YouTube2Text

The Nvidia Situation Is Exactly What We Thought... — Transcript

by Spawn Wave · 2,881 words · 408 segments · language en · Watch on YouTube

Full transcript

  1. 0:00All right, this Nvidia situation is
  2. 0:02getting absolutely wild right now. And
  3. 0:04there are many details and headlines
  4. 0:06around more, I guess, salacious details
  5. 0:10and reporting from this. But there is
  6. 0:12something in here that uh look, I get it
  7. 0:15why gamers are frustrated reading some
  8. 0:17of the recent stories that came out in
  9. 0:18the last 24 hours. And there is
  10. 0:21something I want to go over here today
  11. 0:22that could could
  12. 0:26maybe point to
  13. 0:29cheaper video cards v get this a video
  14. 0:32card that might sell
  15. 0:34at MSRP. But we're going to look at that
  16. 0:37here today. So if you guys enjoy the
  17. 0:38video, hit the like button and subscribe
  18. 0:40if you're new. So, first off, if you
  19. 0:42have been living under a rock for the
  20. 0:44last week, there's been a whole
  21. 0:46situation at Nvidia where Jensen Huang,
  22. 0:49who is the co-founder, he's been there
  23. 0:50for 33 years. He went from what was it?
  24. 0:53Denny's co-founded Nvidia and then if
  25. 0:56you look on like his LinkedIn, it was
  26. 0:58like Denny's then to Nvidia as like this
  27. 1:01like a CEO. So, it's it's it's a
  28. 1:03humorous looking LinkedIn account um for
  29. 1:06his like his overall resume, but that's
  30. 1:09the way it went. and he's been running
  31. 1:11Nvidia for decades now. And many people
  32. 1:14will relate him to even some of their
  33. 1:17favorite video cards that have existed
  34. 1:18throughout like all of like gaming
  35. 1:21history for PC. And trust me, Nvidia has
  36. 1:24had some awesome video cards. They've
  37. 1:26also had some missteps here and there,
  38. 1:28we'll say, but he's been he's been with
  39. 1:30them the whole way and up until recently
  40. 1:33has really had a major focus on the
  41. 1:36gaming side. Now it has indeed been
  42. 1:39about uh AI data center buildouts uh
  43. 1:44like deep deep learning all kinds of
  44. 1:46things kind of next to gaming right it
  45. 1:49it does still relate to it in different
  46. 1:51ways and they have worked on some of the
  47. 1:52research and development on like uh
  48. 1:55machine learning and stuff and then
  49. 1:56moving it over through ways of DLSS but
  50. 1:59uh yeah it's there's a reason that
  51. 2:01recently you've heard less about new
  52. 2:03video cars from Nvidia and more about
  53. 2:05these massive chips
  54. 2:07that he held up on stage as a shield
  55. 2:10that would go into a data center
  56. 2:12somewhere supposedly. U but the whole
  57. 2:15thing that really the the reason a lot
  58. 2:18of people are are figuring right now
  59. 2:20that he resigned from Nvidia has more to
  60. 2:23do with the the idea of smuggling ships
  61. 2:26and international relations and in
  62. 2:29China. Not necessarily all the like the
  63. 2:32I think it's three affairs now that have
  64. 2:34been highlighted. Look, the the whole
  65. 2:36affair thing, that's obviously a
  66. 2:38situation of like of itself, right? Like
  67. 2:40on its own over there. And that would be
  68. 2:42something that yeah, the board would
  69. 2:43probably still look at, but let's face
  70. 2:46it, the money is the most important
  71. 2:48thing to shareholders and people on the
  72. 2:52board that would vote. So, they probably
  73. 2:53would overlook some of uh Jensen's
  74. 2:56personal transgressions there. I think
  75. 2:59it does come down to, oh, there there
  76. 3:01are investigations happening. Oh, this
  77. 3:03actually might be a lot worse than we
  78. 3:05were thinking and how fast all this
  79. 3:07happened. I know there's a lot of
  80. 3:09conspiracy theories out there around
  81. 3:10this. I think it's because there is just
  82. 3:13a pretty large absence of information
  83. 3:16for us like right now. So, we're just
  84. 3:18kind of observing this from the outside
  85. 3:19looking in. Um, but no doubt there are
  86. 3:22people like uh people who are high up at
  87. 3:24other companies who were friends with
  88. 3:26Jensen, who had only good experiences
  89. 3:27with him, and so they're probably trying
  90. 3:28to back him up. It's like, "No, no,
  91. 3:30don't don't let him walk away from
  92. 3:32Nvidia." But then there are also people
  93. 3:33who are high up at some of these
  94. 3:34companies that probably had deals
  95. 3:37getting done by Jensen that are in the
  96. 3:39tens of billions of dollars and are
  97. 3:40linked to AI and data centers. and like
  98. 3:43hold on kind of need him for those deals
  99. 3:46to get done because I don't know what
  100. 3:48his replacement will be doing at Nvidia.
  101. 3:51So that I think is what a lot of this is
  102. 3:53around and he's being called potentially
  103. 3:55to Congress to testify over. So a lot of
  104. 3:58questions, a lot of information still
  105. 3:59needs to get out there. But there was
  106. 4:01one report that popped up, okay, and
  107. 4:03this is covered, I mean is going viral
  108. 4:06online. Nvidia allegedly engineered
  109. 4:08scarcity by funding AI startups to buy
  110. 4:10its GPUs. This has been discussed for
  111. 4:14quite some time now is the idea of
  112. 4:16Nvidia with their circular financing.
  113. 4:19There seems to be like actual evidence
  114. 4:21of things happening in terms of like
  115. 4:24internal meetings or transcripts. So
  116. 4:27they say following the departure of the
  117. 4:29former Nvidia CEO Jensen Hong, new
  118. 4:31report by Wall Street Journal has
  119. 4:33emerged alleging Nvidia artificially
  120. 4:34inflated the ongoing GPU and memory
  121. 4:37shortage by investing in AI startups
  122. 4:40that would go on to buy Nvidia GPUs and
  123. 4:42build data centers with the reasoning
  124. 4:46to keep up the the perception of high
  125. 4:49demand and to move an overupp of
  126. 4:51semiconductors from TSMC. So yeah, seems
  127. 4:55like they themselves may have
  128. 4:56overordered on it and they have to
  129. 4:58figure out how to disperse it to the
  130. 4:59market. And yeah, it kind of comes down
  131. 5:01to Nvidia buying the stuff from
  132. 5:03themselves. But the Wall Street Journal
  133. 5:06report cites an August meeting between
  134. 5:08Huang and his senior finance team,
  135. 5:10including Envia's CFO and acting CEO
  136. 5:13now, Colette Crest, during which Crest
  137. 5:16pushed for a reduction in circular
  138. 5:18financing agreements in response to
  139. 5:20which Hang allegedly laid into Crest,
  140. 5:22calling her a deer in the headlights
  141. 5:24after she had pushed to scale back the
  142. 5:28$250 billion deal with Open AI. So here
  143. 5:32was here's basically the idea is Nvidia
  144. 5:35has x amount of over supply we'll say
  145. 5:38for semiconductors from TFMC has to do
  146. 5:41something with them. So they loan money
  147. 5:45we'll say to smaller AI companies that
  148. 5:48are startups and then they take that
  149. 5:50money and hand it back to Nvidia in
  150. 5:52exchange for chips that would be going
  151. 5:54into data centers that they'd be
  152. 5:56standing up because right now one of the
  153. 5:58big things behind the scenes is selling
  154. 6:00compute. Basically, if you can build a
  155. 6:02data center that functions and is able
  156. 6:04to to generate compute that would then
  157. 6:06be sold to different companies like say
  158. 6:09OpenAI or Anthropic or others. That's
  159. 6:12kind of been the plan. And so data
  160. 6:14centers are like the hot thing right
  161. 6:16now. Everyone's trying to build them.
  162. 6:18But the issue that many are running into
  163. 6:21is one obviously local governance. And
  164. 6:23if you know people want a data center
  165. 6:26somewhat nearby their neighborhood or
  166. 6:28just within you know vicinity of it
  167. 6:30there's a lot of drawbacks to that. Yes.
  168. 6:31But two seems to be more about the
  169. 6:35energy grid in general like being able
  170. 6:37to support a ton of these data centers.
  171. 6:40All right. So with that in mind, there
  172. 6:43have been reports and allegations that
  173. 6:45many of the chips that Nvia then sells
  174. 6:47to some of these companies that like
  175. 6:49these smaller companies that can't get
  176. 6:51these data centers set up right now,
  177. 6:53those chips, GPUs, boards, or just RAM
  178. 6:58are just sitting in warehouses.
  179. 7:01And that's kind of it. Like a big one
  180. 7:03right now that was brought up was I saw
  181. 7:05Microsoft being brought up, Amazon, and
  182. 7:08they're just sitting on stock that
  183. 7:10they've purchased maybe from Nvidia
  184. 7:12here. And why is that frustrating for
  185. 7:14gamers?
  186. 7:16Well, what's what's the weird thing
  187. 7:18that's been happening
  188. 7:20over the last several years now? things
  189. 7:22have gotten more expensive when it comes
  190. 7:24to electronics, especially things that
  191. 7:27well, if we're going through a
  192. 7:28generation, say for a PS5 or uh I mean,
  193. 7:31you really say PS5, Xbox, Switch, any of
  194. 7:34them, right? They go down in price and
  195. 7:36it's been like that every single time
  196. 7:38until now. Video cards, video cards used
  197. 7:42to depreciate very, very quickly. I
  198. 7:44mean, you you were a generation or two
  199. 7:46behind and you were picking cards that
  200. 7:49would have been like several times more
  201. 7:52expensive, you know, years prior, and
  202. 7:54now they're they're 100 bucks. It was
  203. 7:56very common if you wanted to stay a
  204. 7:58generation behind for GPUs to have a
  205. 8:00pretty good deal on them. And that's
  206. 8:02just not the case anymore. But this
  207. 8:04would directly impact
  208. 8:06I mean just removing supply obviously
  209. 8:09from a market that maybe demand is
  210. 8:11heightened a bit because we are seeing
  211. 8:13PC gaming continue to build and build
  212. 8:15and build but not enough of a demand
  213. 8:17heighten to uh or demand raise to get to
  214. 8:21this position where it's just
  215. 8:22overwhelming like we can't keep up. It's
  216. 8:24like, well, actually, no. If if we do
  217. 8:26see this significant amount of circular
  218. 8:29financing to then buy chips and and
  219. 8:32everything from Nvidia, we know where
  220. 8:34their focus is right now. They used to
  221. 8:35be primarily gaming revenue for their
  222. 8:38GPUs. Now, that is a small sliver of the
  223. 8:41pie compared to everything they're doing
  224. 8:43with data centers. So, when you look up
  225. 8:45things like here's video cards. This is
  226. 8:47Best Buy
  227. 8:49over in Delaware. And this is uh this is
  228. 8:51what they have listed on the East Coast
  229. 8:53right now. And this would be a 5070. Not
  230. 8:57an old card. The 50 series is uh I don't
  231. 9:00it's almost two years old now. So, think
  232. 9:02about it that way, right? But still, you
  233. 9:04feel like, okay, we're going to be
  234. 9:05getting to the next series of cards and
  235. 9:07these will start to drop in price. Well,
  236. 9:09the 5070
  237. 9:11it it released what was it? $500ish
  238. 9:13dollars and here it is now for $879
  239. 9:19900. It's got 12 gigabytes GDDR7 on
  240. 9:21board and then you think about what
  241. 9:23about 570 Ti remember that came up and
  242. 9:25it was a big thing and it was $579 I
  243. 9:27think. I mean it's there's a 570 Ti for
  244. 9:32$1,500 which is ridiculous. The 9070 XT
  245. 9:35from AMD was even being viewed as, okay,
  246. 9:38maybe that's a card we can go to because
  247. 9:40Nvidia is just out to lunch when it
  248. 9:41comes to providing anything somewhat
  249. 9:44affordable for video cards. And the a
  250. 9:46the secondhand market, the aftermarket
  251. 9:48is just going crazy with the lack of
  252. 9:50supply. The 970 XT is up to over $1,000.
  253. 9:54That was a $600 card
  254. 9:56like a year and a half ago.
  255. 9:59So this this is what I mean and this is
  256. 10:00why gamers right now are seeing the
  257. 10:02fallout of Nvidia and Jensen Hong. And
  258. 10:05while there is still some back and forth
  259. 10:07online, you're going to have many people
  260. 10:08who are just do not feel bad about it.
  261. 10:10Like, oh, finally, maybe we can have a
  262. 10:12shakeup. Cuz I think that's the thing. I
  263. 10:13think many gamers who are trying to keep
  264. 10:15up with PC gaming and maybe they want to
  265. 10:17upgrade. They're just ready for a
  266. 10:19shakeup because what's happening right
  267. 10:21now was just not working for them, you
  268. 10:22know? So like the 5090 is the most
  269. 10:24egregious thing I've ever seen. And a
  270. 10:26lot of it does have to do with the 32 GB
  271. 10:29of GDR7 that it has on board, but it was
  272. 10:31a $2,000 card less than two years ago.
  273. 10:35And they started to come back in stock,
  274. 10:37but you can see there here's $8,000 for
  275. 10:40a a 5090. It's like, what what is that?
  276. 10:44The 5080 is a 16 gigabyte card GTR7.
  277. 10:48Sure, but that's $2,500 now. That was
  278. 10:51$1,000, you know, less than two years
  279. 10:53ago.
  280. 10:54you think, "All right, well, let's go
  281. 10:56back to the used market." Here's the
  282. 10:58thing, though. Because the overall
  283. 11:00market for new cards have has climbed so
  284. 11:03much, even cards that are a generation
  285. 11:06or two older really haven't gone down in
  286. 11:08price that much. The 4070,
  287. 11:12I mean, here's a Dell Alienware 4070 you
  288. 11:16can get from Nvidia at $665,
  289. 11:19Founders Edition at at 700. And look,
  290. 11:22you could say, "Well, these people
  291. 11:23shouldn't be selling the the card for
  292. 11:25that much." I mean,
  293. 11:27at this point, they're probably
  294. 11:28thinking, "Hey, maybe I can take
  295. 11:29advantage of the market myself before it
  296. 11:31crashes or whatever else happens." So,
  297. 11:33you really can't blame people who are
  298. 11:35kind of f following along with market
  299. 11:37trends because their own product has
  300. 11:39just appreciated overall. But $750 for a
  301. 11:444070, like come on. And I thought, you
  302. 11:46know what, the 3080 still a pretty good
  303. 11:47card. Somehow it's still $500. Not even
  304. 11:51the 3080 Ti, just the 3080. Do you
  305. 11:53remember how excited people were for
  306. 11:55that card six years ago when Nvidia
  307. 11:59showed the price tag on stage and it was
  308. 12:01700? And you know what? Everything they
  309. 12:03showed. This is a good card, but it's
  310. 12:07not $500 good six years later. It's not.
  311. 12:11This would have been a situation where
  312. 12:12you would have found this card kicking
  313. 12:14around for like $200ish dollars, you
  314. 12:17know, six years later. That would have
  315. 12:19made more sense. And it just goes on and
  316. 12:21on and on and I just think back to older
  317. 12:22cards that you would look to as like I
  318. 12:25was just going through the history some
  319. 12:26of the older video cards that I remember
  320. 12:28as being awesome. The 970 it went
  321. 12:30through some struggles. I'll say there
  322. 12:33was a whole controversy around the 4 GB
  323. 12:35of GDDR5. Technically it was 3 12 GB but
  324. 12:38it launched back in 2014 at a price of
  325. 12:41$330. So some people will kind of
  326. 12:44mention the inflation thing. Go ahead
  327. 12:46and put it through your inflation
  328. 12:47calculator. Tell me what that would be
  329. 12:49now. it'd be actually around what the
  330. 12:515070,
  331. 12:52you know, would have been selling for.
  332. 12:54So, that makes sense at its initial
  333. 12:55MSRP, but we're well beyond that at this
  334. 12:58point. I mean, even the 1070, which was
  335. 13:01also a really good card with 8 GB of uh
  336. 13:03of video RAM and everything, that was
  337. 13:05380 back in 2016, going back, you know,
  338. 13:09just over 10 years ago. And one of my
  339. 13:11favorite car, I thought this was an
  340. 13:12awesome move from uh AMD back in 2016.
  341. 13:16Something I really miss
  342. 13:18with sort of the direction of PC gaming,
  343. 13:21but just the again the the market is
  344. 13:24that attempt at a budget card. Because
  345. 13:26even if you are Nvidia or AMD right now,
  346. 13:29there's really no reason for you to
  347. 13:31attack the budget market because again,
  348. 13:35a lot of people who are trying to even
  349. 13:36run log models themselves would aim for
  350. 13:40a more expensive video card because of
  351. 13:42the additional VRAM. But in like a
  352. 13:45normalized world, we probably would see
  353. 13:47more of this where you would have an
  354. 13:50RX480 type card come out and it attempts
  355. 13:53to be at $200. This is the 8 GB card at
  356. 13:55$230. That's probably a better value now
  357. 13:58that I think about it. But there was a 4
  358. 13:59GB card because AMD was really trying to
  359. 14:02get in there at the $199 mark and they
  360. 14:05were able to. an RX480, four gigabytes
  361. 14:08of I think it's GDDDR5 and it was $200
  362. 14:11and that was a very very good 1080p
  363. 14:13card. But there's just no reason to
  364. 14:15attempt this right now because the RAM
  365. 14:18itself due to the shortage would be more
  366. 14:21than that. You like 8 GB of GDR7, it
  367. 14:24probably makes that completely undoable
  368. 14:27or even just GDDR6. And once again, we
  369. 14:30come all the way back to the idea of
  370. 14:32there being an artificial
  371. 14:34uh artificial scarcity being created by
  372. 14:38the circular financing. Now, it seemed
  373. 14:40like this uh the acting CEO right now,
  374. 14:43Colette Crest, was not as excited about
  375. 14:46keeping it going. So, maybe this really
  376. 14:49is the move to where this person would
  377. 14:52come in and maybe cut back on some of
  378. 14:54that. And man, what a scary idea, isn't
  379. 14:57it?
  380. 14:58more supply coming in. Video cards close
  381. 15:01to MSRP from
  382. 15:04two years ago, three years, four years
  383. 15:06ago at this point, years and years ago.
  384. 15:08Uh what a world that would be. But I so
  385. 15:11much stuff is happening with Nvidia.
  386. 15:13Things seem to be unraveling. It's it's
  387. 15:15an it's a crazy situation that only
  388. 15:18continues to get crazier as we go along.
  389. 15:21But I wanted to bring that up because
  390. 15:22for PC gamers out there who have been
  391. 15:25frustrated with the current market and
  392. 15:27the direction of it and maybe you're
  393. 15:29still sitting on a 30 series card
  394. 15:30waiting for prices to become I don't
  395. 15:33know somewhat feasible. This might be a
  396. 15:37signal that we are at least heading in
  397. 15:38that direction. But you got to keep an
  398. 15:40eye on the Nvidia situation to see what
  399. 15:42else comes out and potentially
  400. 15:45I guess what strategy they could have
  401. 15:47for a permanent CEO whether it's this
  402. 15:49Colette Crest person or someone else
  403. 15:51they bring in and what exactly they say
  404. 15:54about the gaming and then also data
  405. 15:56center market. But let me know what you
  406. 15:58guys think about this one down below in
  407. 15:59the comments. Thank you so much for
  408. 16:00watching and I'll see you next time.

About this transcript

This page contains the full transcript of The Nvidia Situation Is Exactly What We Thought... by Spawn Wave, generated from the public captions YouTube serves with the video. The transcript has 2,881 words across 408 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.