WEBVTT 00:00:00.000 --> 00:00:03.819 Welcome to episode 239 of the G2 on 5G, its latest 00:00:03.819 --> 00:00:08.140 insight scoop on everything 5G and 6G. We cover 00:00:08.140 --> 00:00:10.339 six topics in about 20 minutes, and it's brought 00:00:10.339 --> 00:00:12.660 to you by More Insights and Strategy. I'm Will 00:00:12.660 --> 00:00:14.839 Townsend. Joining me again this week is fellow 00:00:14.839 --> 00:00:17.760 analyst Anshul Sag. Let's get started with my 00:00:17.760 --> 00:00:20.160 first topic. And Anshul, I want to talk about 00:00:20.160 --> 00:00:25.429 AT &T. And just yesterday, on October 8th, the 00:00:25.429 --> 00:00:28.850 company declared its nationwide 5G standalone 00:00:28.850 --> 00:00:32.189 deployment. This is obviously big news. You and 00:00:32.189 --> 00:00:34.670 I for years have talked about the true promise 00:00:34.670 --> 00:00:38.869 of 5G matching RAN and 5G core infrastructure. 00:00:39.450 --> 00:00:44.369 And so AT &T has finally accomplished this major 00:00:44.369 --> 00:00:47.270 milestone. This also follows recently on the 00:00:47.270 --> 00:00:49.969 heels of the company's announcement around 5G 00:00:49.969 --> 00:00:53.770 REDCap technology, and it's available nationwide 00:00:53.770 --> 00:00:58.219 as well. So this is a great milestone for AT 00:00:58.219 --> 00:01:00.840 &T. We knew it was going to come. I do recall 00:01:00.840 --> 00:01:02.799 that we predicted that this would happen for 00:01:02.799 --> 00:01:06.379 AT &T in the second half of this year. So we 00:01:06.379 --> 00:01:09.560 can take our victory laps there. I'm expecting 00:01:09.560 --> 00:01:12.819 now that sort of paves the way for some very 00:01:12.819 --> 00:01:14.700 innovative services, especially when it comes 00:01:14.700 --> 00:01:17.280 to slicing. When you look at what T -Mobile has 00:01:17.280 --> 00:01:21.780 done with T -Priority, with T -IoT, their SASE 00:01:21.780 --> 00:01:24.609 solution that has a security slice. and most 00:01:24.609 --> 00:01:28.129 recently super mobile so what are your thoughts 00:01:28.129 --> 00:01:30.209 what do you expect to see from at t now that 00:01:30.209 --> 00:01:32.069 they've established this huge milestone i have 00:01:32.069 --> 00:01:34.890 a lot of thoughts i think this is a long time 00:01:34.890 --> 00:01:39.890 coming i'll say for a future topic i'm surprised 00:01:39.890 --> 00:01:43.409 that it took this long yeah i was really expecting 00:01:43.409 --> 00:01:45.689 them to launch it last year honestly even before 00:01:45.689 --> 00:01:48.090 that but i think it was so closely tied to their 00:01:48.090 --> 00:01:51.090 mid -band rollout that it took a little bit longer 00:01:51.549 --> 00:01:53.670 than I would have liked. They do have a very 00:01:53.670 --> 00:01:57.689 strong enterprise customer base, and they're 00:01:57.689 --> 00:01:59.450 going to have lots of business opportunities 00:01:59.450 --> 00:02:03.890 here. I think the one thing that people don't 00:02:03.890 --> 00:02:07.230 really realize is that application developers, 00:02:07.609 --> 00:02:10.789 ISVs, and a bunch of other companies that are 00:02:10.789 --> 00:02:13.949 doing software, they can't really optimize for 00:02:13.949 --> 00:02:17.169 standalone until there's enough carriers to support 00:02:17.169 --> 00:02:19.689 it. And when T -Mobile was going out on their 00:02:19.689 --> 00:02:22.689 loan, and being ahead of everybody it wasn't 00:02:22.689 --> 00:02:26.129 enough for application developers to want to 00:02:26.129 --> 00:02:29.129 build towards standalone and with the expectation 00:02:29.129 --> 00:02:31.949 that standalone would be available so i think 00:02:31.949 --> 00:02:34.569 we're getting to the point where maybe people 00:02:34.569 --> 00:02:37.349 actually start building for standalone more and 00:02:37.349 --> 00:02:41.960 obviously verizon needs to come to reality and 00:02:41.960 --> 00:02:45.060 launch 5GSA. But it's quite clear that AT &T 00:02:45.060 --> 00:02:47.159 was very close, like you said, with the standalone 00:02:47.159 --> 00:02:49.979 thing, REDCap, because you can't do REDCap without 00:02:49.979 --> 00:02:52.360 standalone. It's not possible. It was quite clear 00:02:52.360 --> 00:02:55.639 AT &T was close. I'm glad that they have gotten 00:02:55.639 --> 00:02:57.580 over it. Their fixed wireless services are also 00:02:57.580 --> 00:02:59.500 standalone only. It was quite clear they were 00:02:59.500 --> 00:03:01.759 close, but they hadn't reached that nationwide 00:03:01.759 --> 00:03:04.780 claim yet. And yeah, I just think it's super 00:03:04.780 --> 00:03:07.020 important. It's going to be huge for business 00:03:07.020 --> 00:03:10.389 applications. I think we're going to see... a 00:03:10.389 --> 00:03:14.030 lot more use of the 5G network in ways that 4G 00:03:14.030 --> 00:03:17.330 didn't really deliver. A lot of people have this 00:03:17.330 --> 00:03:20.610 negative sentiment towards 5G because we didn't 00:03:20.610 --> 00:03:21.990 have standalone networks. So we weren't really 00:03:21.990 --> 00:03:25.430 able to create new applications and new use cases 00:03:25.430 --> 00:03:28.349 for 5G and also take advantage more of the network. 00:03:28.590 --> 00:03:32.710 So this is huge. This is really huge. And it's 00:03:32.710 --> 00:03:35.310 a really big milestone for AT &T. It's important 00:03:35.310 --> 00:03:37.509 for the industry. And now we just need Verizon 00:03:37.509 --> 00:03:39.229 to catch up because once they do, I think we're 00:03:39.229 --> 00:03:42.969 going to see, a lot of uptake on 5G, which might 00:03:42.969 --> 00:03:45.909 end up rolling into 6G, but it will have been 00:03:45.909 --> 00:03:48.849 5G who kind of started it. Yeah, no, I agree. 00:03:49.169 --> 00:03:52.930 And I think your insight on the developer community 00:03:52.930 --> 00:03:56.389 is spot on. There have been sandboxes available 00:03:56.389 --> 00:04:01.669 with 5G Core and 5G RAN, but now that... You've 00:04:01.669 --> 00:04:04.870 got AT &T declaring nationwide standalone. And 00:04:04.870 --> 00:04:07.870 I expect Verizon is going to be very close by 00:04:07.870 --> 00:04:09.830 the end of the year as well. But I think to your 00:04:09.830 --> 00:04:12.569 point, that's really going to provide the incentive 00:04:12.569 --> 00:04:16.610 for developers to really focus on it. And competition 00:04:16.610 --> 00:04:18.350 is a good thing. I always like to say competition 00:04:18.350 --> 00:04:21.720 breeds innovation. And I'm excited to see we're 00:04:21.720 --> 00:04:24.839 just starting to begin to see some of the red 00:04:24.839 --> 00:04:27.139 cap solutions in the market. And you've talked 00:04:27.139 --> 00:04:29.339 quite a bit about that portable modem that you 00:04:29.339 --> 00:04:32.259 have there and just the opportunity to extend 00:04:32.259 --> 00:04:35.459 connectivity to wearables and do that very cost 00:04:35.459 --> 00:04:37.300 effectively. I think it's going to be an exciting 00:04:37.300 --> 00:04:39.519 development as well. I'm good friends with Egal 00:04:39.519 --> 00:04:42.639 Bez that basically manages the entire network 00:04:42.639 --> 00:04:46.839 as the CTO for AT &T. So I just want to congratulate 00:04:46.839 --> 00:04:49.629 Egal. And the team, this is a huge milestone. 00:04:49.790 --> 00:04:52.350 And I'm hoping to spend more time with him in 00:04:52.350 --> 00:04:54.750 the future just to learn about the roadmap. You 00:04:54.750 --> 00:04:58.490 and I had that opportunity with T -Mobile several 00:04:58.490 --> 00:05:01.350 weeks ago. Now, a lot of that, most of it, if 00:05:01.350 --> 00:05:03.329 not all of it, we cannot share because that's 00:05:03.329 --> 00:05:06.670 under nondisclosure. But suffice it to say, I 00:05:06.670 --> 00:05:07.709 think you're going to see some pretty exciting 00:05:07.709 --> 00:05:10.540 things from T -Mobile. And I do agree that with 00:05:10.540 --> 00:05:12.360 AT &T, we're going to see the same thing as well. 00:05:12.500 --> 00:05:14.180 But hey, let's move to your first topic. And 00:05:14.180 --> 00:05:17.459 you want to talk about Intel and some node advancements 00:05:17.459 --> 00:05:19.899 and also just some general updates on Panther 00:05:19.899 --> 00:05:23.579 Lake and Clearwater Forest. Yeah, these aren't 00:05:23.579 --> 00:05:26.860 inherently 5G products. Yeah, they're CPUs, right? 00:05:27.000 --> 00:05:29.579 Yeah. Yeah. So interestingly enough, Panther 00:05:29.579 --> 00:05:32.399 Lake is a consumer product running on Intel's 00:05:32.399 --> 00:05:35.720 new 18A foundry, Fab 52 in Ocotillo, Arizona. 00:05:35.920 --> 00:05:39.990 So that helps bring leading edge. node manufacturing 00:05:39.990 --> 00:05:43.230 to the United States, but also Clearwater Forest 00:05:43.230 --> 00:05:49.329 is a server CPU and it is a 288 core CPU with 00:05:49.329 --> 00:05:51.589 tiny little cores that are highly efficient. 00:05:51.750 --> 00:05:55.449 And there was a demo of it helping run a 5G network. 00:05:55.689 --> 00:05:58.430 So this is going to be targeted towards telecom 00:05:58.430 --> 00:06:00.769 applications because of the really power efficient 00:06:00.769 --> 00:06:04.009 and numerous cores. So that's the 5G angle there, 00:06:04.050 --> 00:06:07.430 but also there will be Panther Lake. based systems 00:06:07.430 --> 00:06:10.449 that will have 5G modos attached to them. Intel 00:06:10.449 --> 00:06:13.110 will not provide them, but they will have partners 00:06:13.110 --> 00:06:17.230 using the PCIe lanes to do 5G PCs. But really 00:06:17.230 --> 00:06:19.850 what this means is that there's going to be a 00:06:19.850 --> 00:06:24.009 lot more performance and lower power, both in 00:06:24.009 --> 00:06:27.370 the server and at the client side for 5G. And 00:06:27.370 --> 00:06:30.310 they did show us some edge IoT applications as 00:06:30.310 --> 00:06:35.069 well. specifically for AI at the edge using Panther 00:06:35.069 --> 00:06:38.290 Lake for a more embedded application as opposed 00:06:38.290 --> 00:06:42.730 to something like a consumer PC. Robotics or 00:06:42.730 --> 00:06:46.050 edge gateways, but extremely low power, but also 00:06:46.050 --> 00:06:49.790 high performance with new GPU cores, new NPUs, 00:06:49.790 --> 00:06:52.449 and new CPU cores, which are built on 1880. A 00:06:52.449 --> 00:06:55.089 quick breakdown is Intel has this chiplet architecture 00:06:55.089 --> 00:06:57.889 where it will combine different types of cores 00:06:57.889 --> 00:07:01.000 from different foundries to... make the most 00:07:01.000 --> 00:07:03.060 cost efficient, high performance chip possible. 00:07:03.360 --> 00:07:06.519 This was kind of forced upon them based on their 00:07:06.519 --> 00:07:09.860 lack of leadership in Foundry, which is why 18A 00:07:09.860 --> 00:07:11.560 is such a big deal because they're actually able 00:07:11.560 --> 00:07:14.680 to use 18A to build their CPU cores for both 00:07:14.680 --> 00:07:17.480 products. And then also Mixed between Intel and 00:07:17.480 --> 00:07:20.980 TSMC for the GPU and some of the IO tiles as 00:07:20.980 --> 00:07:23.600 well. So it's very tiled, chiplet -like approach. 00:07:23.879 --> 00:07:26.120 And it's going to be good for everybody, I think, 00:07:26.120 --> 00:07:28.759 in terms of competition, cost, and also giving 00:07:28.759 --> 00:07:30.839 OEMs more choice in how they build their systems. 00:07:31.360 --> 00:07:34.319 Yeah. Hey, so did you do that tour with our colleague 00:07:34.319 --> 00:07:37.480 Matt Kimball? Yeah. So I was there longer than 00:07:37.480 --> 00:07:41.000 he was, actually. So I did the fab tour with 00:07:41.000 --> 00:07:43.459 him. The first day was client day. Second day 00:07:43.459 --> 00:07:47.240 was server day. Okay. Q &A and NDA discussions 00:07:47.240 --> 00:07:49.920 and panels and things like that. But yeah, it 00:07:49.920 --> 00:07:52.399 was a long day and I went straight from Hawaii 00:07:52.399 --> 00:07:55.920 home for less than 12 hours and then went to 00:07:55.920 --> 00:07:58.379 Arizona for three or four days. So that was my 00:07:58.379 --> 00:08:00.899 last trip of the year. Yeah, before you were 00:08:00.899 --> 00:08:03.399 expecting your new little one. Matt was quite 00:08:03.399 --> 00:08:06.579 impressed just walking some of these fabs and 00:08:06.579 --> 00:08:08.980 just the expanse. And one of the comments he 00:08:08.980 --> 00:08:12.649 made was just... the really high degrees of automation 00:08:12.649 --> 00:08:17.050 and he said that on his little one mile sprint 00:08:17.050 --> 00:08:20.870 around it may have been 18a that that he said 00:08:20.870 --> 00:08:23.209 that he could count the number of actual human 00:08:23.209 --> 00:08:27.029 beings on both of his hands yeah very highly 00:08:27.029 --> 00:08:30.550 i'll explain that the lower the process note 00:08:30.550 --> 00:08:34.029 the bigger the foundry the fewer the people yeah 00:08:34.029 --> 00:08:38.980 because the precision required needs more automation. 00:08:39.220 --> 00:08:41.820 They have these carriers going around with 25 00:08:41.820 --> 00:08:46.179 wafers each, 30 mile track around the foundry, 00:08:46.240 --> 00:08:50.039 walk through 52, which has these gigantic EUV 00:08:50.039 --> 00:08:53.320 ASML machines. They've got lines of them in rows 00:08:53.320 --> 00:08:56.500 because they want to have that capacity to build 00:08:56.500 --> 00:08:58.299 and they have room for expansion. It's also, 00:08:58.360 --> 00:09:01.759 it's like a class 10 clean room. So the air gets 00:09:01.759 --> 00:09:04.960 recirculated every six minutes and it's designed 00:09:04.960 --> 00:09:09.970 to be leading edge, top of the line for Intel 00:09:09.970 --> 00:09:11.990 or their customers. And that's really the message 00:09:11.990 --> 00:09:13.649 you're trying to send is like, they're ready 00:09:13.649 --> 00:09:16.669 to do business. And I think having these products 00:09:16.669 --> 00:09:19.570 come out and be successful will ultimately cause, 00:09:19.830 --> 00:09:22.830 give people cause to move to Intel as a foundry. 00:09:23.029 --> 00:09:27.690 Yeah. Yeah. I believe so too. And we need very 00:09:27.690 --> 00:09:31.759 efficient, high production semiconductor. capacity 00:09:31.759 --> 00:09:34.600 in manufacturing in the United States. We can't 00:09:34.600 --> 00:09:38.080 continue to rely on TSMC to do everything. I 00:09:38.080 --> 00:09:39.860 don't subscribe to the belief that Intel is too 00:09:39.860 --> 00:09:42.039 big to fail, but they're on a roll when you look 00:09:42.039 --> 00:09:44.259 at the most recent news with what's been going 00:09:44.259 --> 00:09:46.620 on with them. And the stock has done quite well. 00:09:46.779 --> 00:09:48.320 There have been some major changes that have 00:09:48.320 --> 00:09:50.720 had to be made. And unfortunately, we both know 00:09:50.720 --> 00:09:54.279 some folks that have had to retire early or move 00:09:54.279 --> 00:09:58.299 on. But it's going to be fun to watch. And I 00:09:58.299 --> 00:10:00.659 believe that Intel has a lot of potential. And 00:10:00.659 --> 00:10:02.039 I'm not going to say that they're too big to 00:10:02.039 --> 00:10:05.299 fail, but we desperately need what they can provide 00:10:05.299 --> 00:10:07.659 from a manufacturing and capacity perspective. 00:10:07.740 --> 00:10:09.940 And just domesticating the supply chain, right, 00:10:10.000 --> 00:10:14.519 given the warranted concerns with TSMC and Taiwan 00:10:14.519 --> 00:10:17.779 and escalating tensions with China, with all 00:10:17.779 --> 00:10:20.639 the geopolitics that are going on there. But 00:10:20.639 --> 00:10:22.240 I don't want to get too far into politics, man. 00:10:22.299 --> 00:10:24.179 We're going to stick to technology on this podcast. 00:10:24.580 --> 00:10:27.600 So I'm going to go to my second topic. And I 00:10:27.600 --> 00:10:30.940 caught the news recently about Vodafone. They're 00:10:30.940 --> 00:10:33.759 claiming that they have launched the first low 00:10:33.759 --> 00:10:36.779 Earth orbit satellite mobile broadband research 00:10:36.779 --> 00:10:41.429 hub. This is at the University of Malaga. And 00:10:41.429 --> 00:10:45.450 they are partnering, no surprise, with AST Space 00:10:45.450 --> 00:10:48.730 Mobile as well to do this. And as I read through 00:10:48.730 --> 00:10:51.909 the press release on the Vodafone website, it's 00:10:51.909 --> 00:10:54.429 interesting. The vision here is to streamline 00:10:54.429 --> 00:10:56.950 the process of testing and validating hardware 00:10:56.950 --> 00:10:59.309 and software required for basically combining 00:10:59.309 --> 00:11:04.070 non -terrestrial network solutions with terrestrial 00:11:04.070 --> 00:11:06.190 ones. There are other third parties that are 00:11:06.190 --> 00:11:08.490 going to be involved in this. This is great for 00:11:08.490 --> 00:11:11.490 Europe, from my perspective. Obviously, when 00:11:11.490 --> 00:11:13.509 you look at the infrastructure providers, you've 00:11:13.509 --> 00:11:15.889 got Nokia, you've got Ericsson, they're obviously 00:11:15.889 --> 00:11:18.909 European as well. But I really like what Vodafone's 00:11:18.909 --> 00:11:20.250 doing. Obviously, they have a vested interest 00:11:20.250 --> 00:11:23.769 in this because they have invested in AST Space 00:11:23.769 --> 00:11:27.090 Mobile. I'm still sticking by my guns with the 00:11:27.090 --> 00:11:30.210 assertion that I do believe that the AST Space 00:11:30.210 --> 00:11:34.509 Mobile overall tech stack is superior to what 00:11:34.509 --> 00:11:36.330 exists in the market today. That doesn't mean 00:11:36.330 --> 00:11:38.779 it can't change tomorrow. There have been some 00:11:38.779 --> 00:11:41.820 delays in getting their birds in the air. They're 00:11:41.820 --> 00:11:43.779 working through that right now. I think I've 00:11:43.779 --> 00:11:47.000 mentioned on a podcast a few weeks ago that I'm 00:11:47.000 --> 00:11:48.980 hoping to have some additional time with Abel 00:11:48.980 --> 00:11:51.120 and his team in Midland. And I'll even be in 00:11:51.120 --> 00:11:52.820 Barcelona in December, and I know that they have 00:11:52.820 --> 00:11:55.139 an operation there as well. But hoping to get 00:11:55.139 --> 00:11:56.519 some time with him and the team just to get an 00:11:56.519 --> 00:11:59.340 update on where they're headed and their vision 00:11:59.340 --> 00:12:01.539 for what's going on. But I don't know if you 00:12:01.539 --> 00:12:03.559 caught this news. Any thoughts before we move 00:12:03.559 --> 00:12:06.419 to your second topic? I didn't really catch this 00:12:06.419 --> 00:12:08.750 news, but... It's not really a surprise considering 00:12:08.750 --> 00:12:11.870 Vodafone's interactions with AST Space Mobile. 00:12:12.090 --> 00:12:14.690 I think it makes sense for them to have something 00:12:14.690 --> 00:12:17.570 like this established, but it will be interesting 00:12:17.570 --> 00:12:21.129 to see how this evolves and how it expands within 00:12:21.129 --> 00:12:24.750 Europe. Yeah, I agree. I agree. Let's go to your 00:12:24.750 --> 00:12:27.389 second topic. And I'm going to talk about Verizon 00:12:27.389 --> 00:12:30.049 too, but I'm going to let you start off. And 00:12:30.049 --> 00:12:31.870 you want to talk about a strategic acquisition 00:12:31.870 --> 00:12:34.309 and what that might mean for the company, I believe 00:12:34.309 --> 00:12:36.450 for its fixed wireless access service, right? 00:12:37.240 --> 00:12:43.059 Yes, Verizon acquired Starry. And Starry is a 00:12:43.059 --> 00:12:47.159 fixed wireless provider. Not to be confused with 00:12:47.159 --> 00:12:50.240 the lemon -lime soda drink that I see in Vegas 00:12:50.240 --> 00:12:55.100 all the time. Yeah, and Starry is still fairly 00:12:55.100 --> 00:12:59.559 small. They're mostly focused on multi -dwelling 00:12:59.559 --> 00:13:04.720 units in Boston, Denver, LA, New York, and DC. 00:13:05.370 --> 00:13:08.830 yeah and it also recently pulled out of columbus 00:13:08.830 --> 00:13:11.289 ohio they're not really in the best financial 00:13:11.289 --> 00:13:13.830 situation yeah because they laid off half their 00:13:13.830 --> 00:13:16.549 workers in 2022 and filed for bankruptcy in 2023 00:13:16.549 --> 00:13:21.610 they currently serve 100 000 customers and was 00:13:21.610 --> 00:13:24.730 that you or me that's fire alarm testing yeah 00:13:24.730 --> 00:13:29.250 during a podcast of course but yeah they're fairly 00:13:29.250 --> 00:13:32.490 small but verizon thinks that they have the ability 00:13:32.490 --> 00:13:36.039 to grow their fixed wireless space to 8 million 00:13:36.039 --> 00:13:39.159 customers, 8 to 9 million customers using Starry 00:13:39.159 --> 00:13:42.899 as like a launchpad because multi -dwelling units 00:13:42.899 --> 00:13:45.120 are a great way to get lots of customers quickly. 00:13:45.460 --> 00:13:47.759 But it seems like maybe Starry didn't really 00:13:47.759 --> 00:13:51.279 have the resources to grow in a way that I believe 00:13:51.279 --> 00:13:53.759 Verizon could actually help them. And the truth 00:13:53.759 --> 00:13:55.659 is that Verizon's fixed wireless business has 00:13:55.659 --> 00:13:58.340 been good, but it's not been as great as T -Mobile's 00:13:58.340 --> 00:14:01.120 and they've lagged behind. So I think it's a 00:14:01.120 --> 00:14:03.490 good opportunity for them to grow. ironically 00:14:03.490 --> 00:14:05.809 their fixed wireless business is the oldest because 00:14:05.809 --> 00:14:07.610 they started it with millimeter wave if you remember 00:14:07.610 --> 00:14:11.149 that yeah and yeah i think there's a this is 00:14:11.149 --> 00:14:15.309 just a natural synergy here they didn't say what 00:14:15.309 --> 00:14:18.049 the acquisition price was but i can't imagine 00:14:18.049 --> 00:14:19.629 was very much they probably have a lot of debt 00:14:19.629 --> 00:14:22.549 still yeah and the import interesting thing is 00:14:22.549 --> 00:14:25.649 that they say they expect this acquisition to 00:14:25.649 --> 00:14:27.289 be expected to be closed by the first quarter 00:14:27.289 --> 00:14:31.450 of 2026 but i think that there's good synergy 00:14:31.450 --> 00:14:35.070 and i think star customers will probably be happier 00:14:35.070 --> 00:14:36.889 that they have a bigger more reliable company 00:14:36.889 --> 00:14:41.490 behind their ifp and yeah it's 100 000 customers 00:14:41.490 --> 00:14:44.870 isn't much but i could see verizon easily doubling 00:14:44.870 --> 00:14:47.129 that or tripling that in that first year or two 00:14:47.129 --> 00:14:49.990 just because they already have a footprint in 00:14:49.990 --> 00:14:52.090 place and verizon has lots of spectrum as well 00:14:52.090 --> 00:14:55.309 yeah i saw the news but i didn't dig into it 00:14:55.309 --> 00:14:58.190 so this is clearly this is not verizon acquiring 00:14:58.190 --> 00:15:01.580 this company for any sort of intellectual property 00:15:01.580 --> 00:15:03.659 it's acquiring this company as a distribution 00:15:03.659 --> 00:15:09.179 Channel right yeah and they specifically said 00:15:09.179 --> 00:15:12.460 this architecture is less expensive to build 00:15:12.460 --> 00:15:16.460 so they're using stars architecture which is 00:15:16.460 --> 00:15:18.879 low cost so there is some like so there's some 00:15:18.879 --> 00:15:21.980 IP and so they say this architecture is less 00:15:21.980 --> 00:15:24.580 expensive to build quicker to deploy and uniquely 00:15:24.580 --> 00:15:27.620 addresses the complexities of urban settings 00:15:28.159 --> 00:15:30.419 where we can leverage our existing fiber and 00:15:30.419 --> 00:15:33.220 millimeter wave assets. So they're looking for, 00:15:33.299 --> 00:15:35.899 this is like a good, easy way for them to get 00:15:35.899 --> 00:15:39.220 that last mile cheaply, leveraging their already 00:15:39.220 --> 00:15:41.899 existing network, because let's be honest, Verizon's 00:15:41.899 --> 00:15:46.019 present in most major markets. This would be 00:15:46.019 --> 00:15:50.279 a way for them to roll out these MDU, multi -tooling 00:15:50.279 --> 00:15:54.419 unit, fixed wireless solutions that are cheaper 00:15:54.419 --> 00:15:57.019 and they can... charge less yeah no that okay 00:15:57.019 --> 00:15:59.120 so i get it so it sounds like it's some sort 00:15:59.120 --> 00:16:03.360 of mesh it's probably got a radio and then yes 00:16:03.360 --> 00:16:07.039 the cpes are probably less expensive and yeah 00:16:07.039 --> 00:16:09.740 and this kind of reminds me when google fiber 00:16:09.740 --> 00:16:12.240 bought web pass i think it's very similar right 00:16:12.240 --> 00:16:15.250 where it's like you're buying a fairly small 00:16:15.250 --> 00:16:17.269 customer base but you're also buying technology 00:16:17.269 --> 00:16:19.750 that's cheaper and easier to deploy got it got 00:16:19.750 --> 00:16:22.429 it and i think yeah okay that makes sense and 00:16:22.429 --> 00:16:24.570 but i think yeah this could potentially be from 00:16:24.570 --> 00:16:27.289 a distribution perspective yeah verizon could 00:16:27.289 --> 00:16:30.110 pump some bucks into this and yeah and to your 00:16:30.110 --> 00:16:32.629 point support it and it solves their last mile 00:16:32.629 --> 00:16:35.529 challenge and i like it okay so it's a combination 00:16:35.529 --> 00:16:38.350 of some ip that's definitely a distribution channel 00:16:38.350 --> 00:16:40.529 play as well cool Yeah, it'll be interesting 00:16:40.529 --> 00:16:43.629 to see. Verizon has lagged, not only with its 00:16:43.629 --> 00:16:47.090 standalone rollout with 5G, but it's had a big 00:16:47.090 --> 00:16:51.350 hole, obviously, in its spectrum footprint. They 00:16:51.350 --> 00:16:53.470 filled that. They did that very expensively, 00:16:53.470 --> 00:16:56.529 as you and I have discussed on numerous podcasts, 00:16:56.769 --> 00:16:59.529 but interesting stuff. I'm going to end with 00:16:59.529 --> 00:17:02.230 my third topic about Verizon, too, and I want 00:17:02.230 --> 00:17:06.009 to talk about Hans Vestberg's somewhat hasty 00:17:06.009 --> 00:17:10.019 exit. One of my favorite journalists, Ian Morris, 00:17:10.099 --> 00:17:11.960 that writes for Light Reading, he's a cheeky 00:17:11.960 --> 00:17:15.039 Brit and he's always got something controversial 00:17:15.039 --> 00:17:18.539 to say, but he posted something on Light Reading 00:17:18.539 --> 00:17:21.839 recently. It's talked about, hey, this was a 00:17:21.839 --> 00:17:25.460 pretty sudden announcement and there wasn't a 00:17:25.460 --> 00:17:28.559 lot of information provided other than it was 00:17:28.559 --> 00:17:31.440 just time for the company and the AIR to move 00:17:31.440 --> 00:17:33.960 on. And what he did is he went back and looked 00:17:33.960 --> 00:17:37.269 at... Several years ago, when Hans was named 00:17:37.269 --> 00:17:40.289 CEO, there was a lead up. So it wasn't super 00:17:40.289 --> 00:17:43.630 sudden. And we have talked about the transition 00:17:43.630 --> 00:17:47.289 at the CEO position at T -Mobile, how that's 00:17:47.289 --> 00:17:49.849 been fairly telegraphed out. There appeared to 00:17:49.849 --> 00:17:53.109 be a succession plan around that. It's not out 00:17:53.109 --> 00:17:54.930 of the ordinary, right, given their cadence. 00:17:55.710 --> 00:17:58.309 What Ian talks about in his article is that when 00:17:58.309 --> 00:18:01.150 Hans took over several years ago, it was telegraphed 00:18:01.150 --> 00:18:02.750 out. It was orchestrated. There was a lot of 00:18:02.750 --> 00:18:05.890 time. He's intimating that maybe Hans, maybe 00:18:05.890 --> 00:18:07.589 he was pushed out for other reasons. And it's 00:18:07.589 --> 00:18:09.410 interesting. You and I have talked about him. 00:18:09.869 --> 00:18:12.329 Hey, listen, I've met him. He's a very charismatic 00:18:12.329 --> 00:18:16.710 guy. He's got friends in very high places. And 00:18:16.710 --> 00:18:18.609 I won't name the infrastructure company, but 00:18:18.609 --> 00:18:21.190 there's someone at a very large cellular infrastructure 00:18:21.190 --> 00:18:24.009 company that he and - Where did he come from? 00:18:24.440 --> 00:18:26.579 Yeah, yeah, exactly. And where he came from as 00:18:26.579 --> 00:18:29.579 well, but has remained friends there. But long 00:18:29.579 --> 00:18:32.339 and short of it, Verizon's missed. They've missed 00:18:32.339 --> 00:18:34.500 on their execution with their 5G deployment. 00:18:34.920 --> 00:18:37.359 They went from hero, I don't want to say hero 00:18:37.359 --> 00:18:41.319 to zero. In the LTE world, they absolutely, it's 00:18:41.319 --> 00:18:43.759 hard to argue that they didn't have the most 00:18:43.759 --> 00:18:47.140 reliable, the most expansive LTE network. That's 00:18:47.140 --> 00:18:50.700 not been the case with 5G. And I follow the financial 00:18:50.700 --> 00:18:53.730 somewhat with Verizon. They've been, and you 00:18:53.730 --> 00:18:55.990 and I have covered their financials on many podcasts. 00:18:56.349 --> 00:18:58.910 They've been good, but they haven't been like 00:18:58.910 --> 00:19:01.329 just kick ass, knock it out of the park, like 00:19:01.329 --> 00:19:04.210 T -Mobile or even AT &T as well. When you like, 00:19:04.210 --> 00:19:07.210 especially look with the mobility ads that T 00:19:07.210 --> 00:19:10.450 -Mobile has been able to bring in with its investment 00:19:10.450 --> 00:19:12.930 in its 5G standalone network. And when you look 00:19:12.930 --> 00:19:17.569 at the investment that John Stanky. and team 00:19:17.569 --> 00:19:20.849 have made it to AT &T with respect to basically 00:19:20.849 --> 00:19:24.269 overbuilding fiber, going into parts of the country 00:19:24.269 --> 00:19:27.769 that were underserved and basically investing 00:19:27.769 --> 00:19:32.150 forward and not expecting the typical ROI that 00:19:32.150 --> 00:19:36.170 they see on fiber deployments in less dense populated 00:19:36.170 --> 00:19:38.730 areas. There's a lot to be said here. So it'll 00:19:38.730 --> 00:19:40.609 be interesting. I don't know a lot about the 00:19:40.609 --> 00:19:45.609 Verizon, the new Verizon CEO, but - Yeah, he's 00:19:45.609 --> 00:19:48.400 ex -PayPal. PayPal's had its ups and downs as 00:19:48.400 --> 00:19:51.579 well. But yeah, it'll be interesting to see if 00:19:51.579 --> 00:19:55.299 Dan Schulman can turn, I don't want to say turn 00:19:55.299 --> 00:19:57.480 things around, but accelerate the company on 00:19:57.480 --> 00:19:59.420 the right trajectory. So what do you think? Am 00:19:59.420 --> 00:20:02.339 I being overly harsh? Is Ian being overly harsh 00:20:02.339 --> 00:20:09.660 about Hans? No. Hans is a CEO, was CEO, correction. 00:20:10.299 --> 00:20:16.039 He should have pivoted. when it was clear that 00:20:16.039 --> 00:20:18.819 the millimeter wave push was not it. Was not 00:20:18.819 --> 00:20:20.440 going to happen. That it was not going to be 00:20:20.440 --> 00:20:25.420 economically viable, right? Correct. The problem 00:20:25.420 --> 00:20:28.259 is they doubled down on it. Instead of just saying, 00:20:28.380 --> 00:20:30.299 we need spectrum, let's get an auction going. 00:20:30.759 --> 00:20:32.740 Because that's the truth. It's like they needed 00:20:32.740 --> 00:20:35.859 the spectrum. They weren't going to be able to 00:20:35.859 --> 00:20:37.400 buy Sprint. No one was ever going to let them 00:20:37.400 --> 00:20:40.440 buy Sprint. TNT tried to buy T -Mobile. That 00:20:40.440 --> 00:20:43.690 didn't happen, if you remember. they couldn't 00:20:43.690 --> 00:20:45.089 acquire their way out of the situation so they 00:20:45.089 --> 00:20:47.390 clearly needed a spectrum auction the c -band 00:20:47.390 --> 00:20:50.650 auction was the answer to that question admittedly 00:20:50.650 --> 00:20:55.190 it came late so i can't fully blame him for not 00:20:55.190 --> 00:20:58.829 having the spectrum that he needed but also that 00:20:58.829 --> 00:21:01.289 there were other ways there are other paths and 00:21:01.289 --> 00:21:03.910 they fumbled millimeter wave they fumbled dss 00:21:03.910 --> 00:21:06.369 they really pissed off a lot of their customers 00:21:06.369 --> 00:21:08.730 and i think that's really the ultimate problem 00:21:08.730 --> 00:21:12.940 is instead of being open with their customers. 00:21:13.140 --> 00:21:15.079 They were just admitting, they were basically 00:21:15.079 --> 00:21:17.420 saying that they've got the fastest, best network 00:21:17.420 --> 00:21:21.140 and they just did it. And with mid band, I think 00:21:21.140 --> 00:21:23.420 in C band, they're doing a lot better. I just 00:21:23.420 --> 00:21:25.700 actually was at the park with my daughter and 00:21:25.700 --> 00:21:28.420 I ran a speed test in the park and I was getting 00:21:28.420 --> 00:21:31.079 over 500 megabits per second. Although in the 00:21:31.079 --> 00:21:33.779 same spot, I got 800 on T -Mobile. They're still 00:21:33.779 --> 00:21:35.720 not in that leadership position, but they've 00:21:35.720 --> 00:21:37.359 plugged a lot of the holes that they had before. 00:21:37.759 --> 00:21:40.299 I just think that honestly, I think Hans should 00:21:40.299 --> 00:21:43.619 have been removed a while ago. And I don't know 00:21:43.619 --> 00:21:45.279 if that's more harsh or less harsh than you, 00:21:45.380 --> 00:21:47.900 but I think that this was a long time coming 00:21:47.900 --> 00:21:50.799 actually. And the timing is surprising mostly 00:21:50.799 --> 00:21:54.519 because like why now? Honestly, it's funny that 00:21:54.519 --> 00:21:57.400 AT &T announced their standalone network like 00:21:57.400 --> 00:21:59.900 the day after, which to me is maybe they fired 00:21:59.900 --> 00:22:01.339 him because they heard AT &T is going to launch 00:22:01.339 --> 00:22:03.299 standalone. I don't think that's actually happening, 00:22:03.420 --> 00:22:07.160 but what it does is it firmly puts Verizon in 00:22:07.160 --> 00:22:09.930 third place in terms of. Oh yeah. What their 00:22:09.930 --> 00:22:12.269 network can do. That said, if you look at a lot 00:22:12.269 --> 00:22:14.589 of the metrics, Verizon's network is a little 00:22:14.589 --> 00:22:17.130 bit faster than AT &T's in a lot of ways. But 00:22:17.130 --> 00:22:18.809 I just think AT &T's network is more capable. 00:22:19.289 --> 00:22:21.589 Another thing is Verizon really invested a lot 00:22:21.589 --> 00:22:24.109 in edge compute. But what are people going to 00:22:24.109 --> 00:22:27.150 do with edge compute if the network isn't capable 00:22:27.150 --> 00:22:29.849 of leveraging that? I think they made a lot of 00:22:29.849 --> 00:22:34.160 the wrong bets and they did not pay off. And 00:22:34.160 --> 00:22:36.380 that's why we are where we are today and why 00:22:36.380 --> 00:22:39.119 the CEO has been replaced. Yeah. Yeah. They just 00:22:39.119 --> 00:22:41.240 they hung their hat on millimeter wave way too 00:22:41.240 --> 00:22:43.980 long. And that was the spectrum that they had 00:22:43.980 --> 00:22:46.759 in their toolbox. Right. But they should have 00:22:46.759 --> 00:22:48.759 let it go. Now, a lot of this was like the FCC 00:22:48.759 --> 00:22:52.539 was dysfunctional and the whole management of 00:22:52.539 --> 00:22:54.839 spectrum farming and that sort of thing has needed 00:22:54.839 --> 00:22:57.960 a total overhaul. And so they had to wait for 00:22:57.960 --> 00:23:00.779 C -band. And then when they got it, they paid 00:23:00.779 --> 00:23:03.259 an astronomical amount of money for it. right? 00:23:03.359 --> 00:23:07.319 Like $40 billion of my recollection serves me. 00:23:07.319 --> 00:23:10.660 52 .9. Okay. I was off by that. With clearing 00:23:10.660 --> 00:23:13.259 costs. With clearing costs on top of that. So 00:23:13.259 --> 00:23:15.660 anyway, we're not going to beat this horse anymore. 00:23:16.279 --> 00:23:19.220 So let's wrap things up with your third topic. 00:23:19.259 --> 00:23:22.859 And you want to frame like all these big AI deals 00:23:22.859 --> 00:23:24.819 that we're seeing and what that means for 5G 00:23:24.819 --> 00:23:26.819 and 6G. And it's interesting that you picked 00:23:26.819 --> 00:23:30.259 this as a topic because Matt Kimball and Paul 00:23:30.259 --> 00:23:33.680 and I recorded our data center podcast this morning, 00:23:33.839 --> 00:23:38.359 and we led with the OpenAI AMD deal and how unconventional 00:23:38.359 --> 00:23:41.200 that is. And so I'm really curious to see how 00:23:41.200 --> 00:23:44.720 you're going to tie that to 5G and 6G. I think 00:23:44.720 --> 00:23:45.980 I know where you're going, but why don't you 00:23:45.980 --> 00:23:48.940 share that? I'm looking at it from the perspective 00:23:48.940 --> 00:23:52.259 of, of course, 5G and 6G, but really it's okay. 00:23:53.240 --> 00:23:56.759 OpenAI is driving a lot of this right now, right? 00:23:57.289 --> 00:24:01.690 whether it's Nvidia, AMD, Oracle, you name it, 00:24:01.710 --> 00:24:03.750 right? Like everybody's getting into the party. 00:24:03.869 --> 00:24:06.410 And the truth is like what a lot of this is focused 00:24:06.410 --> 00:24:10.890 on is agentic AI. And I'm thinking about like, 00:24:10.990 --> 00:24:17.309 how do these deals affect 5G and 6G one in terms 00:24:17.309 --> 00:24:20.230 of how, what does the architecture look like 00:24:20.230 --> 00:24:23.269 to get this compute closer to the users to give 00:24:23.269 --> 00:24:26.039 them a good experience? like how much edge compute 00:24:26.039 --> 00:24:29.720 does open ai need to invest in and what does 00:24:29.720 --> 00:24:32.599 that edge compute look like and then also how 00:24:32.599 --> 00:24:36.359 do the carriers work with open ai to leverage 00:24:36.359 --> 00:24:40.380 this capability because right now the only carrier 00:24:40.380 --> 00:24:44.259 that's like explicitly working with open ai is 00:24:44.259 --> 00:24:47.380 t -mobile verizon has like a me too announcement 00:24:47.380 --> 00:24:50.759 as well but it was very clear that sam altman 00:24:51.160 --> 00:24:54.059 showed up at the T -Mobile. That was strategic. 00:24:54.220 --> 00:24:56.720 That was strategic. I completely agree with you, 00:24:56.740 --> 00:24:59.299 man. I completely agree. So I'm thinking about, 00:24:59.359 --> 00:25:02.599 like, how do all of these investments in OpenAI 00:25:02.599 --> 00:25:06.319 and clearly how OpenAI is accelerating its plans, 00:25:06.519 --> 00:25:10.160 how does that affect what's going on with NVIDIA 00:25:10.160 --> 00:25:12.599 and OpenAI? And there's going to be a little 00:25:12.599 --> 00:25:15.750 piece of compute left over. that's not running 00:25:15.750 --> 00:25:18.869 the network, not running the RAN, not doing orchestration, 00:25:18.950 --> 00:25:21.710 but can be utilized for applications that will 00:25:21.710 --> 00:25:24.349 be at the edge. And what does that look like? 00:25:24.390 --> 00:25:26.589 And I would love to see OpenAI and T -Mobile 00:25:26.589 --> 00:25:28.890 talk about that a little more. They haven't really 00:25:28.890 --> 00:25:31.750 since the capital markets day, but I think that 00:25:31.750 --> 00:25:34.150 everybody's betting on OpenAI right now. I don't 00:25:34.150 --> 00:25:37.349 think they're the only dog to bet on, but I think 00:25:37.349 --> 00:25:39.950 that they're clearly pushing the envelope in 00:25:39.950 --> 00:25:43.140 everywhere that they can. yeah i agree and just 00:25:43.140 --> 00:25:46.420 the creativity of this deal with amd for six 00:25:46.420 --> 00:25:49.599 gigawatts of gpu that's massive that's billions 00:25:49.599 --> 00:25:53.019 of dollars potentially for amd and then amd in 00:25:53.019 --> 00:25:56.920 turn issuing equity as i read through the deal 00:25:56.920 --> 00:25:59.960 and our fearless leader patrick moorhead has 00:25:59.960 --> 00:26:02.180 weighed in on on social media on this as well 00:26:02.180 --> 00:26:04.380 there's certain triggers that have to be met 00:26:04.380 --> 00:26:07.519 for that equity to flow but what i really like 00:26:07.519 --> 00:26:10.849 about it is It allows AMD that's been in kind 00:26:10.849 --> 00:26:12.769 of second position, not that they've been in 00:26:12.769 --> 00:26:15.470 second position relative to NVIDIA on the compute 00:26:15.470 --> 00:26:19.730 side. It provides them some tailwind to really 00:26:19.730 --> 00:26:22.990 drive the volume on their GPU platform and what 00:26:22.990 --> 00:26:25.690 they're in the GPU platform because they're accelerators 00:26:25.690 --> 00:26:29.869 that leverage Nsando as well. And like I always 00:26:29.869 --> 00:26:32.950 like to say, competition breeds innovation. And 00:26:32.950 --> 00:26:35.390 what this means is that there's going to be more 00:26:35.390 --> 00:26:37.660 choice. And to your point, There are lots of 00:26:37.660 --> 00:26:40.619 other models. There's OpenAI, there's IBM Granite, 00:26:40.680 --> 00:26:43.640 there's ChatGPT. But I really like this deal. 00:26:43.700 --> 00:26:46.599 We live in very unconventional times. And that 00:26:46.599 --> 00:26:48.319 was what I shared on the Data Center podcast 00:26:48.319 --> 00:26:51.299 earlier this morning. But I do like it. And I 00:26:51.299 --> 00:26:54.319 agree with you that there is a potential impact 00:26:54.319 --> 00:26:58.519 on telecommunications services. Not the infrastructure, 00:26:58.799 --> 00:27:01.099 we've talked about that, AI RAN, that sort of 00:27:01.099 --> 00:27:03.599 thing, but really on the applications and the 00:27:03.599 --> 00:27:06.500 services. And I'm glad I had forgotten about 00:27:06.500 --> 00:27:09.559 the OpenAI T -Mobile announcement. We need to 00:27:09.559 --> 00:27:12.059 hear some more about how that is progressing, 00:27:12.220 --> 00:27:15.000 right? Because it has a lot of potential. It 00:27:15.000 --> 00:27:18.039 was a little squishy. There weren't a ton of 00:27:18.039 --> 00:27:20.400 concrete details, but that's how partnerships 00:27:20.400 --> 00:27:25.359 start. I can't imagine that partnership is just 00:27:25.359 --> 00:27:29.549 like... on paper because they had jensen they 00:27:29.549 --> 00:27:32.569 had sam altman like you don't get those guys 00:27:32.569 --> 00:27:35.710 to show up to your capital markets day to talk 00:27:35.710 --> 00:27:38.210 about what they're doing in ai unless they're 00:27:38.210 --> 00:27:40.170 actually doing something because that could be 00:27:40.170 --> 00:27:42.650 a press release yeah they could send someone 00:27:42.650 --> 00:27:45.309 who's not the ceo so it was very clear that there's 00:27:45.309 --> 00:27:47.250 something going on there And I hope we get an 00:27:47.250 --> 00:27:49.690 update soon. Yeah, I do too, buddy. Hey, we went 00:27:49.690 --> 00:27:51.190 a little bit long, but it was a great conversation. 00:27:51.410 --> 00:27:53.069 Why don't you take us home? Absolutely. We hope 00:27:53.069 --> 00:27:54.710 our viewers and listeners found this week's topic 00:27:54.710 --> 00:27:56.589 interesting. If anyone would like to provide 00:27:56.589 --> 00:27:59.769 a topic for a future podcast, please reach out 00:27:59.769 --> 00:28:02.049 to us on social media. Will's at WillTownTech 00:28:02.049 --> 00:28:04.470 and I'm at Anshul Saag. We hope you have a great 00:28:04.470 --> 00:28:06.730 weekend and please tune in again next week. And 00:28:06.730 --> 00:28:08.130 don't forget to rate us and subscribe.