceeding supply, price goes up. It may be that simple.

Uh, there have been some statements from AWS saying, prices go up and down depending on demand and supply. And there was a pre-announcement, there was a warning that prices were going to be adjusted. Hmm.

So this was in, uh, last year, late last year. This, uh, AWS announced that prices on some instances were going to be adjusted. They just didn't say whether it was adjusted up or down.

This is kind of that we'll be very quiet when we tell you things are gonna go up in price, but we'll try and tell you very loudly when things are going down. Uh, these particular P five instances, P five E and P five E in with the H 200 GPUs, they've been a avail available since, since September, 2024. And they're really high performance, uh, instance types and buying them from AWS and paying only for the time that you use them is probably a good plan.

So long as you have that cloud friendly bursty workload where your peak workload is maybe four or five times your average workload. But as I said, Tom, we might see this driving some return to on-premises. If your workloads are much more even over time, if you don't have big peaks and troughs and utilization, maybe your AI is gonna be better off on premises.

And I think we'll see increasingly the inference component of AI being run on premises, uh, because it's a much more consistent workload, whereas the big bursty stuff is with training and fine tuning. And we may well continue to see that running on cloud platforms, and particularly AWS of course, would like you to have all of your cloud applications on the one true cloud. If you remember the way AWS talks about the world, some sad news came in over the break as well.

September, uh, December 27th, we got the sad news that Lou Gerstner had passed away. Lou had been the IBM chairman and CEO from 1993 to 2002. And that's a pretty significant time of change within IBM and his decisive leadership saved IBM from Yeah, being a has been to being a still very relevant to give a company that focused on customers and customer needs rather than just on the, the things that they built themselves and relying on the fact that nobody ever got fired for buying IBM, uh, Lou worked very hard to make sure that that remained true as IBM changed over time.

And so his legacy endures within IBM's values and strategy and the global impact of IBM as an organization. Now, Tom, I believe you, you previously were at IBM. Did you ever get invited up to the boardroom?

No, I did not. Uh, Lou did not sit in the area where I was an intern, uh, but Lou was always there with us. And, and that's one of the things that I think people don't really understand.

In 1993, IBM was effectively dying. John Akers knew that the only way to save the company was to split it up because the only way that IBM could stay relevant was that if you basically took the parts and turned them into their own little organizations and, and tried to get them to, you know, I don't know, fight with everybody else. Uh, remember this is when Microsoft was so big that the US federal government was thinking it was time to break them up.

Like, like this was the heyday of all of the upstarts that were fighting against IBM. All you gotta know is that when Lou took over, they were still making Blue Lightning 4 86 processors and OS two. And when Lou stepped in, first thing he did was say, we're not breaking up IBM.

There's no way that having IBM in pieces is going to benefit the company. And so he quietly went about reversing everything John Akers was trying to do, get rid of Blue Lightning, get rid of OS two focus on this new thing called the internet. And this idea that he came up with called E-Business.

And when he did that, that marked two points. One, the turnaround of IBM. That was the moment when they got off of life support and started growing again.

And two, and this is something that I have said for a number of years that in my mind was the end of Tom Watson's, IBM. And that is something that has, was repeated at the company when I was there. And I know for a fact it's still repeated there today.

This entity that you see today is not Tom Watson's, IBM. It is not the group that made the ass 400 or, um, ranac or any of that other stuff. This was the organization that started turning on IBM Global Services.

That was the group that I was actually an intern for, even though I worked at the ass 400 manufacturing plant in Rochester, Minnesota, shout out to Mayo, um, I was more, uh, aligned with the IT department. And Lou's changes were radical at the time because there were still a lot of people that worked at IBM who weren't sure that this upstart knew what he was doing. Fast forward now 24 years after he's gone, I'd say Lou was the reason why we still talk about IBM today in the way that we talk about it and not the way that we talk about something like weighing computers.

He knew what had to be done. He was brave enough to make those changes. And when he finally stepped aside in 2002 and he, he retired, he was not forced out like Aker was, he left the company in better shape than he found it.

com bust area. Like I was there like that quarter when the stock market just started sliding. And, and everyone was like, well, what does this really mean?

Lou Gerstner is one of those people that when we put up the Mount Rushmore of CEOs, people tend to put up the, you know, Kokas and the Jack Welchs of the world. Lou should be on that mountain. And I want the somebody else to keep this in mind.

Lou was the fifth person that they asked to be the CEO at IBM. They were actively trying to unseat acres and they went through five other people before they got to Lou. He worked at Amex and they're like, well, we need to bring in an outsider.

And I think that it worked. So Lou rest in peace. You, you did a lot to keep it on track.

Um, I, I'm, I'm not saying that I got my internship at IBM and launched my official career because of you specifically, but I did it because of the changes and the environment and the culture that you made at IBM. And for that I'll ever, I'll be eternally grateful. So here's hoping that the current form of IBM sticks around long into the future, and we can talk about the impact that great men like Lou Gerstner and several others have had on the world.

There's a group of news organizations who are suing Open AI for copyright infringement, and they just want a judgment that confirms their access to the logs of 20 million Sample chat GPT sessions. The group of news organizations wanna show that chat GPT is reproducing their articles without attribution, while OpenAI claims that their operation is fair use of published content. This judgment simply gives the news organizations access to the logs and it is not proof of whether or not chat GPT is using their content fairly.

With all of the backlash about unscrupulous data harvesting for AI training, is this case the beginning of the end for the current family of LLMs that are being trained on everything that the internet can see? Or is this an opportunity for a group of news organizations to go on a fishing trip? Yeah, it's kind of hard to tell in here.

Certainly there's a threat to the news organizations and, and to most of the ways that consumers access information. Historically, the threat is that they just use chat GPT and ask a, a simple question and get a, what appears to be a conversational, uh, answer that is accurate and, and reasonable may not be either accurate or reasonable, but it's a nice conversational answer. I get specifics of this particular, um, situation is that, uh, a magistrate had granted the request to say that, uh, these news organizations, uh, led by New York Times can get access to some sample data.

Now this is some sample logs of 20 million interactions with chat GPT and the objective here is to see where the chat GPT is actually using data that was harvested off these news sites and is cleansing out any sign that it came from the news sites and making it appear as if this is information that chat GPT itself knew of without really having to attribute back where it came. Uh, this case, there was a district court, um, Sydney sign who just upheld the previous magistrate's judgment that open AI must release a sample 20 million, uh, set of logs and deliver these so they can be analyzed for whether this is fair use, whether there is any direct attempt to hide the source of the information that is being, being presented out in these results to users. Uh, it's fun and interesting to see the sort of back and forth fight here.

Naturally open AI was trying to minimize the amount of information that they release out. Uh, now there's bits of it where OpenAI looks good. It's like we don't wanna have to work too hard on compliance for this case because while working compliance and discovery for cases is always an expensive and uh, not very profitable activity, but there's also some behaviors in open AI where they immediately started deleting more logs after there was a judgment saying you need to hold onto your logs.

Um, because, or at least after the litigation begun and they were supposed to have held, held onto their logs. And so a whole lot of logs got deleted very early in this process, uh, that are therefore not available. Mm.

Is that unscrupulous of OpenAI not to comply with the best practice of not deleting things that might be relevant to a call case? Yes. Yes.

That seems to be a bad thing. Uh, and then it's taken a long time for this to, to roll through. It's been months, it's been around a year since this began.

These, uh, news organizations just wanted to find out for sure whether the content that they have, uh, created as being misused and whether maybe they could charge open ai because realistically, those news organizations are about generating coverage. Uh, and either they want somebody to pay for the news that they've, uh, they've generated or they want attribution so that people will come back to their own websites and can be shown the ads that historically have always funded our news organizations. Is this gonna run?

Is this uh, court case gonna be settled quickly? No, no, this is gonna drag on for a long time. Open AI is gonna continue to follow the same process.

They've gone against all of these copyright infringement cases and that other, uh, ai, um, foundation model builders have also done of let's delay this as much as possible. Let's minimize the information we release. Uh, let's claim that everything is fair use and that if a human had taken exactly the same actions, it would be considered fair use.

Ignoring the fact that they've acted at a scale that is so much vastly larger than any human could act. We'll just have to keep watching this space because the story will continue slowly, slowly unfolding throughout this year. Just before Christmas, ServiceNow announced that they were going to acquire amis for seven and three quarter billion dollars and add cyber physical security and exposure management.

Uh, this will expand ServiceNow's existing security and risk business and it's gonna be a significant focus for ServiceNow as they expand beyond IT service management into managing the IT services that they manage. Does ServiceNow have a lot of credibility with security teams to replace the existing tools? So I dunno that they have it right now, but I think they're hoping they'll have it pretty soon because if there's one thing that stakeholders love, it is a one stop shop, right?

You may have heard it as, you know, one group to yell at one throat to do something with. Um, this is ServiceNow looking around at everybody going, well, what do they have that we don't, that people are still buying? This is very infamously, you know, like Apple's development mantra, right?

Is look at what great third party things are going on out there and integrate those pieces into what we're doing. It's the infamous sherlocking if you're old enough to remember when that was a thing. But I think what ServiceNow is really hoping for here is they want to add some stickiness to things that will allow them to build credibility in the market, but also allow them to expand what they can offer to people to help raise the the pricing, right?

Because one of the things that you run into with a SaaS opportunity is that people really want to use it for what they want to pay for it, right? Like you have to find ways to continually raise the price because if you are working in an organization that is like a restaurant, or if you're working for a company that does a lot of like travel, like raising the cost of something makes a lot of sense, right? Airline tickets aren't getting any cheaper, food costs aren't going down at all.

But if you work in a service where people just kind of look at your dashboard and help you figure out your tickets and stuff like that, other than inflation, why do you raise your prices? Well, in this case, they're raising their prices because they don't have something they can offer you. So if you want to have ServiceNow plus security, then you're gonna pay an extra X number of dollars per seat per month and that makes the investors happy.

But the key there is that you have to be able to integrate what amis is doing quickly into ServiceNow, because what people don't want is to have one company and one bill, but two separate systems. And you know, you'll forgive me for invoking the, the, the pane of single pane of glass, but that's what they want, is they wanna log into one system with one dashboard, see one number that gives them a thumbs up or a thumbs down on what's going on and do that. And I think that this is a great opportunity for ServiceNow to expand their offerings, but I also think that it's a good opportunity for people in the market who have a more robust directed security offering to kind of jump out in front and say, Hey, just know that we've been doing this for a while and we're gonna continue to do it while they're catching up.

So you can either go with what's proven or let them kind of shake out the bugs. Not to say that there's any bugs in arm, but any integration is always going to have friction that you're gonna have to watch out for. And for almost $8 billion, you hope that the friction is kept to a minimum because they're gonna have to start recouping the costs on that pretty soon.

And that's not gonna come from laying off the, the accounting department over there. All right, we wanted to take a closer look this week at everyone's favorite kickoff for the year, and that would be the consumer electronic show. And this year, NVIDIA's, CEO and fashion Maven Jensen Wong unveiled a sweeping vision for the future of AI led by their Ruben platform.

The company's first extreme co-designed six chip AI architecture designed to dramatically reduce AI computing costs while scaling performance across data centers, enterprises, and devices. Nvidia also introduced a broad portfolio of open models that span healthcare, robotics, climate science, and autonomous driving, which was highlighted by the Alpa Mayo platform delivering AI defined driving in the upcoming Mercedes-Benz CLA from supercomputers to personal AI agents and physical AI systems. Nvidia positions itself is a full stack AI company driving intelligence into every domain, device and industry.

And as some of the other news sites have already picked up on, you can't even get away from AI at CES. So Al let's kind of dive into this. Nvidia uses this as a launching platform for all of their latest announcements with Reuben and Alpa Mayo and so many other things.

Is Nvidia basically trying to pull away from the rest of the pack and say, if you're using us, you'll always be on the cutting edge, or are they trying to cover for the fact that people are still a little iffy when it comes to ai? I think fundamentally, uh, NVIDIA's been ahead of the pack for a long time. Uh, this is reflected by the, the huge number of Nvidia GPUs, the B two hundreds that we've previously seen.

Now the, uh, Vera Rubbin combination that we're seeing as the new product, um, huge numbers that isn't being moved. The most of the frameworks that we use for developing AI applications will very easily target these GPUs using the TOKUDA instructions that get the most performance out of them. So, uh, I think this is absolutely Nvidia saying we, we've moved the goalposts even further out in front of where everybody else is, um, when we're seeing them talking about, uh, when we're seeing Jensen talking about reducing the cost per token by a pretty significant factor.

And I don't have the number in front of me. My recollection was there it is five times higher tokens per second, five times better per performance per tco dollar and five times better power efficiency. This is a, a whole lot of interesting things.

This is in the, uh, AI native storage particularly that has announced here. Uh, this is just a, a whole set of hero number announcements at the beginning of the hardware that they're shipping because NVIDIA's job and NVIDIA's purpose is to sell you more of their hardware. Uh, it's interesting that alongside it is a whole lot of open models for ai.

So these are models that you can run on your own Nvidia equipment that are being trained on NVIDIA's own supercomputers, uh, whatever those super computers mean. Uh, these are models that you can use. This is basically to make it easier for you to start using and buying more Nvidia GPUs, whether they're the high performance ones that are sitting in the data center, soaking up all of that power and probably requiring liquid cooling because there's so much power density going on.

But, uh, it might also be in the DGX spark the little units that people are putting on their desks if they've got a spare, uh, six to 10 grand to spend on, on a desktop computer. But beyond that, NVIDIA's also looking towards getting these things in embedded locations. So there's a lot of dialogue around getting AI out to edge locations, whether those edge locations are autonomous vehicles or manufacturing plants or their retail and commercial locations where Nvidia GPUs are gonna be providing some, uh, some acceleration to those models at the edges.

Uh, I think there's a whole lot of this, which is we've got a lot of technology that that allows you as the end customer to buy more Nvidia GPUs or give away the things that allow us to, to sell you more GPUs. Tom, are you planning to upgrade your desktop to A-G-G-D-G-X spark so you can run AI models locally? Oh yeah, totally.

Why not? I mean, it's, it's only six or seven grand. How, how much can that set me back?

I, I think this is Nvidia realizing that they are going to have to start broadening their market, right? You, you see healthcare on the list and you're like, oh, but healthcare should be using AI already, but not for the things that you are thinking they should be using it for. Autonomous driving is another good example.

There is a lot of work that's been done on using AI for autonomous driving, but maybe not with Nvidia because the people who were working on that project were working on it a long time ago. And I, I wanna use an analogy here because I think it might re it might resonate with people. Um, I have a 3D printer that I backed on Kickstarter, uh, a couple years ago, and it's an okay printer.

Um, I probably paid more for it than I should have, but one of the things that holds it back is that a lot of the really cool stuff that runs on it is only enabled in the custom slicer that the company who made it released for it, now they promised that they were gonna release open source slicer support and it's kind of there, but it is like bare bones only. Like it kind of uses the hardware, but not really. And if you wanna do that, why don't you just use our slicer?

The hardware is being driven by the software in the case that we're dealing with here. It's the other way around. The software that Nvidia is releasing works, but it works really well on Nvidia hardware.

It's so much faster, it's so much better. It uses a little less power. I'm sure that's probably in there somewhere and you just need to buy some more GPUs, right?

Just, just buy some more stuff. We got 'em on the truck. We'll, we'll just, we'll, we'll ship 'em out to you today.

And that's really what they want is they want to continue this model where they are selling more and more and more hardware to everybody. And, and like the, the DGX that you mentioned is a perfectly good example. Why do I need a desktop AI system?

Well, to hear Nvidia tell it, it's because it's the future of computing. And I think it's interesting that the way that we've seen kind of the undercurrent in the market, it's being driven to the point where a lot of people are thinking, well, why do you need to have a desktop system at all or a laptop or something? Why can't you just rent your computers from the cloud?

You know, we've, we've seen that a lot, right? Remember when VMware Horizon came out and they're like, why don't you just run your desktop on a, on a cloud system somewhere? And now what they're saying is, well, why don't you run that in our cloud and we'll, it'll be AI powered and it'll be wonderful and you'll be able to use it.

And that's the end goal, right? Is they want to have as much consumption of these things as possible so that people are forced to use them as much as possible, which means they're always gonna be wanting to use new stuff, which means every couple of years they're gonna have to refresh that new stuff. Because the other thing is the Wizz kids on the r and d side can't stop developing because if they ever stop developing, then that will look like a plateau to people and that's a bad signal for the stock market.

So if you're continually making the systems use more power and have more resources available and do more cool stuff, you constantly have to make them do more cool stuff. Because if you don't, people might sit up and say, well, wait a minute, the one that I've got now is just as fast as the one that I had last year. I don't need to upgrade now.

And that could be a problem. And that's why I think what you'll see next is Nvidia continuing to offer a lot of these software solutions and man, they're gonna work great on H two hundreds, but if you had a B 200, they'd work even better. Um, I know that this sounds like a lot of hating on Nvidia and well, maybe a little bit it is, but one thing you won't hate is all of the great stuff that we've got coming up with Tech Field Day, Al your first up in 2026 with your event this month.

What have you got going on? Well, in three weeks time, you can catch us at AI Infrastructure field, day four. Of course, watch us online in all of your favorite locations, particularly on LinkedIn, where you can find us on the Tech Field Day LinkedIn page.

Uh, we've got some great companies coming along. We've got, uh, Cisco, uh, good friends, Cisco, uh, other good friends Hammer Space in there as well. We've got, uh, Ford Networks and exci and Fabrics AI coming in to present.

A few of those are companies we've never heard from, from before at Tech Field Day. So I'm looking forward to a pretty packed time at, uh, AI Infrastructure Field Day four. And it's not too late.

There may be a couple more logos added to those companies as well as we get closer. Uh, of course we also have a, a fun Cloud Field Day coming up in March. That'll be March or 11th and 12th.

And, uh, we're still a little earlier on the companies for that one. I've got a good group of people though. It's shaping up to be an event that's certainly gonna be fun when you're not on camera.

It should be a lot of fun when we're on camera too and later in March. I'm very jealous of Tom because, um, security conferences, um, RSA conferences one of the most interesting conferences that goes on in the industry. Um, and you're gonna be there.

Tom, I'm, this is the very first time that we've ever been at RSA officially as part of Tech Field Day. And, and it's exciting, right? Because you, like you said, RSA is probably the biggest thing in the security space right now.

There's 50, 60, 70,000 people that show up for it, and we are gonna be a part of that. com, you can see what we've got going on. Uh, we actually have presentations that are gonna be lined up from Veeam and Object First, and there's more there, there is definitely more on the, the schedule.

We'll be updating that list as soon as we can, and you'll probably see some familiar faces from the security space joining us as delegates there. And honestly, if you're at RSA, we'd, we'd love to meet you. So make sure you reach out to us on social media, uh, let us know.

I know it's still a couple months away, but it's never too late to start booking all those meetings right now. Uh, get 'em in before they release your name to all of the, the people on the floor and they book your calendar full of stuff. But make sure that you leave some time on your calendar for Wednesday, because that of course would be the new episodes of the Tech Field Day rundown that we release as a YouTube video or a podcast in your favorite podcast application of choice.

And we stream it on Techstrong TV as well. We wanna make sure that you are following along. So make sure that you are subscribed and you're getting notifications.

Leave us a review, leave us a comment. We'd love to hear what you have to say. And don't, uh, forget that Al and I are also on several other Techstrong and future and group programs throughout the week.

Uh, we have lots of great podcasts and lots of other things that we take part in. We'll be back next Wednesday to talk about all the stuff that happened when everybody got back into their office and finally got done checking all of their email. But until then, for myself, Tom Hollingsworth, for Al and everybody else here at Tech Field Day, thank you very much.

Enjoy the rest of your week and we'll see you soon.