The Ciena Insights Podcast

1 00:00:08,870 --> 00:00:11,610 Hello everyone, and welcome to 2 00:00:11,610 --> 00:00:13,784 the Ciena Insights podcast. 3 00:00:13,784 --> 00:00:16,079 I'm Rebecca Smith, senior vice president of 4 00:00:16,079 --> 00:00:19,274 global marketing and communications here at Ciena. 5 00:00:19,274 --> 00:00:24,014 We are going to be discussing OFC 2026, which, 6 00:00:24,014 --> 00:00:25,574 as most of our listeners know, 7 00:00:25,574 --> 00:00:27,210 is the largest and most 8 00:00:27,210 --> 00:00:29,505 influential optical networking event 9 00:00:29,505 --> 00:00:31,349 in our industry. 10 00:00:31,349 --> 00:00:33,389 It's always been where you go to get 11 00:00:33,389 --> 00:00:36,330 a pulse check on networking innovation. 12 00:00:36,330 --> 00:00:39,990 This year the momentum was unmistakable. 13 00:00:39,990 --> 00:00:42,059 It's clear the industry is entering 14 00:00:42,059 --> 00:00:45,089 a new phase where networking, compute, 15 00:00:45,089 --> 00:00:48,194 and AI are all coming together with 16 00:00:48,194 --> 00:00:51,435 optical innovation at the front and center of it all. 17 00:00:51,435 --> 00:00:53,385 This gives us a lot to unpack today, 18 00:00:53,385 --> 00:00:55,680 and I can't think of two better people to do 19 00:00:55,680 --> 00:00:58,410 that with than my esteemed colleagues, 20 00:00:58,410 --> 00:01:00,645 Brodie Gage, senior vice president 21 00:01:00,645 --> 00:01:02,955 of global products and supply chain, 22 00:01:02,955 --> 00:01:07,695 and Dino DiPerna, senior vice president of global R&D. 23 00:01:07,695 --> 00:01:09,315 To get us started, 24 00:01:09,315 --> 00:01:12,179 could each of you briefly share a bit about your role 25 00:01:12,179 --> 00:01:15,134 at Ciena and your journey with the company? 26 00:01:15,134 --> 00:01:16,829 Brodie – Why don't we start with you? 27 00:01:16,829 --> 00:01:19,185 Sure. I've been with Ciena 28 00:01:19,185 --> 00:01:22,740 plus Nortel now for about 28 years. 29 00:01:22,740 --> 00:01:26,129 I started at Nortel in 1998 in engineering. 30 00:01:26,129 --> 00:01:28,980 I would say this OFC was very similar to 31 00:01:28,980 --> 00:01:32,835 the first OFC I went to back in 2000 or 2001. 32 00:01:32,835 --> 00:01:34,170 I started with Nortel in engineering, 33 00:01:34,170 --> 00:01:35,370 moved into product management, 34 00:01:35,370 --> 00:01:38,730 and joined Ciena in March 2010 35 00:01:38,730 --> 00:01:41,640 when Ciena bought the MEN assets off of Nortel. 36 00:01:41,640 --> 00:01:43,440 I've been in product management on 37 00:01:43,440 --> 00:01:46,230 optical basically my entire career and recently 38 00:01:46,230 --> 00:01:49,364 added the supply chain function as well in October 2023. 39 00:01:49,364 --> 00:01:51,420 Awesome. Dino, I think you’ve had 40 00:01:51,420 --> 00:01:53,715 a similar journey coming into Ciena. 41 00:01:53,715 --> 00:01:55,784 Similar, but with a few more years 42 00:01:55,784 --> 00:01:57,480 on the front end of that journey. 43 00:01:57,480 --> 00:02:00,299 Although like Brodie, we have 44 00:02:00,299 --> 00:02:03,750 16 years – very great years by the way – with Ciena. 45 00:02:03,750 --> 00:02:06,674 We joined in 2010. 46 00:02:06,674 --> 00:02:10,839 In fact, day one was actually at OFC. 47 00:02:10,839 --> 00:02:15,364 I remember that the booth was great. 48 00:02:15,364 --> 00:02:17,839 Nothing compared to what we have today, Rebecca, 49 00:02:17,839 --> 00:02:19,070 but they had a box of 50 00:02:19,070 --> 00:02:21,020 business cards waiting for me as soon as I 51 00:02:21,020 --> 00:02:24,434 got in there and that started the journey with Ciena. 52 00:02:24,434 --> 00:02:26,610 Like Brodie, my roots go back 53 00:02:26,610 --> 00:02:29,294 to Bell-Northern Research and Nortel. 54 00:02:29,294 --> 00:02:31,770 I’ve got quite a few more years than Brodie does. 55 00:02:31,770 --> 00:02:34,215 I think I've clocked over four decades now. 56 00:02:34,215 --> 00:02:36,030 I met some of my colleagues back 57 00:02:36,030 --> 00:02:38,144 in the latter part of the 80s. 58 00:02:38,144 --> 00:02:40,814 That's what dragged me to Ottawa and to BNR. 59 00:02:40,814 --> 00:02:43,680 For that, I'm eternally grateful for the journey through 60 00:02:43,680 --> 00:02:47,234 all of the characters and personalities over those years. 61 00:02:47,234 --> 00:02:49,290 Well, we are very grateful and lucky to 62 00:02:49,290 --> 00:02:51,284 have you both at Ciena. 63 00:02:51,284 --> 00:02:53,805 You mentioned OFC was 64 00:02:53,805 --> 00:02:56,069 technically your first day at Ciena. 65 00:02:56,069 --> 00:02:59,040 I want to know how many OFCs have each of you been to? 66 00:02:59,040 --> 00:03:01,330 You want to go first, Brodie? 67 00:03:01,400 --> 00:03:03,810 I think the first one I went 68 00:03:03,810 --> 00:03:07,634 to was either in 2000 or 2001. 69 00:03:07,634 --> 00:03:09,210 That gives me 25 years, 70 00:03:09,210 --> 00:03:11,670 roughly, but I missed a few years. 71 00:03:11,670 --> 00:03:15,779 I’ve probably been to 17 OFCs. 72 00:03:15,779 --> 00:03:18,719 I was just checking online, OFC 73 00:03:18,719 --> 00:03:21,285 celebrated its 50-year anniversary in 2025. 74 00:03:21,285 --> 00:03:22,200 That's right. 75 00:03:22,200 --> 00:03:23,745 I've been to about a third of them. 76 00:03:23,745 --> 00:03:28,994 Wow. A few more for me, I've lost count. 77 00:03:28,994 --> 00:03:31,290 I do remember my first one 78 00:03:31,290 --> 00:03:33,179 was about a decade before Brodie. 79 00:03:33,179 --> 00:03:36,135 I think it was 1990 or 1991. 80 00:03:36,135 --> 00:03:38,174 Let's call it 35 years. 81 00:03:38,174 --> 00:03:40,140 I certainly didn't make every one of those, 82 00:03:40,140 --> 00:03:41,939 but I probably made the vast majority. 83 00:03:41,939 --> 00:03:46,409 It was quite a different machine in those days. 84 00:03:46,409 --> 00:03:48,914 A lot more science and more tech. 85 00:03:48,914 --> 00:03:51,900 35ish years ago, that's for sure. 86 00:03:51,900 --> 00:03:54,509 That's amazing. Both of you have attended 87 00:03:54,509 --> 00:03:57,869 many OFCs and during your tenures, 88 00:03:57,869 --> 00:04:03,269 have seen our industry go through cycles or major shifts. 89 00:04:03,269 --> 00:04:05,100 Looking back over your time at OFC, 90 00:04:05,100 --> 00:04:06,675 what are a couple of the 91 00:04:06,675 --> 00:04:09,959 key technology shifts you remember? 92 00:04:09,959 --> 00:04:13,199 There's a couple of things that come to mind, Rebecca. 93 00:04:13,199 --> 00:04:15,090 Going to the first one was 94 00:04:15,090 --> 00:04:17,579 super exciting and also overwhelming. 95 00:04:17,579 --> 00:04:18,929 I didn't know anybody. 96 00:04:18,929 --> 00:04:22,140 I was hurrying between all the different places, 97 00:04:22,140 --> 00:04:23,820 trying to get calibrated and attend 98 00:04:23,820 --> 00:04:26,294 the different talks and things like that. 99 00:04:26,294 --> 00:04:28,004 As the years went by, 100 00:04:28,004 --> 00:04:29,580 I came to know a lot of people. 101 00:04:29,580 --> 00:04:30,869 In fact, at that this last one, 102 00:04:30,869 --> 00:04:33,090 I'd say probably 1 in 3 or 1 in 103 00:04:33,090 --> 00:04:35,404 4 people I actually know personally, 104 00:04:35,404 --> 00:04:37,550 but I still feel that same sense of 105 00:04:37,550 --> 00:04:40,955 excitement around the tech and the innovation. 106 00:04:40,955 --> 00:04:43,490 I still have so much to learn as we go through it. 107 00:04:43,490 --> 00:04:46,279 There are a couple of inflections that I've been part of, 108 00:04:46,279 --> 00:04:48,729 and then we can see what Brodie is thinking. 109 00:04:48,729 --> 00:04:51,749 Certainly, the boom of the late 90s, 110 00:04:51,749 --> 00:04:56,205 early 2000s – I was there before there was a DWDM. 111 00:04:56,205 --> 00:04:58,379 We were just sending one wavelength 112 00:04:58,379 --> 00:05:00,000 and all we could do was wiggle out 113 00:05:00,000 --> 00:05:01,289 a few hundred megabits per 114 00:05:01,289 --> 00:05:04,499 second or even less on a fiber. 115 00:05:04,499 --> 00:05:07,830 Then came the DWDM era, 2.5G, 116 00:05:07,830 --> 00:05:11,085 10G, Erbium-Doped Fiber Amplifiers. 117 00:05:11,085 --> 00:05:12,960 That was what caused the explosion in 118 00:05:12,960 --> 00:05:15,989 the latter part of the 90s and the 2000s. 119 00:05:15,989 --> 00:05:18,915 The other one that's near and dear to my heart 120 00:05:18,915 --> 00:05:22,604 was the whole coherent optics revolution. 121 00:05:22,604 --> 00:05:26,070 We were among the very first to be doing that kind of 122 00:05:26,070 --> 00:05:30,089 lead technology work and on mixing radio optics, 123 00:05:30,089 --> 00:05:31,980 as well as advanced CMOS 124 00:05:31,980 --> 00:05:33,839 and electronic components together. 125 00:05:33,839 --> 00:05:35,399 We stirred that pot. 126 00:05:35,399 --> 00:05:38,460 From there coherent optics was born. 127 00:05:38,460 --> 00:05:42,389 We unveiled in 2005/2006, 128 00:05:42,389 --> 00:05:46,184 the first DSP-powered optical system 129 00:05:46,184 --> 00:05:49,210 with 10G on-off keying. 130 00:05:50,270 --> 00:05:54,959 20 years later, we now have 1.6TB on a wavelength, 131 00:05:54,959 --> 00:05:59,234 30-40TB, 50TB per fiber, you name it. 132 00:05:59,234 --> 00:06:02,789 I've seen that evolution over those years. 133 00:06:02,789 --> 00:06:05,520 Those innovation cycles are definitely getting shorter 134 00:06:05,520 --> 00:06:08,325 and shorter in between these major breakthroughs. 135 00:06:08,325 --> 00:06:11,999 Brodie – How about you? In your time attending OFC, 136 00:06:11,999 --> 00:06:14,459 what are the big industry milestones you've seen? 137 00:06:14,459 --> 00:06:16,409 I was just reflecting back on my career. 138 00:06:16,409 --> 00:06:18,434 I started off as an engineer working on 139 00:06:18,434 --> 00:06:21,360 Nortel’s OC 48 classic, 140 00:06:21,360 --> 00:06:23,025 just to give you an idea of 141 00:06:23,025 --> 00:06:26,429 the terminology that we were using at the time. 142 00:06:26,429 --> 00:06:27,749 I was just looking at 143 00:06:27,749 --> 00:06:29,100 the transitions that I've been through. 144 00:06:29,100 --> 00:06:31,575 It's been 2.5G to 10G, 145 00:06:31,575 --> 00:06:34,439 10G to DWDM, DWDM to IMDD, 146 00:06:34,439 --> 00:06:36,750 then IMDD going to coherent. 147 00:06:36,750 --> 00:06:41,039 Coherent was challenged from a cost perspective at first, 148 00:06:41,039 --> 00:06:43,260 but it was clear that it was eventual. 149 00:06:43,260 --> 00:06:45,719 It started off in long-haul DWDM 150 00:06:45,719 --> 00:06:47,175 and then it went to submarine. 151 00:06:47,175 --> 00:06:49,424 Over a multi-year period, 152 00:06:49,424 --> 00:06:52,139 we cost reduced and power reduced coherent modems 153 00:06:52,139 --> 00:06:54,060 by 25% on average per year. 154 00:06:54,060 --> 00:06:55,949 What that then allowed us to do is get 155 00:06:55,949 --> 00:06:58,649 coherent from the long-haul into Metro. 156 00:06:58,649 --> 00:07:03,119 Now Metro into pluggable for Metro DCI type use cases. 157 00:07:03,119 --> 00:07:05,489 Then coherent is going inside the data center, 158 00:07:05,489 --> 00:07:08,040 and optical is becoming even more important 159 00:07:08,040 --> 00:07:09,329 inside the data center 160 00:07:09,329 --> 00:07:11,460 with things like co-packaged optics. 161 00:07:11,460 --> 00:07:13,424 That's a great way to think about it, Brodie. 162 00:07:13,424 --> 00:07:16,140 The capacity and speed in those increments 163 00:07:16,140 --> 00:07:18,630 and how we've been able to innovate 164 00:07:18,630 --> 00:07:21,555 so much to bring coherent optics from 165 00:07:21,555 --> 00:07:23,355 one application to across 166 00:07:23,355 --> 00:07:26,025 every part of the network and inside the data center. 167 00:07:26,025 --> 00:07:28,019 If you think about today, we're in 168 00:07:28,019 --> 00:07:30,780 this AI supercycle, as some call it. 169 00:07:30,780 --> 00:07:32,520 In that context, Dino, 170 00:07:32,520 --> 00:07:37,005 what stood out in terms of innovation at OFC this year? 171 00:07:37,005 --> 00:07:38,774 That's a good question. 172 00:07:38,774 --> 00:07:40,940 Despite having seen so many OFCs, 173 00:07:40,940 --> 00:07:43,379 and the frenzy and the pace of 174 00:07:43,379 --> 00:07:46,574 innovation across multiple technologies, 175 00:07:46,574 --> 00:07:48,779 what Brodie and I both talked about 176 00:07:48,779 --> 00:07:52,709 was primarily the transceiver evolution. 177 00:07:52,709 --> 00:07:55,544 We've certainly been hellbent on a journey 178 00:07:55,544 --> 00:07:58,470 to drive capacity up while driving cost, 179 00:07:58,470 --> 00:08:00,585 power, and space down. 180 00:08:00,585 --> 00:08:04,179 The report card is stellar, as Brodie outlined. 181 00:08:05,480 --> 00:08:08,910 What we're seeing now on that front is a push to 182 00:08:08,910 --> 00:08:11,669 get deeper inside the data center, 183 00:08:11,669 --> 00:08:14,100 and closer to the GPUs, 184 00:08:14,100 --> 00:08:16,575 CPUs, and the memory. 185 00:08:16,575 --> 00:08:19,830 There's a frenzy of activity and innovation and using 186 00:08:19,830 --> 00:08:23,100 new materials and systems to try to get that space done. 187 00:08:23,100 --> 00:08:25,004 Number two, I see 188 00:08:25,004 --> 00:08:27,884 long overdue innovation again in amplification. 189 00:08:27,884 --> 00:08:29,969 I mentioned that in my early days, 190 00:08:29,969 --> 00:08:35,205 EDFAs became a workhorse that unleashed DWDM. 191 00:08:35,205 --> 00:08:37,140 It's been a workhorse for decades 192 00:08:37,140 --> 00:08:38,684 and hasn't changed much. 193 00:08:38,684 --> 00:08:40,244 Now, with the scale, 194 00:08:40,244 --> 00:08:42,915 especially with applications like scale across, 195 00:08:42,915 --> 00:08:44,534 amplifiers have to change. 196 00:08:44,534 --> 00:08:47,099 We need to see new technologies in that area. 197 00:08:47,099 --> 00:08:49,889 There's a fervor going on in terms of trying to 198 00:08:49,889 --> 00:08:52,979 harness new techniques and new materials in that space. 199 00:08:52,979 --> 00:08:55,259 The third one, which is not 200 00:08:55,259 --> 00:08:57,720 our direct wheelhouse, is fiber. 201 00:08:57,720 --> 00:08:59,835 In the last couple of years, 202 00:08:59,835 --> 00:09:02,040 we've seen the emergence of 203 00:09:02,040 --> 00:09:05,550 a modern generation of hollow core fiber that 204 00:09:05,550 --> 00:09:08,490 has tackled the loss challenges 205 00:09:08,490 --> 00:09:11,715 of previous incarnations of hollow core fiber. 206 00:09:11,715 --> 00:09:13,379 I see a lot more activity, 207 00:09:13,379 --> 00:09:15,960 a lot more technologies emerging, 208 00:09:15,960 --> 00:09:19,050 but also a lot more interest from our big customers, 209 00:09:19,050 --> 00:09:21,344 asking “How does this work? 210 00:09:21,344 --> 00:09:22,980 And if it works, how will it 211 00:09:22,980 --> 00:09:25,680 influence our technology and our product evolutions into 212 00:09:25,680 --> 00:09:28,400 the future?” This hyperscale is 213 00:09:28,400 --> 00:09:31,884 really driving innovation on multiple fronts here. 214 00:09:31,884 --> 00:09:34,050 Everybody's talking about AI 215 00:09:34,050 --> 00:09:35,549 and everyone's building up these networks. 216 00:09:35,549 --> 00:09:41,639 But OFC started off as a very scientific, 217 00:09:41,639 --> 00:09:46,295 technology-rooted event focused on proof of concepts. 218 00:09:46,295 --> 00:09:47,689 In talking to customers, is 219 00:09:47,689 --> 00:09:49,070 there that sense of urgency that we’ve 220 00:09:49,070 --> 00:09:52,324 got to get this this innovation into our networks today? 221 00:09:52,324 --> 00:09:56,329 It's a great question. I’ll break 222 00:09:56,329 --> 00:09:58,360 it down into multiple segments. 223 00:09:58,360 --> 00:10:00,540 If you just look at the service providers, 224 00:10:00,540 --> 00:10:02,834 the traditional tier one service providers, 225 00:10:02,834 --> 00:10:04,349 what they're really focused on 226 00:10:04,349 --> 00:10:06,210 is shifting their investment away 227 00:10:06,210 --> 00:10:09,885 from their 5G buildouts back to optical. 228 00:10:09,885 --> 00:10:11,249 They realized that they 229 00:10:11,249 --> 00:10:13,950 underinvested in their optical infrastructure. 230 00:10:13,950 --> 00:10:15,299 They're trying to get that, along with 231 00:10:15,299 --> 00:10:17,429 their broadband access deployments, going. 232 00:10:17,429 --> 00:10:18,750 They're shifting their focus and 233 00:10:18,750 --> 00:10:19,979 their investment back into 234 00:10:19,979 --> 00:10:22,814 IP and optical deployments. That's number one. 235 00:10:22,814 --> 00:10:25,125 The urgency is they've underinvested, 236 00:10:25,125 --> 00:10:25,859 and they need to get 237 00:10:25,859 --> 00:10:27,300 more capacity in the ground to support 238 00:10:27,300 --> 00:10:28,499 the services that they're trying to put 239 00:10:28,499 --> 00:10:30,164 on top of that infrastructure. 240 00:10:30,164 --> 00:10:32,070 Then I look at the hyperscalers. 241 00:10:32,070 --> 00:10:33,809 For the hyperscalers, 242 00:10:33,809 --> 00:10:35,849 it's really about their AI build out. 243 00:10:35,849 --> 00:10:37,035 I think right now it's about 244 00:10:37,035 --> 00:10:38,939 AI data ingestion and 245 00:10:38,939 --> 00:10:40,859 scale across for training use cases. 246 00:10:40,859 --> 00:10:43,544 For the scale across use case, 247 00:10:43,544 --> 00:10:46,289 where they're deploying data centers across 248 00:10:46,289 --> 00:10:50,294 multiple locations and spread out across the country, 249 00:10:50,294 --> 00:10:52,214 they need to be able to interconnect it. 250 00:10:52,214 --> 00:10:54,659 The sense of urgency is that there's so much 251 00:10:54,659 --> 00:10:56,340 bandwidth that's required to 252 00:10:56,340 --> 00:10:57,734 interconnect these data centers. 253 00:10:57,734 --> 00:11:00,014 They need to solve two fundamental problems. 254 00:11:00,014 --> 00:11:03,449 One, getting fiber infrastructure in place. 255 00:11:03,449 --> 00:11:05,205 They've been doing that for a couple of years now. 256 00:11:05,205 --> 00:11:07,949 And two, solving their amplifier hut challenges. 257 00:11:07,949 --> 00:11:09,749 That's where the densification 258 00:11:09,749 --> 00:11:11,760 of amplifiers and the investment that we, 259 00:11:11,760 --> 00:11:14,310 and others, are making in hyper-rail technology 260 00:11:14,310 --> 00:11:15,944 to enable that scale across. 261 00:11:15,944 --> 00:11:18,134 It's now becoming foundational 262 00:11:18,134 --> 00:11:20,459 to their AI training and their infrastructure. 263 00:11:20,459 --> 00:11:22,140 The third group that I'll 264 00:11:22,140 --> 00:11:24,074 talk about is the wholesale carriers. 265 00:11:24,074 --> 00:11:27,284 The wholesale carriers are trying to 266 00:11:27,284 --> 00:11:28,890 win as much neoscaler and 267 00:11:28,890 --> 00:11:30,869 hyperscaler business as humanly possible. 268 00:11:30,869 --> 00:11:33,120 They recognize that the best way to do that is 269 00:11:33,120 --> 00:11:35,414 to have the lowest cost and lowest power per bit. 270 00:11:35,414 --> 00:11:37,815 The cycle of technology, 271 00:11:37,815 --> 00:11:40,859 both from an investment perspective on our side and how 272 00:11:40,859 --> 00:11:43,769 the customers consume it, is increasing drastically. 273 00:11:43,769 --> 00:11:46,334 That time cycle is shrinking. 274 00:11:46,334 --> 00:11:50,460 That’s great. You were mentioning inside the data center. 275 00:11:50,460 --> 00:11:53,159 During OFC, there were some debates 276 00:11:53,159 --> 00:11:54,899 happening in the industry 277 00:11:54,899 --> 00:11:58,410 around copper versus coherent 278 00:11:58,410 --> 00:11:59,910 optical inside the data center. 279 00:11:59,910 --> 00:12:02,849 Dino - I wonder if you could comment on that. 280 00:12:02,849 --> 00:12:06,900 How should our customers be thinking about the role of 281 00:12:06,900 --> 00:12:10,214 coherent optics versus other technologies 282 00:12:10,214 --> 00:12:12,135 inside the data center? 283 00:12:12,135 --> 00:12:16,815 Could you comment on what bets we have made in our R&D, 284 00:12:16,815 --> 00:12:18,209 thinking about some of 285 00:12:18,209 --> 00:12:20,730 our more recently announced products like 286 00:12:20,730 --> 00:12:22,949 Vesta and Nitro and 287 00:12:22,949 --> 00:12:25,155 how they fit into this bigger picture? 288 00:12:25,155 --> 00:12:26,999 That's a good point, Rebecca. 289 00:12:26,999 --> 00:12:29,250 In the frenzy, there's so many announcements 290 00:12:29,250 --> 00:12:31,920 and so many demos and booths. 291 00:12:31,920 --> 00:12:34,185 It's easy to get lost in the myriad. 292 00:12:34,185 --> 00:12:37,935 I’d step back and think about it as a continuum. 293 00:12:37,935 --> 00:12:40,860 It really is a continuum as to what 294 00:12:40,860 --> 00:12:43,860 technologies fit for which part of the application. 295 00:12:43,860 --> 00:12:46,109 When you think of within a shelf or up 296 00:12:46,109 --> 00:12:48,899 a rack all the way to across the Pacific Ocean, 297 00:12:48,899 --> 00:12:51,210 those are the application spaces. 298 00:12:51,210 --> 00:12:53,490 Now, as the capacity goes up, 299 00:12:53,490 --> 00:12:56,100 you still have to look at it as a continuum, 300 00:12:56,100 --> 00:12:59,519 but there is a bit of a shifting of technologies. 301 00:12:59,519 --> 00:13:02,940 Brodie alluded to that as he talked about early coherent 302 00:13:02,940 --> 00:13:05,129 starting in the long-haul subsea 303 00:13:05,129 --> 00:13:06,585 and then finding its way down. 304 00:13:06,585 --> 00:13:09,299 It's all part of that continuum in the journey. 305 00:13:09,299 --> 00:13:11,819 If I start deep in the data center, 306 00:13:11,819 --> 00:13:16,845 there's copper and basic IMDD signaling over copper. 307 00:13:16,845 --> 00:13:19,529 You stay there as long as you can because 308 00:13:19,529 --> 00:13:21,899 it's lower power and lower cost – and this 309 00:13:21,899 --> 00:13:24,254 is coming from the optics guy – 310 00:13:24,254 --> 00:13:26,070 because that's the right solution 311 00:13:26,070 --> 00:13:28,539 for transmission in there. 312 00:13:28,700 --> 00:13:31,304 That's where Nitro fits in. 313 00:13:31,304 --> 00:13:33,719 One of our technologies that is 314 00:13:33,719 --> 00:13:36,434 emerging from the Nubis acquisition is this 315 00:13:36,434 --> 00:13:39,269 re-driver that helps extend the reach of 316 00:13:39,269 --> 00:13:41,505 a copper interconnect or a cable interconnect 317 00:13:41,505 --> 00:13:44,025 while keeping the power as low as possible. 318 00:13:44,025 --> 00:13:45,464 That's the first step. 319 00:13:45,464 --> 00:13:47,579 At some point, though, the physics gets you 320 00:13:47,579 --> 00:13:49,770 and then you need to move to optical. 321 00:13:49,770 --> 00:13:54,119 Then you use optical but with simpler coating like IMDD. 322 00:13:54,119 --> 00:13:56,084 That's where you see things like 323 00:13:56,084 --> 00:13:58,229 co-packaged and near-packaged optics, 324 00:13:58,229 --> 00:14:01,290 even some of the basic face plate optics. 325 00:14:01,290 --> 00:14:04,799 That's where our Vesta product line, also from Nubis, 326 00:14:04,799 --> 00:14:07,320 comes in: Co-packaged optics offering 327 00:14:07,320 --> 00:14:09,510 interconnect componentry that allows 328 00:14:09,510 --> 00:14:11,669 6.4 terabits per second to be 329 00:14:11,669 --> 00:14:14,894 connected from close to a GPU or a CPU, 330 00:14:14,894 --> 00:14:17,340 up and down racks and across the data centers 331 00:14:17,340 --> 00:14:20,009 themselves to a couple of kilometers. 332 00:14:20,009 --> 00:14:22,199 At some point that breaks, 333 00:14:22,199 --> 00:14:26,535 and then enters coherent-assisted or Coherent-Lite. 334 00:14:26,535 --> 00:14:29,160 That's where investments and technologies we've 335 00:14:29,160 --> 00:14:31,919 been developing for the last couple of seasons will 336 00:14:31,919 --> 00:14:35,070 start to play a role as you get to longer reaches in 337 00:14:35,070 --> 00:14:38,550 the campus as well as outside some of the campuses, 338 00:14:38,550 --> 00:14:40,245 but in close proximity. 339 00:14:40,245 --> 00:14:42,509 Ultimately, you take the next step 340 00:14:42,509 --> 00:14:44,009 to go over distance in the Metro, 341 00:14:44,009 --> 00:14:46,305 regional, long-haul and subsea. 342 00:14:46,305 --> 00:14:48,869 Then you put on full-bore coherent 343 00:14:48,869 --> 00:14:51,839 with varying modulation schemes and 344 00:14:51,839 --> 00:14:53,940 signal processing enhancements that 345 00:14:53,940 --> 00:14:55,259 allow you to always get 346 00:14:55,259 --> 00:14:56,609 the highest capacity for 347 00:14:56,609 --> 00:15:00,179 the lowest cost power and space that you can. 348 00:15:00,179 --> 00:15:02,115 One of the things that surprised me at 349 00:15:02,115 --> 00:15:03,929 OFC was the debate of whether it 350 00:15:03,929 --> 00:15:07,350 was going to be copper or optical or IMDD or coherent. 351 00:15:07,350 --> 00:15:08,850 It's an “and” – the question is at 352 00:15:08,850 --> 00:15:10,830 what rate and what distance? 353 00:15:10,830 --> 00:15:13,125 As you increase distance or increase rate, 354 00:15:13,125 --> 00:15:14,639 copper will go to optical and 355 00:15:14,639 --> 00:15:17,789 IMDD will go to coherent. But it's an “and.” 356 00:15:17,789 --> 00:15:20,249 Really well said, Brodie. 357 00:15:20,249 --> 00:15:22,919 What's interesting about Ciena’s strategy 358 00:15:22,919 --> 00:15:24,689 is we are setting ourselves 359 00:15:24,689 --> 00:15:26,429 up to serve the spectrum of 360 00:15:26,429 --> 00:15:29,714 the different technology needs, as you mentioned. 361 00:15:29,714 --> 00:15:31,769 The Nubis acquisition, 362 00:15:31,769 --> 00:15:33,794 which was just this past fall for us, 363 00:15:33,794 --> 00:15:36,599 has set us up well to be able to serve 364 00:15:36,599 --> 00:15:40,245 those different needs inside the data center. 365 00:15:40,245 --> 00:15:44,684 The Ciena family has grown and expanded with Nubis. 366 00:15:44,684 --> 00:15:47,400 Dino – How has it been working with a great set 367 00:15:47,400 --> 00:15:50,309 of engineers that came to us? How is that going? 368 00:15:50,309 --> 00:15:52,020 I'm a little biased, but it's 369 00:15:52,020 --> 00:15:54,819 going fabulously well, Rebecca. 370 00:15:55,160 --> 00:15:57,195 They're a great team. 371 00:15:57,195 --> 00:15:58,695 I'll speak from the heart. 372 00:15:58,695 --> 00:16:01,170 I've known some of those guys for many years. 373 00:16:01,170 --> 00:16:03,299 We were collegiate in some ways. 374 00:16:03,299 --> 00:16:06,074 We fiercely competed on the playing field. 375 00:16:06,074 --> 00:16:08,235 Many of them are from Bell Labs, 376 00:16:08,235 --> 00:16:10,259 while we were north of the border 377 00:16:10,259 --> 00:16:11,489 in Bell-Northern Research. 378 00:16:11,489 --> 00:16:13,964 I have immense respect for them technically, 379 00:16:13,964 --> 00:16:15,840 and deeper than that, 380 00:16:15,840 --> 00:16:17,729 is the character and that sense of 381 00:16:17,729 --> 00:16:20,565 pride and commitment to the cause. 382 00:16:20,565 --> 00:16:22,349 I couldn't be happier. 383 00:16:22,349 --> 00:16:24,000 They're a fantastic team. 384 00:16:24,000 --> 00:16:25,620 I know both you and Brodie saw them 385 00:16:25,620 --> 00:16:27,314 beaming with pride as they were 386 00:16:27,314 --> 00:16:29,535 demonstrating that first Vesta 387 00:16:29,535 --> 00:16:31,965 in our Ciena booth just a few weeks ago. 388 00:16:31,965 --> 00:16:33,749 I couldn't be happier for them. 389 00:16:33,749 --> 00:16:36,360 We want the tech and we want to continue to lead, 390 00:16:36,360 --> 00:16:39,119 but to see the pride and the efforts that 391 00:16:39,119 --> 00:16:41,909 they put in to be part of it, I'm absolutely thrilled. 392 00:16:41,909 --> 00:16:44,475 We love having these guys and gals with us. 393 00:16:44,475 --> 00:16:46,769 It's already starting to stir the pot for 394 00:16:46,769 --> 00:16:50,880 new innovations beyond what's already on the track. 395 00:16:50,880 --> 00:16:53,609 It's great to have them as part of Ciena, 396 00:16:53,609 --> 00:16:55,200 and it's going to be amazing to see 397 00:16:55,200 --> 00:16:57,435 what we do with that technology. 398 00:16:57,435 --> 00:16:59,745 Let's move outside of the data center 399 00:16:59,745 --> 00:17:01,290 and talk about scale across. 400 00:17:01,290 --> 00:17:02,820 Brodie – I want to go back to that comment 401 00:17:02,820 --> 00:17:04,350 you made earlier about how 402 00:17:04,350 --> 00:17:05,639 you see the different segments of 403 00:17:05,639 --> 00:17:07,875 the market and the different technologies. 404 00:17:07,875 --> 00:17:10,619 What are the architectural changes 405 00:17:10,619 --> 00:17:11,790 we should be thinking about, 406 00:17:11,790 --> 00:17:13,800 what our customers should be thinking about, 407 00:17:13,800 --> 00:17:15,989 and how are we approaching the whole scale 408 00:17:15,989 --> 00:17:19,004 across opportunity from a roadmap perspective? 409 00:17:19,004 --> 00:17:20,700 That’s a great question. I'll start with 410 00:17:20,700 --> 00:17:23,264 describing how I define scale across. 411 00:17:23,264 --> 00:17:25,019 It's different in my mind from 412 00:17:25,019 --> 00:17:27,255 Metro DCI, Data Center Interconnect. 413 00:17:27,255 --> 00:17:29,069 When I look at scale across, 414 00:17:29,069 --> 00:17:30,629 the difference between DCI is that it's 415 00:17:30,629 --> 00:17:32,415 connecting the back end 416 00:17:32,415 --> 00:17:36,434 of our hyperscalers’ fabrics, GPU to GPU. 417 00:17:36,434 --> 00:17:39,389 They cannot get enough GPUs because of 418 00:17:39,389 --> 00:17:42,269 power and real estate in one data center, 419 00:17:42,269 --> 00:17:45,570 so they have to spread them across multiple data centers. 420 00:17:45,570 --> 00:17:47,474 When they spread them across multiple data centers, 421 00:17:47,474 --> 00:17:48,675 they have to interconnect them. 422 00:17:48,675 --> 00:17:51,149 The difference between DCI and scale 423 00:17:51,149 --> 00:17:53,459 across at a foundational level 424 00:17:53,459 --> 00:17:55,739 is the amount of bandwidth that is required. 425 00:17:55,739 --> 00:17:57,420 We're now moving the connectivity 426 00:17:57,420 --> 00:17:58,620 that traditionally sat within 427 00:17:58,620 --> 00:18:02,250 the data center and extending it across distance. 428 00:18:02,250 --> 00:18:04,890 There are, for many of the use cases, 429 00:18:04,890 --> 00:18:06,629 tens of petabytes of 430 00:18:06,629 --> 00:18:09,044 traffic that need to go from one data center to another. 431 00:18:09,044 --> 00:18:12,885 It's high-capacity point-to-point connectivity. 432 00:18:12,885 --> 00:18:14,399 Initial deployments were on 433 00:18:14,399 --> 00:18:16,680 the 80-kilometer type distance range. 434 00:18:16,680 --> 00:18:18,179 You could argue that was simple, 435 00:18:18,179 --> 00:18:20,504 but it's hundreds of fiber pairs, 436 00:18:20,504 --> 00:18:22,634 and you’ve got to multiplex all those signals 437 00:18:22,634 --> 00:18:25,290 onto a common infrastructure. 438 00:18:25,290 --> 00:18:28,829 Real deployments there were either embedded modems, 439 00:18:28,829 --> 00:18:31,439 high performance to optimize spectra efficiency 440 00:18:31,439 --> 00:18:33,450 or pluggable coherent optics 441 00:18:33,450 --> 00:18:35,144 across a photonic line system. 442 00:18:35,144 --> 00:18:39,194 The scale of those line systems is unreal. 443 00:18:39,194 --> 00:18:41,040 Being able to handle the software, 444 00:18:41,040 --> 00:18:43,049 being able to handle protection and switching, 445 00:18:43,049 --> 00:18:45,329 being able to handle events that are happening 446 00:18:45,329 --> 00:18:47,864 in the network are absolutely critical. 447 00:18:47,864 --> 00:18:50,219 As soon as our customers want to 448 00:18:50,219 --> 00:18:52,920 start extending that across further distance, 449 00:18:52,920 --> 00:18:54,390 the big problems become 450 00:18:54,390 --> 00:18:57,840 fiber and solving that amplifier hut problem. 451 00:18:57,840 --> 00:19:01,300 That's where hyper-rail comes into play. 452 00:19:01,300 --> 00:19:04,820 The densification of amplifier technology is such that 453 00:19:04,820 --> 00:19:08,525 now they can amplify hundreds of fiber pairs over 1000km. 454 00:19:08,525 --> 00:19:10,010 What I would think about from 455 00:19:10,010 --> 00:19:13,744 an architecture perspective is very high-capacity 456 00:19:13,744 --> 00:19:15,670 point-to-point connectivity 457 00:19:15,670 --> 00:19:17,984 with both photonics-embedded modems 458 00:19:17,984 --> 00:19:20,505 and coherent pluggables. That's the winning end. 459 00:19:20,505 --> 00:19:22,260 We've talked about a lot of technologies. 460 00:19:22,260 --> 00:19:24,510 We talked about technologies required in 461 00:19:24,510 --> 00:19:27,045 innovations in and around the data center. 462 00:19:27,045 --> 00:19:29,550 There's huge opportunity around scale across. 463 00:19:29,550 --> 00:19:33,495 Beyond these innovations that we've talked about, 464 00:19:33,495 --> 00:19:36,690 what other key architectural changes that drive 465 00:19:36,690 --> 00:19:38,910 this AI-level scale and 466 00:19:38,910 --> 00:19:40,590 operational efficiency should we 467 00:19:40,590 --> 00:19:42,464 be thinking about moving forward? 468 00:19:42,464 --> 00:19:45,089 It's a good point. There's the usual blocking 469 00:19:45,089 --> 00:19:47,369 and tackling we've been talking about. Three things. 470 00:19:47,369 --> 00:19:48,900 One is picking up on 471 00:19:48,900 --> 00:19:52,949 Brodie's theme on the amplifiers and the multi-rail. 472 00:19:52,949 --> 00:19:56,280 It's ripe for innovation inside how 473 00:19:56,280 --> 00:19:59,130 we do optical amplification. 474 00:19:59,130 --> 00:20:02,714 You mentioned that the huts are a problem. 475 00:20:02,714 --> 00:20:05,969 We never would have thought that many years ago. 476 00:20:05,969 --> 00:20:10,875 There's more revolution required inside amplification. 477 00:20:10,875 --> 00:20:14,250 Two, I touched on it before: fiber. 478 00:20:14,250 --> 00:20:15,269 It’s sacred. 479 00:20:15,269 --> 00:20:16,619 Historically, people put it in 480 00:20:16,619 --> 00:20:18,135 the ground and never wanted to touch it. 481 00:20:18,135 --> 00:20:21,839 With this level of expansion and scale, 482 00:20:21,839 --> 00:20:24,359 they're happy to fill conduits and pull 483 00:20:24,359 --> 00:20:26,009 old fiber and put the new stuff 484 00:20:26,009 --> 00:20:28,110 in if it gives them an advantage. 485 00:20:28,110 --> 00:20:29,865 That's why I mentioned earlier, 486 00:20:29,865 --> 00:20:33,509 innovation around hollow core fiber is really attractive. 487 00:20:33,509 --> 00:20:36,330 The original allure was latency. 488 00:20:36,330 --> 00:20:40,574 Light travels faster in an air than it does in glass, 489 00:20:40,574 --> 00:20:41,999 but the kinds of losses 490 00:20:41,999 --> 00:20:43,409 we're seeing now and that are being 491 00:20:43,409 --> 00:20:47,009 reported are making it very attractive. 492 00:20:47,009 --> 00:20:50,565 That drives innovation back on the transceivers, 493 00:20:50,565 --> 00:20:52,215 because if the medium's different, 494 00:20:52,215 --> 00:20:54,780 what else can you do with the transceivers? 495 00:20:54,780 --> 00:20:56,339 What else can you do with the amplifier? 496 00:20:56,339 --> 00:20:58,034 That's also very exciting. 497 00:20:58,034 --> 00:21:01,095 The last thing I'll say is cooling. 498 00:21:01,095 --> 00:21:02,729 The biggest challenge is not 499 00:21:02,729 --> 00:21:05,219 all these photons and high-speed electrons. 500 00:21:05,219 --> 00:21:06,690 It’s getting the heat out of 501 00:21:06,690 --> 00:21:09,270 these products that we're developing. 502 00:21:09,270 --> 00:21:12,719 We're using absolutely state-of-the-art technology, 503 00:21:12,719 --> 00:21:15,975 such as 2-nanometer CMOS-based devices. 504 00:21:15,975 --> 00:21:17,609 There's still a need to get 505 00:21:17,609 --> 00:21:19,874 heat out of these very small spots. 506 00:21:19,874 --> 00:21:21,495 I've lived long enough to make 507 00:21:21,495 --> 00:21:22,949 sure we're finally going to get 508 00:21:22,949 --> 00:21:24,840 liquid cooling into 509 00:21:24,840 --> 00:21:26,864 both transmission and switching products. 510 00:21:26,864 --> 00:21:28,739 Long overdue, but it will open 511 00:21:28,739 --> 00:21:30,779 a new frontier of how to better optimize 512 00:21:30,779 --> 00:21:32,399 the designs to handle 513 00:21:32,399 --> 00:21:35,100 this insatiable demand for capacity. 514 00:21:35,100 --> 00:21:37,110 I remember when your team 515 00:21:37,110 --> 00:21:40,409 demonstrated that at our Vectors event two years ago, 516 00:21:40,409 --> 00:21:44,324 and it was a last minute add to the demo lineup, 517 00:21:44,324 --> 00:21:45,465 and it was probably 518 00:21:45,465 --> 00:21:49,090 the most sought-after demo that we had. 519 00:21:50,510 --> 00:21:53,444 The demand is there in the industry. 520 00:21:53,444 --> 00:21:55,424 We'll shift gears a little bit. 521 00:21:55,424 --> 00:21:58,784 I want to talk about customer segments with both of you. 522 00:21:58,784 --> 00:22:02,595 A lot of the technologies we've talked about are around 523 00:22:02,595 --> 00:22:04,290 the AI opportunity: the 524 00:22:04,290 --> 00:22:06,585 build out of these massive data centers, 525 00:22:06,585 --> 00:22:08,939 and the distribution of these data centers 526 00:22:08,939 --> 00:22:11,790 necessitating higher speeds and longer distances, 527 00:22:11,790 --> 00:22:13,590 as well as higher, 528 00:22:13,590 --> 00:22:17,430 more power efficient speeds inside the data center. 529 00:22:17,430 --> 00:22:19,004 That's the hyperscalers. 530 00:22:19,004 --> 00:22:21,659 There's also these emerging neoscalers 531 00:22:21,659 --> 00:22:23,279 that both of you have touched on. 532 00:22:23,279 --> 00:22:26,624 Then there are the communication service providers. 533 00:22:26,624 --> 00:22:28,275 Brodie – from your perspective, 534 00:22:28,275 --> 00:22:32,370 how do you see the needs across these different segments? 535 00:22:32,370 --> 00:22:33,614 How do they differ? 536 00:22:33,614 --> 00:22:35,984 Where are the growth opportunities for each of them? 537 00:22:35,984 --> 00:22:37,529 Hyperscalers seem to dominate, 538 00:22:37,529 --> 00:22:39,330 but service providers still have 539 00:22:39,330 --> 00:22:43,649 a really viable opportunity in this AI value chain. 540 00:22:43,649 --> 00:22:45,885 It's a great question. I think I mentioned it earlier. 541 00:22:45,885 --> 00:22:48,060 That's how I break down the market segments. 542 00:22:48,060 --> 00:22:49,919 I'd say the service providers are in two segments – 543 00:22:49,919 --> 00:22:51,659 the tier one operators and 544 00:22:51,659 --> 00:22:53,129 then the wholesale carriers – and 545 00:22:53,129 --> 00:22:54,735 then neo scalers and hyperscalers. 546 00:22:54,735 --> 00:22:57,779 That's the way I break down the customer segmentation. 547 00:22:57,779 --> 00:23:00,374 About 42% of Ciena’s business 548 00:23:00,374 --> 00:23:02,459 is direct to hyperscalers and neoscalers. 549 00:23:02,459 --> 00:23:04,709 The service providers continue to be 550 00:23:04,709 --> 00:23:07,830 a very critical segment for Ciena. 551 00:23:07,830 --> 00:23:09,885 If I look at the hyperscalers, 552 00:23:09,885 --> 00:23:12,045 the bandwidth is being driven by 553 00:23:12,045 --> 00:23:15,314 AI ingestion and scale across that we've spoken about. 554 00:23:15,314 --> 00:23:17,204 The neoscalers, same thing. 555 00:23:17,204 --> 00:23:20,895 On the service provider side, there's two things. 556 00:23:20,895 --> 00:23:23,385 One is their infrastructure buildout 557 00:23:23,385 --> 00:23:26,609 and trying to capture what we call MOFN, 558 00:23:26,609 --> 00:23:28,274 managed optical fiber networks, 559 00:23:28,274 --> 00:23:30,569 and capture the build outs on behalf of 560 00:23:30,569 --> 00:23:34,079 the hyperscalers and behalf of the neoscalers as well. 561 00:23:34,079 --> 00:23:36,900 One of the biggest opportunities for 562 00:23:36,900 --> 00:23:40,095 Ciena is in our international service provider segment. 563 00:23:40,095 --> 00:23:42,704 Many of the international regions 564 00:23:42,704 --> 00:23:44,100 won't allow the hyperscalers 565 00:23:44,100 --> 00:23:45,570 to build out their own infrastructure. 566 00:23:45,570 --> 00:23:48,254 They have to go to an existing operator. 567 00:23:48,254 --> 00:23:50,414 In places like India, 568 00:23:50,414 --> 00:23:52,874 Middle East, or Southeast Asia, 569 00:23:52,874 --> 00:23:55,499 it's opening up new opportunities for Ciena to 570 00:23:55,499 --> 00:23:56,940 participate with carriers that 571 00:23:56,940 --> 00:23:58,184 traditionally would have gone with, 572 00:23:58,184 --> 00:24:00,315 say, a Huawei or a ZTE. 573 00:24:00,315 --> 00:24:02,009 That's really driving 574 00:24:02,009 --> 00:24:04,769 our international service provider business as well. 575 00:24:04,769 --> 00:24:08,639 It is amazing, this AI supercycle is 576 00:24:08,639 --> 00:24:10,109 creating opportunities across 577 00:24:10,109 --> 00:24:11,790 the board for all operators, 578 00:24:11,790 --> 00:24:14,355 whether it's a hyperscaler or a service provider. 579 00:24:14,355 --> 00:24:16,349 That is why we're seeing 580 00:24:16,349 --> 00:24:17,759 this sustained demand and 581 00:24:17,759 --> 00:24:19,570 it feels very different than 1999, 582 00:24:19,570 --> 00:24:22,049 as one of you mentioned earlier, 583 00:24:22,049 --> 00:24:24,809 around the first time of one of your first OFCs, 584 00:24:24,809 --> 00:24:26,939 where that maybe was a tech bubble. 585 00:24:26,939 --> 00:24:28,889 This feels very different. 586 00:24:28,889 --> 00:24:30,675 One of those reasons is because there's 587 00:24:30,675 --> 00:24:32,039 great opportunities across 588 00:24:32,039 --> 00:24:33,390 the board for all of the players, 589 00:24:33,390 --> 00:24:34,814 so it’s really exciting. 590 00:24:34,814 --> 00:24:36,450 Absolutely. The last thing I 591 00:24:36,450 --> 00:24:37,979 would state is we're still in 592 00:24:37,979 --> 00:24:41,670 the early innings of the AI build out. 593 00:24:41,670 --> 00:24:44,070 As we go to inferencing 594 00:24:44,070 --> 00:24:46,305 and inferencing moves to the edge of the network, 595 00:24:46,305 --> 00:24:47,489 the service provider is going to have 596 00:24:47,489 --> 00:24:49,620 a very critical role because they're the ones that have 597 00:24:49,620 --> 00:24:52,139 the infrastructure to connect the users and 598 00:24:52,139 --> 00:24:55,815 the enterprises into those inferencing locations. 599 00:24:55,815 --> 00:24:57,689 The service providers are going to have 600 00:24:57,689 --> 00:25:00,240 a huge role to play as we move forward here. 601 00:25:00,240 --> 00:25:02,760 As you rightly said, we are just getting started. 602 00:25:02,760 --> 00:25:06,959 This is early days of the whole AI networking buildout. 603 00:25:06,959 --> 00:25:09,000 It does beg the question, Dino, 604 00:25:09,000 --> 00:25:10,215 as you think about 605 00:25:10,215 --> 00:25:12,779 developing this technology – and we already talked 606 00:25:12,779 --> 00:25:14,760 about how the pace of technology is getting faster 607 00:25:14,760 --> 00:25:17,490 – but if you think about the diversity of requirements, 608 00:25:17,490 --> 00:25:20,250 you've got hyperscalers, these emerging cloud players, 609 00:25:20,250 --> 00:25:23,040 service providers, how do you navigate that? 610 00:25:23,040 --> 00:25:25,499 How do you think about product development or 611 00:25:25,499 --> 00:25:29,070 technology development from an R&D strategy perspective, 612 00:25:29,070 --> 00:25:31,170 when there's a lot of 613 00:25:31,170 --> 00:25:33,884 different requirements that we're trying to address? 614 00:25:33,884 --> 00:25:35,699 That's a good point, Rebecca, 615 00:25:35,699 --> 00:25:37,829 and it’s something that’s top of mind. 616 00:25:37,829 --> 00:25:39,479 By the way, did you notice how Brodie 617 00:25:39,479 --> 00:25:41,910 switched to baseball already? 618 00:25:41,910 --> 00:25:44,910 It's gone from hockey to baseball with his analogy. 619 00:25:44,910 --> 00:25:47,219 Hopefully the listeners picked up on that theme. 620 00:25:47,219 --> 00:25:49,409 I'll leave it there for now. 621 00:25:49,409 --> 00:25:53,354 I'm a big Leafs fan. It’s terrible right now. 622 00:25:53,354 --> 00:25:56,110 All right, Brodie, it couldn't help it. 623 00:25:56,960 --> 00:26:01,424 All kidding aside, as Brodie said, 624 00:26:01,424 --> 00:26:03,900 service providers are extremely important to us. 625 00:26:03,900 --> 00:26:06,689 It's easy for people to think that everything we 626 00:26:06,689 --> 00:26:09,929 do is just for the hyperscaler. That's not true. 627 00:26:09,929 --> 00:26:12,720 Having said that, what the hyperscalers bring is 628 00:26:12,720 --> 00:26:16,140 that “hyper” part and the “scale,” as the name suggests. 629 00:26:16,140 --> 00:26:20,100 That huge volume and opportunity is what allows 630 00:26:20,100 --> 00:26:24,435 us to plow heavy investment in foundational technology. 631 00:26:24,435 --> 00:26:27,674 As I said before, with transceivers and amplifiers, 632 00:26:27,674 --> 00:26:30,329 the price tag for developing these technologies, 633 00:26:30,329 --> 00:26:33,390 especially out at the leading edge, is incredible. 634 00:26:33,390 --> 00:26:35,429 Without a large volume 635 00:26:35,429 --> 00:26:37,049 opportunity – you've got to win it, 636 00:26:37,049 --> 00:26:40,454 of course – but without that prize of high volume, 637 00:26:40,454 --> 00:26:41,699 it's very difficult to 638 00:26:41,699 --> 00:26:44,385 justify those levels of investment. 639 00:26:44,385 --> 00:26:47,220 They help us underpin the development 640 00:26:47,220 --> 00:26:49,860 of the core technologies and the engines. 641 00:26:49,860 --> 00:26:52,379 That's what gets you capacity, 642 00:26:52,379 --> 00:26:53,759 while driving cost per bit, 643 00:26:53,759 --> 00:26:56,294 power per bit, and space per bit down. 644 00:26:56,294 --> 00:26:59,160 Who doesn't want that across the patch? 645 00:26:59,160 --> 00:27:01,019 Once you have the engines though, 646 00:27:01,019 --> 00:27:02,820 how we instantiate them in 647 00:27:02,820 --> 00:27:05,384 products and solutions will vary. 648 00:27:05,384 --> 00:27:06,719 I would argue that's 649 00:27:06,719 --> 00:27:09,329 the smaller part of the puzzle, not insignificant, 650 00:27:09,329 --> 00:27:12,419 but it allows us to package those engines into 651 00:27:12,419 --> 00:27:14,429 more hyperscaler-friendly types 652 00:27:14,429 --> 00:27:16,109 of products and appliances 653 00:27:16,109 --> 00:27:18,299 with software interfaces accordingly but 654 00:27:18,299 --> 00:27:20,940 also take the same core ingredients and 655 00:27:20,940 --> 00:27:23,429 package solutions that are better fit for 656 00:27:23,429 --> 00:27:26,369 service providers and wholesale carriers as well. 657 00:27:26,369 --> 00:27:28,830 So that's how I look at the investment. 658 00:27:28,830 --> 00:27:31,350 It's not one thing only. 659 00:27:31,350 --> 00:27:34,230 Make no mistake, without the scale, 660 00:27:34,230 --> 00:27:37,425 it's very hard to justify these investments. 661 00:27:37,425 --> 00:27:41,895 This whole system is hidden everywhere. 662 00:27:41,895 --> 00:27:43,920 That's the great opportunity 663 00:27:43,920 --> 00:27:46,380 for us to have enough volume to 664 00:27:46,380 --> 00:27:48,839 justify investments while being able to take 665 00:27:48,839 --> 00:27:50,220 the emerging technologies and 666 00:27:50,220 --> 00:27:52,424 feed all the consumers of it. 667 00:27:52,424 --> 00:27:54,764 I love the way you think about that. 668 00:27:54,764 --> 00:27:57,315 How does one segment benefit the other, 669 00:27:57,315 --> 00:28:00,255 and how does one customer segment 670 00:28:00,255 --> 00:28:02,579 take advantage of the scale, 671 00:28:02,579 --> 00:28:05,624 the flexibility and architecture from another? 672 00:28:05,624 --> 00:28:08,579 Not only is it smart from an R&D strategy perspective, 673 00:28:08,579 --> 00:28:10,889 but all customers benefit. 674 00:28:10,889 --> 00:28:12,645 They get better technology, 675 00:28:12,645 --> 00:28:15,120 very future forward-looking technology that 676 00:28:15,120 --> 00:28:17,969 can help them evolve their networks over time. 677 00:28:17,969 --> 00:28:20,010 Because it's so fast-paced, 678 00:28:20,010 --> 00:28:23,325 the need for innovation is accelerating with AI. 679 00:28:23,325 --> 00:28:25,709 Before we close, I'd like to 680 00:28:25,709 --> 00:28:28,229 ask one final question of both of you. 681 00:28:28,229 --> 00:28:31,649 If you could leave our listeners today with 682 00:28:31,649 --> 00:28:35,834 one key takeaway about the future of networks, 683 00:28:35,834 --> 00:28:37,709 something they should be paying 684 00:28:37,709 --> 00:28:41,354 close attention to right now, what would it be? 685 00:28:41,354 --> 00:28:44,954 That's a pretty broad, tough question, Rebecca. 686 00:28:44,954 --> 00:28:46,515 I think about how, 687 00:28:46,515 --> 00:28:48,210 despite how proud we are of 688 00:28:48,210 --> 00:28:50,519 our core technologies and competencies, 689 00:28:50,519 --> 00:28:52,515 that nobody can do this alone. 690 00:28:52,515 --> 00:28:55,844 It doesn't matter who you are, even hyperscalers. 691 00:28:55,844 --> 00:28:58,650 Maybe it sounds a little quaint, 692 00:28:58,650 --> 00:29:00,300 the notion of the partnership, 693 00:29:00,300 --> 00:29:02,624 but you have to choose 694 00:29:02,624 --> 00:29:05,939 your teammates and your partners carefully. 695 00:29:05,939 --> 00:29:09,300 If I'm looking at how we comport ourselves with 696 00:29:09,300 --> 00:29:11,489 our supply partners and 697 00:29:11,489 --> 00:29:13,680 how they think about our customers, 698 00:29:13,680 --> 00:29:16,350 they're not just buying a product or 699 00:29:16,350 --> 00:29:19,990 a point solution because the pace is too fast. 700 00:29:20,840 --> 00:29:23,669 The pace of innovation is so quick, 701 00:29:23,669 --> 00:29:26,925 requirements change, opportunities pop up. 702 00:29:26,925 --> 00:29:29,610 I like to think that you’ve got to hitch your wagon 703 00:29:29,610 --> 00:29:33,180 to a partner that's got that innovative spirit, 704 00:29:33,180 --> 00:29:34,919 and is able to make 705 00:29:34,919 --> 00:29:39,195 bold decisions from the technology standpoint. 706 00:29:39,195 --> 00:29:41,129 Sometimes they're scary, but you have to be 707 00:29:41,129 --> 00:29:43,155 able to do that to get the winning hand. 708 00:29:43,155 --> 00:29:44,580 You’ve got to have the deep subject 709 00:29:44,580 --> 00:29:46,409 matter expertise and technology, 710 00:29:46,409 --> 00:29:48,510 and you’ve got to have a proven track record 711 00:29:48,510 --> 00:29:51,285 of being able to execute and deliver. 712 00:29:51,285 --> 00:29:52,874 Choose your partners. 713 00:29:52,874 --> 00:29:54,690 You're not picking a point solution. 714 00:29:54,690 --> 00:29:56,100 Technology moves too fast; 715 00:29:56,100 --> 00:29:57,645 products move too fast. 716 00:29:57,645 --> 00:30:00,090 Pick your partners carefully. 717 00:30:00,090 --> 00:30:02,669 One thing you would want people to think about 718 00:30:02,669 --> 00:30:04,710 is who are you picking in your ecosystem? 719 00:30:04,710 --> 00:30:06,644 Who are you partnering with in the long term? 720 00:30:06,644 --> 00:30:06,929 Yes. 721 00:30:06,929 --> 00:30:09,105 Perfect. Brodie – thoughts? 722 00:30:09,105 --> 00:30:11,579 I think it’s the importance of optics. 723 00:30:11,579 --> 00:30:13,650 If you look at optics, 724 00:30:13,650 --> 00:30:15,375 it's the lowest cost, 725 00:30:15,375 --> 00:30:17,849 lowest power and fastest way to 726 00:30:17,849 --> 00:30:20,519 deliver bits across a network infrastructure. 727 00:30:20,519 --> 00:30:22,229 Nothing's faster than the speed of light. 728 00:30:22,229 --> 00:30:24,254 The importance of optics is 729 00:30:24,254 --> 00:30:26,805 becoming more and more critical. 730 00:30:26,805 --> 00:30:28,755 Optical is going to have a bigger play 731 00:30:28,755 --> 00:30:30,120 within the data center and 732 00:30:30,120 --> 00:30:32,429 more optical technologies are going to go inside 733 00:30:32,429 --> 00:30:35,369 the data center and into the WAN as we move forward. 734 00:30:35,369 --> 00:30:40,544 That definitely would sum up a key takeaway from OFC. 735 00:30:40,544 --> 00:30:41,759 Optical was very front and 736 00:30:41,759 --> 00:30:43,185 center, as we said at the beginning. 737 00:30:43,185 --> 00:30:46,904 We're seeing all of these things converge together, 738 00:30:46,904 --> 00:30:48,194 the need for scale, 739 00:30:48,194 --> 00:30:50,400 these critical latency requirements 740 00:30:50,400 --> 00:30:53,039 for AI and high capacity. 741 00:30:53,039 --> 00:30:55,875 Optical really is sitting at the center of all of it. 742 00:30:55,875 --> 00:30:59,445 That's a really good way of summarizing that, Brodie. 743 00:30:59,445 --> 00:31:01,619 I want to thank both of you for sharing 744 00:31:01,619 --> 00:31:03,764 your insights and perspectives today. 745 00:31:03,764 --> 00:31:06,030 Thank you to our listeners for joining us on 746 00:31:06,030 --> 00:31:08,114 the Ciena Insights Podcast. 747 00:31:08,114 --> 00:31:10,230 You can subscribe on Apple Podcasts, 748 00:31:10,230 --> 00:31:12,314 Spotify or Radio Public. 749 00:31:12,314 --> 00:31:14,100 To find all of our episodes, 750 00:31:14,100 --> 00:31:17,804 visit us at ciena.com/podcast. 751 00:31:17,804 --> 00:31:21,249 Thanks for listening and we'll see you next time.