WEBVTT 00:00:00.000 --> 00:00:03.459 Welcome to episode 243 of the 6G podcast. It's 00:00:03.459 --> 00:00:06.400 the inside scoop on everything 6G, and we cover 00:00:06.400 --> 00:00:09.500 six topics in about 30 minutes. And it's brought 00:00:09.500 --> 00:00:11.939 to you by More Insights and Strategy. I'm Anshul 00:00:11.939 --> 00:00:14.140 Sog, and joining me again this week is Foul Analyst 00:00:14.140 --> 00:00:16.899 Mike Dano from Ookla. This week, we're doing 00:00:16.899 --> 00:00:21.679 a special host MWC 2026 edition. Both of us attended 00:00:21.679 --> 00:00:24.539 the show last week. How did things go for you, 00:00:24.579 --> 00:00:28.679 Mike? So busy. This is my first show where I 00:00:28.679 --> 00:00:31.640 attended as an analyst, not a journalist. Much 00:00:31.640 --> 00:00:33.939 different for me. I'm with Ookla now, different 00:00:33.939 --> 00:00:35.679 companies. And we were just talking about the 00:00:35.679 --> 00:00:38.420 time change hit me especially hard at this show. 00:00:38.579 --> 00:00:40.320 So it was just a lot. This is the first time 00:00:40.320 --> 00:00:43.780 I stayed through Thursday. So I actually spent 00:00:43.780 --> 00:00:45.719 the weekend recovering. This is just a lot. I 00:00:45.719 --> 00:00:48.159 almost had the exact same experience, although 00:00:48.159 --> 00:00:51.159 this has probably been my fifth or sixth as an 00:00:51.159 --> 00:00:55.009 analyst. Jet lag still kicked my ass. and there's 00:00:55.009 --> 00:00:57.189 proof of it. I won't post where there's proof 00:00:57.189 --> 00:00:59.969 of it, but there's proof. And yeah, I stayed 00:00:59.969 --> 00:01:02.189 till Wednesday, but I probably would have stayed 00:01:02.189 --> 00:01:03.750 till Thursday if I didn't have my family with 00:01:03.750 --> 00:01:07.349 me. Ah, yeah, yeah. And Barcelona's great for 00:01:07.349 --> 00:01:10.430 bringing a family with you, not with an MWC schedule. 00:01:10.769 --> 00:01:12.870 Were you able to go out and go see stuff with 00:01:12.870 --> 00:01:16.209 your family? Yeah, a little bit on Sunday before 00:01:16.209 --> 00:01:19.329 the show, and then sporadically, mostly for dinner 00:01:19.329 --> 00:01:23.299 after the show, because to your point. My meetings 00:01:23.299 --> 00:01:27.180 are basically 8 a .m. till 5 to 6 p .m. every 00:01:27.180 --> 00:01:30.140 day, and then clawing my way back from the convention 00:01:30.140 --> 00:01:33.159 to my hotel, which was nowhere near the show, 00:01:33.260 --> 00:01:36.219 which is actually ideal, but not for the end 00:01:36.219 --> 00:01:39.519 of the day. Exactly, yeah. The travel in and 00:01:39.519 --> 00:01:41.760 out of there is just challenging, no matter which 00:01:41.760 --> 00:01:43.819 route you take. I was stuck in traffic in a taxi. 00:01:43.879 --> 00:01:48.040 I got crammed into the tightest subway in the 00:01:48.040 --> 00:01:50.079 history of the world, and both times it took 00:01:50.079 --> 00:01:52.019 me an hour to get to the show, so it just... 00:01:52.250 --> 00:01:55.329 There's no way to win. Only lose. Yes. So with 00:01:55.329 --> 00:01:58.310 that, I'll kick things off. I think I'll start 00:01:58.310 --> 00:02:01.909 with Qualcomm. And normally I wouldn't just pick 00:02:01.909 --> 00:02:06.450 one company for an entire topic, but they had 00:02:06.450 --> 00:02:10.330 a very busy MWC. They had the new Snapdragon 00:02:10.330 --> 00:02:13.509 Wear Elite, which is a wearable chip specifically 00:02:13.509 --> 00:02:17.710 for AI purposes. It will be used in a next generation 00:02:17.710 --> 00:02:19.969 Galaxy Watch, which is the first time they're 00:02:19.969 --> 00:02:23.219 in a Galaxy Watch. and pre -announcing such a 00:02:23.219 --> 00:02:26.240 product announcement on its own. So that's a 00:02:26.240 --> 00:02:28.159 big one for them. All of the new AI wearables 00:02:28.159 --> 00:02:30.639 will most likely use this chip in some way or 00:02:30.639 --> 00:02:33.319 shape. They also announced the FastConnect 8800, 00:02:33.500 --> 00:02:36.280 which is a Wi -Fi 8 chip. This is the company's 00:02:36.280 --> 00:02:39.439 first Wi -Fi 8 chip. It's also the first chip 00:02:39.439 --> 00:02:41.840 from the company to have four different radios 00:02:41.840 --> 00:02:47.259 on it. So Wi -Fi, Bluetooth, Thread, and Matter, 00:02:47.379 --> 00:02:52.080 and then UWV. So it's quad. radio technology. 00:02:52.319 --> 00:02:56.340 It's ultimately 4x4 Wi -Fi. So it's the first 00:02:56.340 --> 00:03:00.560 client Wi -Fi to have 4x4 physical antennas, 00:03:00.719 --> 00:03:03.560 which doubles the throughput from the theoretical 00:03:03.560 --> 00:03:09.919 of 5 .8 in 2x2 up to 10 .6. If your router supports 00:03:09.919 --> 00:03:11.960 it, it doesn't have to be a Wi -Fi 8 router, 00:03:12.039 --> 00:03:13.960 it could be a Wi -Fi 7 router as long as it supports 00:03:13.960 --> 00:03:16.780 those four spatial streams. And then in addition 00:03:16.780 --> 00:03:18.819 to that, they also had a data center announcement 00:03:18.819 --> 00:03:22.289 with the AI200, which is a rack scale AI solution. 00:03:22.550 --> 00:03:24.569 I'm sure everybody who went to their booth saw 00:03:24.569 --> 00:03:27.169 that rack goes then center there. And then last 00:03:27.169 --> 00:03:32.009 but not least, the Snapdragon x 105 modem, they 00:03:32.009 --> 00:03:37.370 skipped 95. And this is their 5g advanced chipset 00:03:37.370 --> 00:03:41.409 that will allow them to also test some 6g capabilities, 00:03:41.669 --> 00:03:45.169 but it's not explicitly being called a 6g modem. 00:03:45.169 --> 00:03:47.569 And it will come to phones at the end of this 00:03:47.569 --> 00:03:50.680 year. and potentially other devices next year 00:03:50.680 --> 00:03:53.520 that they'll probably use for trials for 6G. 00:03:54.139 --> 00:03:58.439 And yeah, that's just Qualcomm. Yeah, tons of 00:03:58.439 --> 00:04:00.159 stuff out of Qualcomm. And they talked about 00:04:00.159 --> 00:04:02.599 their concept of physical AI, which I know they're 00:04:02.599 --> 00:04:04.340 not the first one to talk about that. But I know 00:04:04.340 --> 00:04:06.180 they talked a lot about their wearables play. 00:04:06.620 --> 00:04:09.699 And I saw Meta had a booth at the show. They 00:04:09.699 --> 00:04:11.879 were showing off their glasses. There's several 00:04:11.879 --> 00:04:14.219 other companies that have smart glasses. Did 00:04:14.219 --> 00:04:16.519 you feel like, are we turning a corner in terms 00:04:16.519 --> 00:04:18.360 of wearables? Are we starting to get to the point 00:04:18.360 --> 00:04:21.620 where these are going to be a little less sideshow 00:04:21.620 --> 00:04:24.779 and a little more main stage type things? Because 00:04:24.779 --> 00:04:26.740 that certainly I thought was the message that 00:04:26.740 --> 00:04:29.540 Qualcomm was giving. Yeah, I think they're the 00:04:29.540 --> 00:04:32.300 one company that's positioned to make that mainstream. 00:04:32.600 --> 00:04:35.439 All of Netta's products run their chipsets. I 00:04:35.439 --> 00:04:38.500 would say I also saw Google's AI classes with 00:04:38.500 --> 00:04:40.939 a display in it. And I actually posted that on 00:04:40.939 --> 00:04:43.930 Twitter. comparing the two with the MetaRay bands. 00:04:44.449 --> 00:04:46.970 And the truth is they have a very mature product 00:04:46.970 --> 00:04:49.730 and they have a very mature model. So I think 00:04:49.730 --> 00:04:51.990 really it comes down to are the models ready 00:04:51.990 --> 00:04:54.649 to take advantage of this form factor? Because 00:04:54.649 --> 00:04:56.269 the truth is we're going to get all kinds of 00:04:56.269 --> 00:04:59.649 hearable AI or that's clipped to your ears or 00:04:59.649 --> 00:05:01.810 something you wear around your neck or something 00:05:01.810 --> 00:05:05.009 you wear on your face. But in my experience, 00:05:05.129 --> 00:05:07.769 I find that glasses are like the ultimate form 00:05:07.769 --> 00:05:11.189 factor for AI. whether so AI can perceive the 00:05:11.189 --> 00:05:13.490 world that you see it. And I think that's important. 00:05:14.009 --> 00:05:15.689 I don't think everybody's going to want to wear 00:05:15.689 --> 00:05:17.949 glasses. So there's going to be different solutions 00:05:17.949 --> 00:05:20.370 for different people. But ultimately, to your 00:05:20.370 --> 00:05:22.889 point, we're going to see a lot of AI wearables 00:05:22.889 --> 00:05:25.209 in the next year or two. And I feel like a lot 00:05:25.209 --> 00:05:26.610 of the ones that came out last year were too 00:05:26.610 --> 00:05:29.629 early for a lot of reasons. And I think that 00:05:29.629 --> 00:05:32.750 this Snapdragon Wear Elite product is optimized 00:05:32.750 --> 00:05:37.149 specifically for that kind of use case, especially 00:05:37.149 --> 00:05:39.279 when you consider that they're saying it can 00:05:39.279 --> 00:05:42.800 run a 2 billion parameter model on device, which 00:05:42.800 --> 00:05:45.399 I think people care about because it's going 00:05:45.399 --> 00:05:47.500 to affect battery relations. It's always connecting 00:05:47.500 --> 00:05:49.839 to the internet and privacy because you're going 00:05:49.839 --> 00:05:52.100 to have to have these things on pretty often, 00:05:52.139 --> 00:05:55.459 pretty frequently. And that's going to be, I 00:05:55.459 --> 00:05:57.660 think, the biggest challenge, that privacy factor, 00:05:57.819 --> 00:05:59.920 which is going to be a challenge for companies 00:05:59.920 --> 00:06:01.459 like Meta, especially considering the latest 00:06:01.459 --> 00:06:04.699 news. Yeah, we had the Google Glasses when it 00:06:04.699 --> 00:06:07.170 was at least five years. 10 years ago when you 00:06:07.170 --> 00:06:10.389 had the glass holes for just that problem. So 00:06:10.389 --> 00:06:12.110 I'm not sure that's been addressed, but it really 00:06:12.110 --> 00:06:14.750 feels like this time around, I'm interested to 00:06:14.750 --> 00:06:16.389 hear you say that because I'm wondering if this 00:06:16.389 --> 00:06:18.649 time around, if it's going to get real and it 00:06:18.649 --> 00:06:22.449 really feels like there's a lot more pieces of 00:06:22.449 --> 00:06:25.550 the puzzle in place to make those wearables actually 00:06:25.550 --> 00:06:28.810 happen this time and maybe not suffer from the 00:06:28.810 --> 00:06:31.829 glass hole that those initial ones fell into. 00:06:33.350 --> 00:06:36.939 Yeah, definitely. I think that. might lead us 00:06:36.939 --> 00:06:40.139 into your topic. Oh, yeah. Yeah. For my topic, 00:06:40.199 --> 00:06:42.379 my first topic that I chose, my big takeaway 00:06:42.379 --> 00:06:45.399 from Mobile World Congress this year is all about 00:06:45.399 --> 00:06:47.480 satellites. I can't help it. I'm just interested 00:06:47.480 --> 00:06:50.060 in satellites personally. And I also think that 00:06:50.060 --> 00:06:53.399 we got a lot of interesting satellite news last 00:06:53.399 --> 00:06:56.120 week and the week before. And so I think just 00:06:56.120 --> 00:06:58.959 to summarize the satellite news, I think Starlink 00:06:58.959 --> 00:07:01.779 obviously had a huge announcement at the show 00:07:01.779 --> 00:07:04.839 about the second generation. direct -to -device 00:07:04.839 --> 00:07:06.920 satellites. And they had a whole keynote about 00:07:06.920 --> 00:07:09.920 it. Lots of articles written about that. So that 00:07:09.920 --> 00:07:12.680 isn't really a surprise. They're going to have 00:07:12.680 --> 00:07:15.379 basically 1 ,200 satellites that they're going 00:07:15.379 --> 00:07:17.980 to launch over the next year or two that will 00:07:17.980 --> 00:07:20.540 basically be able to do transmissions direct 00:07:20.540 --> 00:07:23.480 to smartphones. And they said speeds up to 120 00:07:23.480 --> 00:07:26.279 megabits. We'll see how that actually plays out 00:07:26.279 --> 00:07:28.959 in real life and whether that's for one. customer 00:07:28.959 --> 00:07:31.399 in one area, or if it's for multiple customers 00:07:31.399 --> 00:07:33.860 in one area, we don't know the details yet, but 00:07:33.860 --> 00:07:35.899 we're certainly watching it very carefully. But 00:07:35.899 --> 00:07:38.899 what I thought was my takeaway was that was not 00:07:38.899 --> 00:07:42.199 the only announcement on the satellite internet 00:07:42.199 --> 00:07:46.279 front. We had announcements from SES. We had 00:07:46.279 --> 00:07:48.800 a rumor that SES is selling their ground infrastructure 00:07:48.800 --> 00:07:52.720 in order to raise money. So that is. One development 00:07:52.720 --> 00:07:55.860 that I'm watching, we had Viasat talk about its 00:07:55.860 --> 00:07:58.579 joint venture with Space 42. I think it's called 00:07:58.579 --> 00:08:01.100 Equities, I think is how you pronounce that. 00:08:01.199 --> 00:08:02.939 But that was pretty prominent at their booth 00:08:02.939 --> 00:08:05.699 at Mobile World Congress. They had a big drone 00:08:05.699 --> 00:08:08.480 that they talked about potentially connecting 00:08:08.480 --> 00:08:12.620 to that network. We had Utilsat, just in the 00:08:12.620 --> 00:08:15.899 last few days, got financing to refresh its OneWeb 00:08:15.899 --> 00:08:17.959 satellite constellation and participate with 00:08:17.959 --> 00:08:21.600 Europe's IRIS constellation effort. We had AST 00:08:21.600 --> 00:08:24.160 Space Mobile announce earnings last week, and 00:08:24.160 --> 00:08:26.639 they also had a big deal with TELUS. And TELUS 00:08:26.639 --> 00:08:30.839 also invested into AST Space Mobile in order 00:08:30.839 --> 00:08:33.480 to create some ground infrastructure there in 00:08:33.480 --> 00:08:36.340 Canada. We had Amazon Leo had an announcement 00:08:36.340 --> 00:08:40.480 with Vodafone for backhaul for cell sites. Samsung, 00:08:40.580 --> 00:08:42.799 of course, had their big Galaxy announcement. 00:08:42.940 --> 00:08:45.620 And their new Galaxy phone supports three different 00:08:45.620 --> 00:08:48.679 types of satellite connections. It's LTE, NB 00:08:48.679 --> 00:08:52.659 -IoT, and... 5G NR, and then Qualcomm had an 00:08:52.659 --> 00:08:54.759 interesting demo with their booth, which was 00:08:54.759 --> 00:08:57.539 satellite connectivity using millimeter wave 00:08:57.539 --> 00:09:00.919 spectrum. So I think my takeaway from all of 00:09:00.919 --> 00:09:02.679 those things, and each one of those things has 00:09:02.679 --> 00:09:04.980 interesting things to say about it. But when 00:09:04.980 --> 00:09:07.679 you step back from all these individual announcements, 00:09:07.879 --> 00:09:10.019 I think what you get to see is that there's just, 00:09:10.080 --> 00:09:13.139 this is a really new emerging market. And the 00:09:13.139 --> 00:09:16.639 fact that there are that many players, that many 00:09:16.639 --> 00:09:20.289 demos, that many funding announcements. All in 00:09:20.289 --> 00:09:21.990 just the span of a week, it really tells you 00:09:21.990 --> 00:09:24.330 that there's just a lot going on in the satellite 00:09:24.330 --> 00:09:28.350 marketplace today. And I fully expect that we'll 00:09:28.350 --> 00:09:30.950 continue to see that pace continue this year. 00:09:31.029 --> 00:09:32.769 I think this is just a market that is really, 00:09:32.850 --> 00:09:35.370 it's really on the upswing and the sky's the 00:09:35.370 --> 00:09:37.490 limit, really, in terms of where this thing is 00:09:37.490 --> 00:09:40.629 going to go. I think 2026 is going to be a pivotal 00:09:40.629 --> 00:09:42.669 year for just the development of the satellite 00:09:42.669 --> 00:09:46.259 internet market. fixed as well as direct -to 00:09:46.259 --> 00:09:48.399 -device mobile. I think this is the area that 00:09:48.399 --> 00:09:51.860 I'm going to be watching this year. It makes 00:09:51.860 --> 00:09:53.919 a lot of sense, especially when you consider 00:09:53.919 --> 00:09:56.200 that I remember talking to MediaTek about this 00:09:56.200 --> 00:10:00.279 two years ago, and they were saying 2027 will 00:10:00.279 --> 00:10:03.879 likely be the year when it's actually available. 00:10:04.539 --> 00:10:07.039 It doesn't seem that far off, actually. It seems 00:10:07.039 --> 00:10:10.039 like we might actually get commercial viability 00:10:10.039 --> 00:10:13.659 from a lot of these companies by 2027. Just considering 00:10:13.659 --> 00:10:16.440 all the activity we're expecting this year. Yeah. 00:10:16.460 --> 00:10:18.419 I mean, we're really, we're seeing it being playing 00:10:18.419 --> 00:10:20.700 out right now. We've got obviously iPhone has 00:10:20.700 --> 00:10:22.179 been doing this with the global star satellites 00:10:22.179 --> 00:10:25.240 for a while, but Starlink is turning on generation 00:10:25.240 --> 00:10:29.139 one service in multiple countries and with a 00:10:29.139 --> 00:10:31.139 lot of different providers and it's at the pace 00:10:31.139 --> 00:10:33.279 is accelerating. They're talking about their 00:10:33.279 --> 00:10:35.320 second generation satellites and there's a lot 00:10:35.320 --> 00:10:37.059 of additional players that look like they're 00:10:37.059 --> 00:10:39.080 going to play in or yeah, just very interesting 00:10:39.080 --> 00:10:42.259 to watch. But. To pass it back to you, I think 00:10:42.259 --> 00:10:44.179 you've got another couple of takeaways from Mobile 00:10:44.179 --> 00:10:48.179 World Congress, right? Yes. So I'm going to focus 00:10:48.179 --> 00:10:53.740 on the NVIDIA push at MWC. A lot of that push 00:10:53.740 --> 00:10:58.480 was in concert with Nokia. They actually had 00:10:58.480 --> 00:11:02.080 an analyst roundtable at Nokia's booth. Interestingly 00:11:02.080 --> 00:11:04.500 enough, they did not have a booth at the show, 00:11:04.639 --> 00:11:06.399 which I think a lot of people paid attention 00:11:06.399 --> 00:11:09.470 to. Considering that they're still relatively 00:11:09.470 --> 00:11:11.409 new to the space compared to everybody else. 00:11:11.850 --> 00:11:15.190 It's not a total shock. And I expect that next 00:11:15.190 --> 00:11:18.169 year they'll most likely have a booth. But anyways, 00:11:18.429 --> 00:11:21.129 they were all over the show. They actually showed 00:11:21.129 --> 00:11:23.289 me a map of all the different booths that they 00:11:23.289 --> 00:11:25.710 were at. And there were definitely some halls 00:11:25.710 --> 00:11:27.409 where they weren't present, but that's normal. 00:11:27.570 --> 00:11:30.409 But they were very present in numerous halls, 00:11:30.470 --> 00:11:34.610 especially hall six, hall two and three. So they're 00:11:34.610 --> 00:11:37.590 all over the place and they're. trying to make 00:11:37.590 --> 00:11:41.789 sure that they have their vision understood by 00:11:41.789 --> 00:11:45.269 the market. Because I think a lot of people anticipate 00:11:45.269 --> 00:11:49.389 that Nvidia's whole play here is to sell AI compute 00:11:49.389 --> 00:11:52.370 at the end. And that's actually their tertiary 00:11:52.370 --> 00:11:55.870 goal. Their primary goal is actually to explicitly 00:11:55.870 --> 00:12:00.370 just run the RAN more efficiently and to do it 00:12:00.370 --> 00:12:03.850 at higher performance than the existing incumbents. 00:12:04.240 --> 00:12:06.460 which I think is fundamentally an important thing 00:12:06.460 --> 00:12:10.039 to establish, but also that they will enable 00:12:10.039 --> 00:12:13.200 AI features within the RAN and power features 00:12:13.200 --> 00:12:17.360 like integrated sensing, ISAC, and then other 00:12:17.360 --> 00:12:20.700 AI enhanced capabilities, which they're working 00:12:20.700 --> 00:12:23.379 with companies that all come on, which is literally 00:12:23.379 --> 00:12:27.320 deploying AI at the modem and network level and 00:12:27.320 --> 00:12:29.679 allowing those models to talk to each other and 00:12:29.679 --> 00:12:33.779 understand network state. in a faster, more efficient 00:12:33.779 --> 00:12:36.179 way. And then the third goal is to take that 00:12:36.179 --> 00:12:39.539 excess capacity and enable it to be used for 00:12:39.539 --> 00:12:41.419 edge computes. It was an illuminating discussion 00:12:41.419 --> 00:12:44.460 with them. It was interesting because I got that 00:12:44.460 --> 00:12:47.120 from one of my one -on -ones with NVIDIA. They 00:12:47.120 --> 00:12:50.700 didn't fully express that three -pillar vision, 00:12:50.840 --> 00:12:53.480 I think, in every place that they should have. 00:12:53.919 --> 00:12:57.139 But to me, that makes a lot more sense than just 00:12:57.139 --> 00:12:59.960 talking about AI at the edge, which is editing. 00:13:00.350 --> 00:13:02.649 how a lot of people have captured Nvidia's strategy. 00:13:03.110 --> 00:13:05.649 And they're not just working with Nokia, they're 00:13:05.649 --> 00:13:08.190 also working with Ericsson and Samsung. And each 00:13:08.190 --> 00:13:10.269 of the vendors has a slightly different approach 00:13:10.269 --> 00:13:14.029 to how they're leveraging Nvidia's GPUs. Also, 00:13:14.169 --> 00:13:17.289 in our conversation with Nokia and NVIDIA, I 00:13:17.289 --> 00:13:18.990 started to get a better understanding of how 00:13:18.990 --> 00:13:21.529 they're going to deploy GPUs at the edge and 00:13:21.529 --> 00:13:23.970 in the infrastructure. And they're actually going 00:13:23.970 --> 00:13:25.710 to have add -in cards. It's not going to be a 00:13:25.710 --> 00:13:28.169 full rip and replace, which what some people 00:13:28.169 --> 00:13:30.750 were expecting would be. So there's a lot more 00:13:30.750 --> 00:13:35.169 nuance to this, of course, but overall, I think 00:13:35.169 --> 00:13:38.529 I left MWC being more confident in what they're 00:13:38.529 --> 00:13:41.690 trying to do and having more detail in what they're 00:13:41.690 --> 00:13:43.750 trying to accomplish. And they gave timelines 00:13:43.750 --> 00:13:48.090 and they gave specific feature readiness dates 00:13:48.090 --> 00:13:50.350 that they want to have for capabilities. People 00:13:50.350 --> 00:13:52.570 were asking about which Nokia products are going 00:13:52.570 --> 00:13:55.389 to get updated and when. This is, I think Nokia 00:13:55.389 --> 00:13:58.059 is the right partner to look at. for what Nvidia 00:13:58.059 --> 00:14:00.039 is trying to do just because there's some tightly 00:14:00.039 --> 00:14:03.299 integrated. But also T -Mobile is also pretty 00:14:03.299 --> 00:14:04.919 bought in as we've talked about in the past. 00:14:05.440 --> 00:14:08.120 But in general, I think that this looks a lot 00:14:08.120 --> 00:14:13.159 more measured and I think more professional than 00:14:13.159 --> 00:14:16.059 I was led to believe in the beginning. To be 00:14:16.059 --> 00:14:17.919 fair, we didn't have that many details. I think 00:14:17.919 --> 00:14:20.240 them coming out with more details of being more 00:14:20.240 --> 00:14:23.000 measured in their answers, I think has made me 00:14:23.000 --> 00:14:25.840 feel a little more confident that This is mostly 00:14:25.840 --> 00:14:29.759 a 5G infrastructure upgrade and that there will 00:14:29.759 --> 00:14:33.559 be software 6G implementations down the road. 00:14:33.960 --> 00:14:36.779 Those are a little bit more murky, but fundamentally 00:14:36.779 --> 00:14:39.080 they just want to be able to run the network 00:14:39.080 --> 00:14:42.279 more efficiently and to do it in a way that enables 00:14:42.279 --> 00:14:44.840 more software -defined features down the road. 00:14:44.940 --> 00:14:47.259 I know you have a lot of opinions. Oh, that's 00:14:47.259 --> 00:14:49.399 super interesting, actually. I'm very interested 00:14:49.399 --> 00:14:52.500 to hear that because, yeah, I obviously... NVIDIA 00:14:52.500 --> 00:14:55.299 is all over the show. My hot take was that it's 00:14:55.299 --> 00:15:00.480 not MWC, it's NWC. It's NVIDIA World Congress 00:15:00.480 --> 00:15:02.740 because, yeah, just NVIDIA was everywhere. And 00:15:02.740 --> 00:15:05.860 during the show, they announced massive investments 00:15:05.860 --> 00:15:09.539 into Momentum. and coherent some optical players 00:15:09.539 --> 00:15:13.620 in order to help design the networks in and around 00:15:13.620 --> 00:15:15.320 data centers to make sure that those networks 00:15:15.320 --> 00:15:18.399 are able to handle the massive capacity for the 00:15:18.399 --> 00:15:20.799 data that's being transmitted in and out of data 00:15:20.799 --> 00:15:22.799 centers and that follows on obviously you mentioned 00:15:22.799 --> 00:15:26.080 the the investment that nvidia has into nokia 00:15:26.080 --> 00:15:29.080 but yeah very interesting to hear that because 00:15:29.870 --> 00:15:32.190 It's hard to understand the scope of what NVIDIA 00:15:32.190 --> 00:15:34.610 wants to do in telecom, whether it's all hardware 00:15:34.610 --> 00:15:36.990 related, all software related, a mix of the two. 00:15:37.549 --> 00:15:40.350 Yeah, that's interesting to hear that they're 00:15:40.350 --> 00:15:43.289 taking one step at a time. And the idea that 00:15:43.289 --> 00:15:46.330 you may upgrade some components at the base station 00:15:46.330 --> 00:15:48.750 as necessary rather than a wholesale upgrade. 00:15:48.830 --> 00:15:52.090 I think that's very telling of the current situation 00:15:52.090 --> 00:15:54.370 in telecom, right? There's not a lot of demand 00:15:54.370 --> 00:15:57.929 for a wholesale. rip and replace update to equipment, 00:15:58.110 --> 00:15:59.490 right? They're just the operators just don't 00:15:59.490 --> 00:16:02.269 have that financial capacity. But a more targeted 00:16:02.269 --> 00:16:04.769 upgrade approach that actually does make sense 00:16:04.769 --> 00:16:07.169 for the locations where it would work. Actually, 00:16:07.330 --> 00:16:09.370 that leads. So I'm going to pivot real quick 00:16:09.370 --> 00:16:11.870 to my next topic, because I think that leads 00:16:11.870 --> 00:16:14.350 into it perfectly, which is the obviously we're 00:16:14.350 --> 00:16:17.090 here talking at the 6G podcast. So what was my 00:16:17.090 --> 00:16:20.330 6G takeaway at Mobile World Congress? My 6G takeaway 00:16:20.330 --> 00:16:24.429 from Mobile World Congress is that I think you've 00:16:24.429 --> 00:16:29.679 got to pay attention to LA Olympics 2028. I talked 00:16:29.679 --> 00:16:32.519 to a number of people and basically everyone 00:16:32.519 --> 00:16:35.620 said, if you're going to put a 6G date on your 00:16:35.620 --> 00:16:40.059 calendar, it is Summer Olympics 2028 for Los 00:16:40.059 --> 00:16:44.480 Angeles Olympics. So I fully expect to see some 00:16:44.480 --> 00:16:49.059 kind of 6G something at that Olympics. And I 00:16:49.059 --> 00:16:52.100 would not be surprised that NVIDIA plays either 00:16:52.100 --> 00:16:55.399 a very serious supporting role. or a starring 00:16:55.399 --> 00:16:58.700 role in that, in that 6g demo, whether it's a 00:16:58.700 --> 00:17:02.519 demo full -fledged network, hard to say the parameters 00:17:02.519 --> 00:17:05.140 of it, what spectrum bands, what capabilities 00:17:05.140 --> 00:17:08.559 it might have unclear as of yet. But I think 00:17:08.559 --> 00:17:10.859 it is, I'd give it very good chances that we're 00:17:10.859 --> 00:17:13.259 going to hear a lot about 6g at the Olympics 00:17:13.259 --> 00:17:18.000 in 2028 for sure. Yeah. And if you remember back 00:17:18.000 --> 00:17:21.240 in Hawaii, Qualcomm already pre -announced they 00:17:21.240 --> 00:17:24.299 were going to be. Part of the LA 28 Olympics. 00:17:25.460 --> 00:17:28.799 I think we're just getting more and more confirmations. 00:17:28.799 --> 00:17:31.339 That's going to be the case. Interesting fun 00:17:31.339 --> 00:17:35.299 fact, the surfing component of LA 28 will not 00:17:35.299 --> 00:17:38.380 be occurring in Los Angeles County. In fact, 00:17:38.440 --> 00:17:41.339 it will be taking place on the border of San 00:17:41.339 --> 00:17:44.160 Diego and Orange County because I know the beach 00:17:44.160 --> 00:17:46.019 where they're going to be playing. It's actually 00:17:46.019 --> 00:17:50.440 a competitive beach for competitive surfing. 00:17:51.130 --> 00:17:53.750 It makes total sense. Yeah. Oh, that's super 00:17:53.750 --> 00:17:55.829 interesting. Yeah, exactly. I think I have to 00:17:55.829 --> 00:17:57.210 say it. The only thing I'm interested in from 00:17:57.210 --> 00:17:59.029 the Olympics is the rock climbing stuff. That's 00:17:59.029 --> 00:18:00.730 my jam right there. So now that the Olympics 00:18:00.730 --> 00:18:02.670 is, or the, now that rock climbing is part of 00:18:02.670 --> 00:18:04.329 the Olympics, that's the one I'm going to be 00:18:04.329 --> 00:18:06.490 watching. So I assume that'll be in, they'll 00:18:06.490 --> 00:18:09.329 have it, they'll have it in LA. So if I get to 00:18:09.329 --> 00:18:12.150 go, that's where I'm headed. Yeah. I haven't 00:18:12.150 --> 00:18:14.230 really thought about attending yet, even though 00:18:14.230 --> 00:18:19.230 it's so close to me. Yes. Yes. Theoretically 00:18:19.230 --> 00:18:21.849 I am, but. It's still a two to three hour drive 00:18:21.849 --> 00:18:24.069 or two to three hour train ride, which honestly, 00:18:24.289 --> 00:18:26.650 I'll take the train run anytime. Yeah, exactly. 00:18:26.750 --> 00:18:28.650 Plus, it'll be just swarming with people anyway. 00:18:28.789 --> 00:18:32.569 So double that for whatever it is. Yes. And they 00:18:32.569 --> 00:18:35.869 still might not have the train line from the 00:18:35.869 --> 00:18:38.789 airport finished in time. So that'll be a fun 00:18:38.789 --> 00:18:42.190 one. But I'll transition to my third and final 00:18:42.190 --> 00:18:47.589 topic, which was a device roundup from MWC. Shocker, 00:18:47.589 --> 00:18:50.619 there were no 6G devices. But there were a lot 00:18:50.619 --> 00:18:53.960 of smartphones and other personal computing devices. 00:18:54.259 --> 00:18:58.940 We had Moto not actually announce any new phones, 00:18:59.099 --> 00:19:02.539 but give a lot more details and do a final dump 00:19:02.539 --> 00:19:05.380 on all the details from the new Moto Razr Fold, 00:19:05.680 --> 00:19:09.440 which is going to be, has a World Cup edition. 00:19:09.819 --> 00:19:14.140 We'll also have a Graphene OS option, which is 00:19:14.140 --> 00:19:17.319 really interesting. And we'll have the best rated, 00:19:17.420 --> 00:19:21.380 according to TXOMark, camera on a foldable, which 00:19:21.380 --> 00:19:24.200 addresses one of my biggest pet peeves of most 00:19:24.200 --> 00:19:28.740 foldables today. And there will be a bigger battery 00:19:28.740 --> 00:19:30.640 and faster charging. So it's going to be like 00:19:30.640 --> 00:19:32.859 a really flagship phone in a lot of ways. Dude, 00:19:32.920 --> 00:19:35.660 those razors are so awesome. I just like, of 00:19:35.660 --> 00:19:37.880 any foldables, I'm very drawn to that one. I 00:19:37.880 --> 00:19:39.680 think that one looks really nice. So to hear 00:19:39.680 --> 00:19:41.160 it is upgraded. That's a good news. Oh, there 00:19:41.160 --> 00:19:44.259 we go. And this is the ultra off the wind back. 00:19:44.940 --> 00:19:47.039 Yeah, I love it. It's just, those are just, that's 00:19:47.039 --> 00:19:49.019 a good one. And the main reason why I don't use 00:19:49.019 --> 00:19:51.640 it, camera's good, but it's not as good as my 00:19:51.640 --> 00:19:53.839 S26 Ultra. That's been one of my biggest pet 00:19:53.839 --> 00:19:57.619 peeves. And then there was also the Honor Magic 00:19:57.619 --> 00:20:00.299 V6, which is another foldable. The thinnest one 00:20:00.299 --> 00:20:03.039 with, I believe, the largest battery in the market, 00:20:03.200 --> 00:20:06.900 600 milliamp hours. And then they productized 00:20:06.900 --> 00:20:09.859 the robot phone, which is like an AI play. It 00:20:09.859 --> 00:20:11.579 was a prototype for a long time, and now they 00:20:11.579 --> 00:20:13.700 say they're going to ship one. I did see that 00:20:13.700 --> 00:20:16.539 with the, it has the camera that comes out. That's 00:20:16.539 --> 00:20:20.390 so weird. It could work for some people. If you 00:20:20.390 --> 00:20:22.250 wanted to show off a crazy phone, if you wanted 00:20:22.250 --> 00:20:24.089 to show that off and get some attention from 00:20:24.089 --> 00:20:25.950 your friends, like that is the thing to get. 00:20:26.650 --> 00:20:30.930 Yes. And then Xiaomi had their Xiaomi 7T Ultra 00:20:30.930 --> 00:20:33.069 do its global launch. So they did the launch 00:20:33.069 --> 00:20:35.630 in China and then a global launch at MWC. And 00:20:35.630 --> 00:20:38.390 they did a special edition Leica lights phone, 00:20:38.569 --> 00:20:42.170 which has a turnable dial for the camera. It's 00:20:42.170 --> 00:20:44.710 one of the best cameras on a phone, period. But 00:20:44.710 --> 00:20:47.089 then they went with an extra Leica special edition. 00:20:47.640 --> 00:20:49.500 I think that's two grand or something like that. 00:20:49.579 --> 00:20:51.779 So it's a lot more expensive than the base model, 00:20:51.900 --> 00:20:54.180 but it's a special edition. And then Xiaomi also 00:20:54.180 --> 00:20:57.519 had that cool prototype car at their booth, which 00:20:57.519 --> 00:21:00.779 is a, they call it the Vision GT. So it's a Gran 00:21:00.779 --> 00:21:03.839 Turismo kind of style prototype, which will never 00:21:03.839 --> 00:21:06.460 ship as a real car, but it's an idea of what 00:21:06.460 --> 00:21:08.059 they're planning on doing for their high -end 00:21:08.059 --> 00:21:11.859 product. And then Lenovo launched a bunch of 00:21:11.859 --> 00:21:14.599 laptops and then showed off a bunch of prototypes, 00:21:14.779 --> 00:21:18.710 including a... convertible tablet. That's a gaming 00:21:18.710 --> 00:21:20.630 tablet that can be used as a gaming handheld 00:21:20.630 --> 00:21:25.529 or as a as a laptop, like a tablet, with a keyboard, 00:21:25.730 --> 00:21:27.609 all that all those doodads. And then they also 00:21:27.609 --> 00:21:31.990 had a highly convertible PC with screens and 00:21:31.990 --> 00:21:34.269 keyboards that can move around inside and out 00:21:34.269 --> 00:21:37.589 the device. And you customize it based on what 00:21:37.589 --> 00:21:39.930 you want from the laptop. So you can like have 00:21:39.930 --> 00:21:41.869 one screen on the front and one screen on the 00:21:41.869 --> 00:21:44.430 back. Or you can have the keyboard on the outside 00:21:44.430 --> 00:21:46.609 and the screen on the inside. If it was just 00:21:46.609 --> 00:21:49.769 like a configurable laptop. I'm less sold on 00:21:49.769 --> 00:21:52.710 that concept, but I think overall Lenovo has 00:21:52.710 --> 00:21:55.250 been very good at showing off their prototypes 00:21:55.250 --> 00:21:57.410 and then trying to productize them if there's 00:21:57.410 --> 00:21:59.609 demand for them. They showed off a bunch of CES 00:21:59.609 --> 00:22:01.869 stuff as well, but none of that's new. But they 00:22:01.869 --> 00:22:03.950 did have iFixit at their booth, which I loved. 00:22:04.069 --> 00:22:07.130 They were shot off, I think the T14S, which now 00:22:07.130 --> 00:22:10.029 has a 10 out of 10 repairability score because 00:22:10.029 --> 00:22:13.809 you can basically repair yourself, the entire 00:22:13.809 --> 00:22:17.160 laptop based on, all the guides and parts that 00:22:17.160 --> 00:22:20.200 are available and the QR codes. So you can like 00:22:20.200 --> 00:22:23.279 basically maintain it yourself if you want. But 00:22:23.279 --> 00:22:26.019 that also enables the laptop in general to be 00:22:26.019 --> 00:22:28.640 reused and repaired more often. So it doesn't, 00:22:28.640 --> 00:22:31.119 all the components are recycled. It's just they 00:22:31.119 --> 00:22:33.420 fix what's broken and then sell it to the next 00:22:33.420 --> 00:22:35.500 person for a slight discount. It's really interesting 00:22:35.500 --> 00:22:39.480 how that might help with all the environmental 00:22:39.480 --> 00:22:42.440 things companies are trying to deploy and make 00:22:42.440 --> 00:22:44.940 sure that they adhere to whatever standards are 00:22:44.940 --> 00:22:48.380 required for the missions. I really thought at 00:22:48.380 --> 00:22:50.160 this show, like we talked about the robot phone, 00:22:50.240 --> 00:22:54.640 like I saw foldable phones, glasses. I know devices 00:22:54.640 --> 00:22:57.740 are not nearly as much of a focus on the show 00:22:57.740 --> 00:22:59.619 as they have been in the past, but I felt like 00:22:59.619 --> 00:23:03.460 this year the device sort of game was at least 00:23:03.460 --> 00:23:04.960 a little bit better than previous years. Like 00:23:04.960 --> 00:23:07.380 I just felt like there was just more stuff to 00:23:07.380 --> 00:23:11.019 look at in terms of phones. Totally. Somebody 00:23:11.019 --> 00:23:13.250 who's been going for a while. It definitely felt 00:23:13.250 --> 00:23:15.730 it was like waning. A lot of companies were just 00:23:15.730 --> 00:23:17.890 doing their own events and not doing anything 00:23:17.890 --> 00:23:20.750 at MWC. So I agree. It was nice to have that. 00:23:20.809 --> 00:23:24.549 I also left off that I did go to the Google Avenue. 00:23:24.690 --> 00:23:27.609 I got my fair share of Android pins, which I 00:23:27.609 --> 00:23:30.069 don't have on me right now. But I also tried 00:23:30.069 --> 00:23:33.829 out the AR glasses or AI glasses from Google 00:23:33.829 --> 00:23:35.609 with the single display, which I think I mentioned 00:23:35.609 --> 00:23:37.890 earlier. That's also a prototype, but will very 00:23:37.890 --> 00:23:41.599 likely be close to the production model. Lots 00:23:41.599 --> 00:23:44.079 of devices. I'm the device guy at the firm, so 00:23:44.079 --> 00:23:48.740 I felt obligated to give a report from MWC. Good. 00:23:48.900 --> 00:23:51.000 Yeah. Interesting. Yeah. Interesting stuff. It 00:23:51.000 --> 00:23:53.539 was good to see that. Yeah. So I have, I guess 00:23:53.539 --> 00:23:56.059 it's a little bit related to devices. My last 00:23:56.059 --> 00:24:02.740 topic is, yeah, I'm still wondering about the 00:24:02.740 --> 00:24:07.980 idea of edge processing for these devices, because 00:24:07.980 --> 00:24:12.430 a lot of the applications, Or some of the applications 00:24:12.430 --> 00:24:14.529 that people have been talking about may require 00:24:14.529 --> 00:24:17.650 edge processing. And especially if you're talking 00:24:17.650 --> 00:24:20.470 about augmented reality on these glasses, that 00:24:20.470 --> 00:24:23.529 computer processing is extremely intensive and 00:24:23.529 --> 00:24:25.789 probably cannot be done on device and probably 00:24:25.789 --> 00:24:28.170 will need to be done somewhere nearby to that 00:24:28.170 --> 00:24:30.150 user in order to get the timing right. Like if 00:24:30.150 --> 00:24:32.809 you turn your head when your augmented reality 00:24:32.809 --> 00:24:35.559 vision... that you're looking through in your 00:24:35.559 --> 00:24:37.799 glasses needs to react to how you're turning 00:24:37.799 --> 00:24:40.099 your head and it displays different information 00:24:40.099 --> 00:24:42.440 in different information as you turn your head. 00:24:43.000 --> 00:24:45.019 Presumably all that has to be done with an edge 00:24:45.019 --> 00:24:48.140 computing setup. But my question is, so I get 00:24:48.140 --> 00:24:51.039 that. NVIDIA is obviously a part of that whole 00:24:51.039 --> 00:24:54.599 story. So if NVIDIA is able to supercharge a 00:24:54.599 --> 00:24:57.759 base station with its GPUs and now the base station 00:24:57.759 --> 00:24:59.980 is near to you, that's doing the edge processing. 00:25:00.240 --> 00:25:02.940 It's powered by an NVIDIA GPU. That all makes 00:25:02.940 --> 00:25:05.839 sense to me. My question. Now is, does it make 00:25:05.839 --> 00:25:07.920 sense to have all these edge computing nodes 00:25:07.920 --> 00:25:12.519 unmanned and unprotected, particularly in an 00:25:12.519 --> 00:25:17.059 era when there's tons of discussion about sovereign 00:25:17.059 --> 00:25:21.220 AI and sovereign data processing? Does it make 00:25:21.220 --> 00:25:25.480 sense to have all of this edge computing conducted 00:25:25.480 --> 00:25:30.740 on a base station that is unmanned and presumably 00:25:30.740 --> 00:25:33.269 unprotected? I don't know. That's just a question 00:25:33.269 --> 00:25:35.529 I'm raising. I feel like there's a lot of questions 00:25:35.529 --> 00:25:38.509 in this industry that we haven't yet thought 00:25:38.509 --> 00:25:41.670 all the way through. Can you afford to upgrade 00:25:41.670 --> 00:25:45.329 a base station to support edge computing? Does 00:25:45.329 --> 00:25:47.690 that financially make sense? Does it make sense 00:25:47.690 --> 00:25:50.769 from a data sovereignty perspective? Does it 00:25:50.769 --> 00:25:52.789 make sense from a literal enterprise security? 00:25:53.049 --> 00:25:55.789 So if enterprises are buying the computing capacity 00:25:55.789 --> 00:25:58.130 at that base station, can they do that with the 00:25:58.130 --> 00:25:59.930 base station is unmanned? There's just no one 00:25:59.930 --> 00:26:02.980 there to protect. A lot of the enterprise applications 00:26:02.980 --> 00:26:06.859 for AI assume that the data is going to be extremely 00:26:06.859 --> 00:26:09.200 well protected, maybe by being on -premises. 00:26:09.460 --> 00:26:12.140 And if it's not on -premises, would they trust 00:26:12.140 --> 00:26:15.900 the computing to be done in a data center or 00:26:15.900 --> 00:26:17.960 an edge computing data center at the base of 00:26:17.960 --> 00:26:19.819 a cell tower that is unmanned? I don't know. 00:26:19.920 --> 00:26:22.740 I just feel like there's so many questions about 00:26:22.740 --> 00:26:25.500 this AI future that I think are not yet fully 00:26:25.500 --> 00:26:28.500 baked. Not yet fully baked. That's my, but I 00:26:28.500 --> 00:26:33.410 don't know. I don't know if you've got a... I 00:26:33.410 --> 00:26:37.170 do have an opinion. I think that there's a lot 00:26:37.170 --> 00:26:40.529 of challenges with hybrid AI. And I think this 00:26:40.529 --> 00:26:41.829 is definitely one of the ones that has to be 00:26:41.829 --> 00:26:45.930 resolved. I think there will have to be some 00:26:45.930 --> 00:26:49.490 level of security that proves like a chain of 00:26:49.490 --> 00:26:52.369 custody. And there will probably be all kinds 00:26:52.369 --> 00:26:55.789 of security layer on top of that. And I think 00:26:55.789 --> 00:26:59.059 because of that, that might even... reduce some 00:26:59.059 --> 00:27:01.359 of the compute capacity that's available at the 00:27:01.359 --> 00:27:05.599 edge just to to do decrypt and encrypt and that 00:27:05.599 --> 00:27:08.920 might be part of the equation of how nvidia deploys 00:27:08.920 --> 00:27:11.440 this and how others do it and maybe that might 00:27:11.440 --> 00:27:14.000 be a big component of different companies approaches 00:27:14.000 --> 00:27:18.579 to this hybrid ai but i think ultimately for 00:27:18.579 --> 00:27:21.960 ai to be successful on device it also has to 00:27:21.960 --> 00:27:25.160 be pretty good at the edge because to your point 00:27:25.160 --> 00:27:27.299 there's gonna be tons of applications that are 00:27:27.710 --> 00:27:32.029 way too big to run on device, but maybe not computationally 00:27:32.029 --> 00:27:34.549 marked to run all the way in the cloud in terms 00:27:34.549 --> 00:27:39.069 of cost. So we have to think about the security 00:27:39.069 --> 00:27:42.750 aspect of things, but also the costs, and what's 00:27:42.750 --> 00:27:44.710 going to work locally, what's going to work at 00:27:44.710 --> 00:27:45.990 the edge of what's going to work in the cloud. 00:27:46.430 --> 00:27:50.670 And I will also require models of varying sizes 00:27:50.670 --> 00:27:55.839 and complexities. So I think I had some of these 00:27:55.839 --> 00:27:58.519 discussions at MWC, which was like, what's going 00:27:58.519 --> 00:28:01.519 to end up, what's going to run on the edge? How 00:28:01.519 --> 00:28:04.359 big is it going to be? And how relevant will 00:28:04.359 --> 00:28:07.740 it be to the user or the application developer? 00:28:08.339 --> 00:28:10.660 Because is it going to be a 7 billion parameter 00:28:10.660 --> 00:28:13.319 model? If it is, maybe you can actually just 00:28:13.319 --> 00:28:15.059 crunch it down to something you can run on device. 00:28:15.660 --> 00:28:17.799 Or is it going to be like a 70 billion parameter 00:28:17.799 --> 00:28:20.740 model? That's maybe actually should end up in 00:28:20.740 --> 00:28:22.859 the cloud. Or maybe we don't actually. end up 00:28:22.859 --> 00:28:25.519 running it on the RAN. And we run totally different 00:28:25.519 --> 00:28:28.460 workloads on the RAN. So I think if there's one 00:28:28.460 --> 00:28:31.380 way that I look at it, it's maybe the RAN ends 00:28:31.380 --> 00:28:34.200 up being something that the operators use to 00:28:34.200 --> 00:28:37.279 run their own services that they offer down the 00:28:37.279 --> 00:28:40.539 road. I think that the live translation feature 00:28:40.539 --> 00:28:43.359 that T -Mobile deploy in my mind for those things 00:28:43.359 --> 00:28:45.880 that make sense, there might be more of those 00:28:45.880 --> 00:28:49.160 types of workloads that make sense that don't 00:28:49.160 --> 00:28:52.099 touch. super sensitive data. But then again, 00:28:52.240 --> 00:28:55.000 translating a live call could actually be sensitive 00:28:55.000 --> 00:28:58.240 data. Yeah, I think you raise a very valid point 00:28:58.240 --> 00:29:01.380 in that we do have concerns about data security. 00:29:01.759 --> 00:29:05.200 I will tell you, really sensitive data, even 00:29:05.200 --> 00:29:07.579 to this day, doesn't really even go to a cloud. 00:29:07.680 --> 00:29:10.200 It might not even be a discussion because it's 00:29:10.200 --> 00:29:12.920 already not part of the discussion already. And 00:29:12.920 --> 00:29:15.279 your only option is basically on -prem or on 00:29:15.279 --> 00:29:18.769 -device. You and I did sit in that T -Mobile 00:29:18.769 --> 00:29:20.849 press conference at Mobile World Congress, and 00:29:20.849 --> 00:29:23.849 they talked about their cloud that Deutsche Telekom 00:29:23.849 --> 00:29:26.849 is building in Europe. And already at 10 ,000 00:29:26.849 --> 00:29:29.869 GPUs, and they only announced it a few months 00:29:29.869 --> 00:29:33.490 ago, and it's already at 30 % utilization. And 00:29:33.490 --> 00:29:36.349 then they're going to build more locations throughout 00:29:36.349 --> 00:29:39.930 Europe for Sovereign. data center play and seemingly 00:29:39.930 --> 00:29:42.430 very successful already. It definitely seems 00:29:42.430 --> 00:29:45.349 like there is demand and operator. And not only 00:29:45.349 --> 00:29:46.990 is there demand, but there are operators out 00:29:46.990 --> 00:29:48.970 there that are investing in this successfully, 00:29:49.150 --> 00:29:51.450 which is a big change even from, I think, just 00:29:51.450 --> 00:29:54.309 last year. And I would also add that I think 00:29:54.309 --> 00:29:57.089 carriers outside the U .S. have more opportunities. 00:29:57.349 --> 00:29:59.569 Right. Because there's a lot more appetite to 00:29:59.569 --> 00:30:02.470 not be dependent on American cloud companies. 00:30:02.710 --> 00:30:06.029 Right. For this capability and the next natural 00:30:06.029 --> 00:30:11.359 tenant. is the operators. So I think non -US 00:30:11.359 --> 00:30:14.559 based operators do have opportunities here. I 00:30:14.559 --> 00:30:15.980 think we mentioned this a little bit in our last 00:30:15.980 --> 00:30:19.720 podcast. I think that's still true. Non -US operators 00:30:19.720 --> 00:30:23.519 are going to be the natural successors to US 00:30:23.519 --> 00:30:26.799 cloud companies if that separation were to occur 00:30:26.799 --> 00:30:30.119 or is already occurring because some countries 00:30:30.119 --> 00:30:33.039 don't necessarily trust the US government. There's 00:30:33.039 --> 00:30:34.920 a possibility the US government could just decide 00:30:34.920 --> 00:30:38.779 to cut. anthropic off from some other country, 00:30:38.880 --> 00:30:42.859 right? ByteDance. Yeah. We could, it's anybody's 00:30:42.859 --> 00:30:45.480 world now. Yeah. I'm not really sure where we're 00:30:45.480 --> 00:30:47.960 going with it, but I think the Europeans aren't 00:30:47.960 --> 00:30:51.339 willing to take too many risks. And fair position 00:30:51.339 --> 00:30:53.259 too. I know. And we'll see what the next week, 00:30:53.359 --> 00:30:54.859 everything will change next week. We'll see. 00:30:54.960 --> 00:30:57.920 We'll see how it goes. A war broke out while 00:30:57.920 --> 00:31:00.680 we were traveling, attending a trade show in 00:31:00.680 --> 00:31:02.619 another country. Who knows what will happen next? 00:31:03.240 --> 00:31:05.640 especially considering what's going on with everything. 00:31:05.839 --> 00:31:12.839 The war with Iran has proven out the ISAC use 00:31:12.839 --> 00:31:16.259 case is very valuable and will likely be at the 00:31:16.259 --> 00:31:18.700 forefront of many countries' 6G deployments, 00:31:18.759 --> 00:31:21.079 especially in the Gulf states. I think that will 00:31:21.079 --> 00:31:24.859 be a very big chip in the negotiations for Spectrum 00:31:24.859 --> 00:31:27.140 and other things with governments. But that was 00:31:27.140 --> 00:31:29.619 just an aside that I had during MWC that I didn't 00:31:29.619 --> 00:31:32.269 mention earlier in the podcast. Great point. 00:31:32.329 --> 00:31:35.150 Yeah. If the focus wasn't on public safety yet, 00:31:35.250 --> 00:31:40.329 it is now. Really? No. I really enjoyed our podcast. 00:31:40.490 --> 00:31:42.589 We didn't actually run as long as I thought we 00:31:42.589 --> 00:31:44.589 would, which. Not too bad. Covered a lot of ground. 00:31:45.869 --> 00:31:48.690 Mixed feelings about that. But overall, I really 00:31:48.690 --> 00:31:51.990 enjoy our second podcast together. Another big 00:31:51.990 --> 00:31:54.309 one. We hope our viewers and listeners found 00:31:54.309 --> 00:31:57.289 this week's topics interesting. And please don't 00:31:57.289 --> 00:31:59.650 forget to rate us and subscribe and tune in again 00:31:59.650 --> 00:32:02.329 next week. Thanks again, Mike, for joining me. 00:32:02.690 --> 00:32:03.009 Thanks.