YouTube2Text

【国策の最深部】ラピダス稼働で覚醒する古河電工。AI半導体バブルの「次」を担う光電融合の正体 — Transcript

by 株の気づきチャンネル · 1,962 words · 350 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Concentration of data centers in
  2. 0:02Hokkaido for Rapidus operations. Many
  3. 0:06investors here flock only to
  4. 0:08semiconductors, power, or recently
  5. 0:10optical fibers, but they are actually
  6. 0:12overlooking the most fatal bottleneck
  7. 0:14in AI design. It is not the volume of
  8. 0:19data communication, but the heat
  9. 0:21generated inside the servers. Now, no
  10. 0:24matter how much AI semiconductors
  11. 0:26evolve or how thick the cables
  12. 0:28connecting Hokkaido to Tokyo become,
  13. 0:30there is the massive heat and power
  14. 0:32consumption generated when converting
  15. 0:34electrical and optical signals within
  16. 0:36data centers. Unless this is solved
  17. 0:39from the roots, Japan's communication
  18. 0:41infrastructure will physically collapse
  19. 0:43. The one holding the key to the
  20. 0:46solution—NTT's national strategy, the
  21. 0:49IOWN initiative, and its core
  22. 0:51technology of photonics-electronics
  23. 0:53convergence—is Furukawa Electric,
  24. 0:55which we will dive into today. In the
  25. 0:58second half of this video, we will
  26. 1:00compare the decisive weaknesses of
  27. 1:02Furukawa Electric with its
  28. 1:03market-dominating rival Sumitomo
  29. 1:05Electric, and examine the unique
  30. 1:07strengths that more than compensate for
  31. 1:09them, using specific figures and charts
  32. 1:11. By watching until the end, you should
  33. 1:14gain the perspective to logically
  34. 1:16identify the winners of the industrial
  35. 1:18structure years from now, without being
  36. 1:20swayed by superficial AI-related
  37. 1:22earnings figures. Why is this major
  38. 1:25wire and cable company the next
  39. 1:27protagonist after the AI semiconductor
  40. 1:29bubble? Today, we will unravel the
  41. 1:31heart of this national infrastructure.
  42. 1:34Hello, everyone. This is the Kabu no Ki
  43. 1:36Channel. Please subscribe and turn on
  44. 1:40notifications now so you don't miss the
  45. 1:42next important structural changes or
  46. 1:44the latest analytical updates. Let's
  47. 1:49first organize the background on why we
  48. 1:51should view Furukawa Electric not just
  49. 1:53as a thematic stock, but from the
  50. 1:55broader perspective of structural
  51. 1:57changes in communication infrastructure
  52. 1:59. Currently in Hokkaido, Rapidus is
  53. 2:03building a state-of-the-art
  54. 2:05semiconductor plant in Chitose, and
  55. 2:07Sakura Internet is launching one huge
  56. 2:09national-project data center after
  57. 2:11another in Ishikari. Hokkaido is now
  58. 2:14becoming the central hub for computing
  59. 2:17resources in Japan and Asia. Why
  60. 2:20Hokkaido instead of Tokyo? It is
  61. 2:25because only Hokkaido has the vast land
  62. 2:27, cool climate to cool servers, and
  63. 2:29massive power required for generative
  64. 2:31AI processing. In Tokyo and Osaka,
  65. 2:36power grid capacity has already reached
  66. 2:39its limit, making it physically
  67. 2:41difficult to even build new data
  68. 2:43centers. The next challenge that arises
  69. 2:46from this is distance. Massive amounts
  70. 2:49of AI data processed in Hokkaido must
  71. 2:52be sent instantaneously and without
  72. 2:54delay to Tokyo, the largest demand hub,
  73. 2:57a straight-line distance of over 800km.
  74. 3:00The installation of terrestrial and
  75. 3:02undersea fiber-optic networks to
  76. 3:04connect these points is currently
  77. 3:06proceeding at a rapid pace. However,
  78. 3:09there is a major bottleneck here that
  79. 3:11the market is overlooking. Many people
  80. 3:13think like this: "In that case, we
  81. 3:15should just lay down plenty of thick,
  82. 3:17high-performance cables that can carry
  83. 3:19more data at once.""Therefore, major
  84. 3:22optical cable companies must be
  85. 3:23profitable." Certainly, that
  86. 3:25perspective is not wrong. But the core
  87. 3:28problem is not inside the cables; it is
  88. 3:30happening inside the data centers,
  89. 3:32specifically at the server entry and
  90. 3:34exit points. Currently, light signals
  91. 3:38traveling at the speed of light through
  92. 3:40fiber optics are converted into
  93. 3:42electrical signals just before entering
  94. 3:44the data center servers. This is
  95. 3:46because current semiconductors, such as
  96. 3:49GPUs and CPUs, can only perform
  97. 3:51calculations using electricity. This
  98. 3:54light-to-electricity conversion process
  99. 3:56, known technically as photoelectric
  100. 3:58conversion, actually generates an
  101. 4:00incredible amount of heat. Currently,
  102. 4:03communication speeds are accelerating
  103. 4:06by orders of magnitude to 400G, 800G,
  104. 4:08and beyond. As the volume of data
  105. 4:12explodes, the conversion work at the
  106. 4:14gateway becomes intense, and heat
  107. 4:16generation skyrockets. There is even
  108. 4:19data suggesting that 30%to 40%of the
  109. 4:21enormous power consumed by current data
  110. 4:24centers is taken up not by server
  111. 4:26calculations themselves, but by air
  112. 4:28conditioning to cool the generated heat
  113. 4:30. In other words, no matter how many
  114. 4:33high-performance cables you lay between
  115. 4:36Hokkaido and Tokyo to send massive
  116. 4:38amounts of data, a massive traffic jam
  117. 4:40caused by heat is occurring at the data
  118. 4:42center entrance. If servers go down due
  119. 4:45to heat, they cannot function as
  120. 4:47infrastructure at all. This is the
  121. 4:49physical limit that current
  122. 4:50communication infrastructure is facing.
  123. 4:53As an aside, the technology to simply
  124. 4:55build high-capacity cables will not
  125. 4:57fundamentally solve this heat problem.
  126. 5:00The structure of infrastructure
  127. 5:02investment money is forced to shift
  128. 5:04from the cables themselves toward
  129. 5:05resolving this heat problem. This is a
  130. 5:09physical bottleneck that simply cannot
  131. 5:11be overcome. This is the very reason
  132. 5:13for the existence of NTT's IOWN
  133. 5:15initiative as a national project, and
  134. 5:17it ties directly into the scenario
  135. 5:19where Furukawa Electric, whose
  136. 5:21technology we will compare in detail
  137. 5:23later, holds the key to the next
  138. 5:25generation. I would like to ask you all
  139. 5:27one question here. As you are all well
  140. 5:30aware, the recent frenzy in the
  141. 5:32electric wire sector, especially
  142. 5:34Fujikura, has created a historic bull
  143. 5:37market driven by the AI boom in North
  144. 5:39America. So, even though both are among
  145. 5:42the world's top-tier electric wire
  146. 5:44companies, why is the market's initial
  147. 5:46capital so heavily concentrated in
  148. 5:48Fujikura's physical cables rather than
  149. 5:50Furukawa Electric's optical devices?
  150. 5:52The answer lies hidden not in the
  151. 5:54superficial technological superiority,
  152. 5:57but in the cold, hard fact of a
  153. 5:58divergence in the time horizons that
  154. 6:00market capital is targeting. Let me
  155. 6:03state the conclusion first. The market
  156. 6:05is currently pouring funds into the
  157. 6:07physical lines that are needed
  158. 6:09immediately, but it has not yet fully
  159. 6:10priced in the value of the converters
  160. 6:12that will solve the heat problems
  161. 6:14several years down the line. On the
  162. 6:17screen, I have displayed a comparison
  163. 6:19summarizing the decisive time-horizon
  164. 6:21gap between Fujikura and Furukawa
  165. 6:22Electric, which are currently splitting
  166. 6:24the telecommunications infrastructure
  167. 6:26demand. The meaning of this table is
  168. 6:30very simple. Fujikura's current
  169. 6:32overwhelming strength comes from its
  170. 6:34ultra-high-density cable technology,
  171. 6:36which meets the immediate demand of
  172. 6:38North American tech giants wanting to
  173. 6:40pack as many optical fibers as possible
  174. 6:43into limited underground conduits when
  175. 6:45building data centers. This is
  176. 6:47certainly a magnificent technology, and
  177. 6:49the current rise in their stock price
  178. 6:51is a justified one backed by their
  179. 6:53performance. However, as I mentioned
  180. 6:56earlier, simply packing more cables
  181. 6:58into the conduits does not solve the
  182. 7:00fundamental heat problem. The more the
  183. 7:02volume of data increases, the more the
  184. 7:04power consumption due to heat
  185. 7:05generation will skyrocket. This is
  186. 7:07where the core technology of Furukawa
  187. 7:10Electric, opto-electronic fusion, comes
  188. 7:12into play. Furukawa Electric is a
  189. 7:15global leader in research on technology
  190. 7:17that directly converts light and
  191. 7:19electricity on a semiconductor chip,
  192. 7:21rather than using the large devices
  193. 7:23traditionally required to convert
  194. 7:25electrical and optical signals. In
  195. 7:29NTT’s IOWN initiative, which is a bet
  196. 7:31on the company's future, there is an
  197. 7:33ambitious goal to reduce power
  198. 7:35consumption to one-hundredth of current
  199. 7:38levels; making this physically possible
  200. 7:40is the optical device technology
  201. 7:42centered on high-power CW lasers, where
  202. 7:44Furukawa Electric boasts world-class
  203. 7:47expertise. If I must mention a concern
  204. 7:50here, it is the fact that for many
  205. 7:52years, Furukawa Electric has been held
  206. 7:54back by losses in its automotive parts
  207. 7:56business, such as wire harnesses, and
  208. 7:58the worsening market for metal
  209. 8:00materials. "Even if communications are
  210. 8:02good, the other businesses are failing,
  211. 8:04so I can't buy the stock." This has
  212. 8:06been the primary reason why many
  213. 8:07institutional and individual investors
  214. 8:10have hesitated to invest in Furukawa
  215. 8:12Electric until now. However, please
  216. 8:14look at this data. Furukawa Electric is
  217. 8:16currently pushing forward with
  218. 8:17structural reforms without any
  219. 8:19exceptions. They are carrying out a
  220. 8:21transformation of their business
  221. 8:22portfolio—literally shedding blood—
  222. 8:24by thoroughly downsizing and
  223. 8:26restructuring the unprofitable
  224. 8:27automotive parts business that was
  225. 8:29eating into profits, while
  226. 8:30concentrating management resources on
  227. 8:32information communications, optical
  228. 8:34infrastructure, and energy
  229. 8:35infrastructure. In other words, the
  230. 8:37factors that were holding them back are
  231. 8:39being eliminated. The company is being
  232. 8:41reborn into a strong, lean structure
  233. 8:43where the high profit margins from
  234. 8:45optical-electrical fusion devices
  235. 8:47directly boost the company’s EPS. To
  236. 8:51harness the tremendous computing power
  237. 8:53of next-generation, high-performance
  238. 8:55semiconductors like those manufactured
  239. 8:57by Rapidus without having it killed by
  240. 9:00heat, CPO (Co-Packaged Optics)
  241. 9:02technology, which converts light and
  242. 9:04electricity right next to the chip, is
  243. 9:06absolutely essential. And some of you
  244. 9:09may have realized this after hearing
  245. 9:11what I've said so far. In fact, the
  246. 9:13next-generation semiconductors from
  247. 9:14Rapidus, which is expanding into
  248. 9:16Hokkaido, and Furukawa Electric’s
  249. 9:18optical-electrical fusion are
  250. 9:20completely linked at the foundational
  251. 9:21level. It is said that due to the
  252. 9:24evolution of AI, semiconductors will
  253. 9:26ultimately evolve into a form called
  254. 9:28CPO. This is a technology that directly
  255. 9:31attaches optical communication
  256. 9:33functions—optical transceivers—
  257. 9:35right next to the computing
  258. 9:36semiconductor on the same package. It
  259. 9:39shortens electrical wiring to the limit
  260. 9:41and transmits data directly from the
  261. 9:43chip using light. In short, the artery
  262. 9:46of light connecting Hokkaido to Tokyo
  263. 9:49begins with NEC's undersea cables,
  264. 9:51relies on foundations built by Hokkaido
  265. 9:53Electric Power and Comsys on land, and
  266. 9:56is ultimately processed within the
  267. 9:58server at the millimeter scale by
  268. 10:00Furukawa Electric’s optical fusion
  269. 10:02devices. This is the true face of the
  270. 10:04complete supply chain for the
  271. 10:06infrastructure being promoted as a
  272. 10:08national project. The market is
  273. 10:11currently fixated on the immediate AI
  274. 10:13semiconductor bubble and the
  275. 10:14extraordinary financial results from
  276. 10:16Fujikura. However, the true fruits of
  277. 10:19investment may only be enjoyed by those
  278. 10:21who place their capital not where the
  279. 10:23masses are enthralled, but on the
  280. 10:25resolution of structural bottlenecks
  281. 10:27silently progressing behind that frenzy
  282. 10:29. How was that? This time, I have
  283. 10:32explained the thermal issues facing
  284. 10:34infrastructure amid the massive
  285. 10:36national projects of Rapidus operations
  286. 10:39and data center concentration in
  287. 10:40Hokkaido, along with the key to solving
  288. 10:43them: Furukawa Electric’s optical
  289. 10:45fusion. The biggest mistake to avoid in
  290. 10:48investing is jumping on the bandwagon
  291. 10:50at the peak of a theme that everyone
  292. 10:52already knows. Fujikura's remarkable
  293. 10:56success deserves proper recognition,
  294. 10:58but what we should focus on now is the
  295. 11:00quiet turning point where the
  296. 11:02infrastructure investment phase shifts
  297. 11:04from laying physical lines to devices
  298. 11:06that manage server heat. As a near-term
  299. 11:11event, I want you all to keep an eye on
  300. 11:13the next quarterly earnings
  301. 11:15announcement. What we should track here
  302. 11:18is not the superficial scale of revenue
  303. 11:20. Whether the losses from drag-on
  304. 11:23businesses like automotive parts are
  305. 11:25shrinking as planned, and whether the
  306. 11:27profit margins in the information and
  307. 11:29communications division for optical
  308. 11:31devices are trending upward. The moment
  309. 11:35these two figures cross is when
  310. 11:37Furukawa Electric’s EPS will see a
  311. 11:39dramatic leap. While much of the market
  312. 11:43is captivated by short-term AI hype and
  313. 11:45easy-to-spot stock surges, you have now
  314. 11:47grasped the inevitable physical
  315. 11:49bottlenecks arriving in a few years and
  316. 11:51the structural flow of capital toward
  317. 11:54resolving them. Do you think current
  318. 11:57investment funds should be weighted
  319. 11:59more toward Fujikura, with its visible
  320. 12:02wiring, or Furukawa Electric, with its
  321. 12:04devices that solve future heat issues?
  322. 12:07Please let me know your thoughts on
  323. 12:09portfolio strategies or any hidden
  324. 12:11infrastructure stocks that you think
  325. 12:13could skyrocket in the comments section
  326. 12:14. And finally, I would like to announce
  327. 12:18that memberships have been reopened. I
  328. 12:21have opened memberships on this channel
  329. 12:23to ensure I can continue independent
  330. 12:25research and analysis without being
  331. 12:27influenced by platform trends.
  332. 12:30Currently, this is not a place for
  333. 12:31trading specific investment information
  334. 12:34, but is operated as a venue for purely
  335. 12:36supporting those who agree with the
  336. 12:38principles of my activities. I would
  337. 12:40like to take this opportunity to
  338. 12:42express my deepest gratitude to
  339. 12:43everyone who has already signed up as a
  340. 12:46new member upon this reopening. First,
  341. 12:48I aim to solidify the foundation of
  342. 12:50these activities and continue to
  343. 12:52explore the structure of Japanese
  344. 12:54stocks from a mid-to-long-term
  345. 12:55perspective together with all of you.
  346. 12:58If you agree with the purpose of these
  347. 13:00activities, please consider joining. I
  348. 13:04will see you all in the next video.
  349. 13:07Thank you for watching until the very
  350. 13:09end.

About this transcript

This page contains the full transcript of 【国策の最深部】ラピダス稼働で覚醒する古河電工。AI半導体バブルの「次」を担う光電融合の正体 by 株の気づきチャンネル, generated from the public captions YouTube serves with the video. The transcript has 1,962 words across 350 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.