Latest / DX Today | No-Hype Podcast & News About AI & DX / AI News Today: SpaceX Seeks $40B to Buy Nvidia Chips

Transcript

AI News Today: SpaceX Seeks $40B to Buy Nvidia Chips · DX Today | No-Hype Podcast & News About AI & DX

• 0:00
  SPEAKER_05: It's Wednesday, October the 7th, 2026.
• 0:03
  SPEAKER_05: You're listening to the DX Today AI Daily Brief.
• 0:07
  SPEAKER_05: Today, SpaceX looks to borrow$40 billion to buy NVIDIA chips.
• 0:12
  SPEAKER_05: France's Mistral unveils a$1 trillion parameter model, and two Chinese AI labs line up giant raises on the road to public listings.
• 0:23
  SPEAKER_05: Let's get into it.
• 0:24
  SPEAKER_00: SpaceX is seeking to raise about$40 billion to buy NVIDIA artificial intelligence chips, according to a Financial Times report published Tuesday and confirmed by other outlets citing people familiar with the matter.
• 0:37
  SPEAKER_00: The financing would be led by the asset manager, Apollo Global Management.
• 0:42
  SPEAKER_00: It reportedly breaks down into roughly$10 billion in bank loans and$30 billion in investment grade debt, and the deal was expected to close next year.
• 0:51
  SPEAKER_00: Neither company has publicly confirmed the terms.
• 0:54
  SPEAKER_00: If it goes ahead, it would be one of the largest debt packages ever assembled for a single chip order, and a sign of how far Elon Musk's companies are prepared to lean on credit markets to secure computing power.
• 1:07
  SPEAKER_02: Now to Europe.
• 1:08
  SPEAKER_02: French AI developer Mistral on Tuesday unveiled Mistral Large 4, a 1 trillion parameter model the company has nicknamed Le Chanc.
• 1:19
  SPEAKER_02: It is the company's first major release in about five months.
• 1:23
  SPEAKER_02: The model is available now as a public preview through Mistral's application programming interface, and the company says it will publish the open weights later this month, with news reports citing October 27th.
• 1:38
  SPEAKER_02: Mistral says the model ranks among the strongest open systems in the world, with particular strength in cybersecurity work, and it is pitching the release as a European answer to leading open models from Chinese developers.
• 1:55
  SPEAKER_02: Those performance claims come from Mistral itself and have not yet been independently tested.
• 2:01
  SPEAKER_02: The launch arrives as European governments push for sovereign AI options that do not depend on American or Chinese suppliers.
• 2:10
  SPEAKER_04: Next, a cloud provider eyes the public markets.
• 2:13
  SPEAKER_04: Lambda, the NVIDIA-backed AI cloud company, is raising up to$4 billion at a pre-money valuation of$14.5 billion.
• 2:22
  SPEAKER_04: The Wall Street Journal reported Tuesday.
• 2:25
  SPEAKER_04: COTU Management and Blackstone are leading the round.
• 2:28
  SPEAKER_04: According to a letter to investors reviewed by the journal, this is expected to be Lambda's final private fundraising before a planned initial public offering in 2027, depending on market conditions.
• 2:40
  SPEAKER_04: Lambda rents access to large clusters of graphics processors to developers and enterprises training and running AI models.
• 2:48
  SPEAKER_04: The company joins a crowded field of so-called neo clouds, competing with the biggest cloud platforms for scarce chips.
• 2:55
  SPEAKER_04: The round has not been officially confirmed by Lambda, and the final size could change before it closes.
• 3:01
  SPEAKER_04: Turning to China.
• 3:02
  SPEAKER_01: Chinese AI startup DeepSeek is set to raise at least 80 billion yuan, or roughly$12 billion, in its latest funding round, Bloomberg reported Tuesday, citing people familiar with the matter.
• 3:14
  SPEAKER_01: Tencent and the battery giant CATL are among the largest backers, and the total could reach as much as$15 billion.
• 3:23
  SPEAKER_01: DeepSeek began the fundraising in July as it builds a war chest for computing capacity ahead of a possible domestic stock listing in early 2027.
• 3:32
  SPEAKER_01: The company shook up the industry last year with low-cost reasoning models that matched American rivals on many benchmarks.
• 3:39
  SPEAKER_01: The round and the listing timeline remain reported plans rather than confirmed deals, and DeepSeek has not commented publicly.
• 3:47
  SPEAKER_05: Another Chinese lab is close behind.
• 3:49
  SPEAKER_03: Moonshot AI, the Beijing company behind the Kimmy chatbot, has closed its final private funding round at a valuation of about$50 billion, Bloomberg reported Tuesday.
• 4:01
  SPEAKER_03: According to the report, Moonshot is now preparing an initial public offering in Hong Kong as early as the first quarter of 2027, and the timeline could still change.
• 4:11
  SPEAKER_03: The company has gained attention for its Kimi models, which compete with leading systems on coding and agent tasks.
• 4:18
  SPEAKER_03: Together with the Deep Seek Rays, the news shows Chinese AI developers racing to secure capital and public market access at the same time that their American counterparts pursue record private rounds.
• 4:30
  SPEAKER_03: Moonshot has not publicly confirmed the valuation or the listing plans.
• 4:35
  SPEAKER_05: Now, a big chip forecast.
• 4:37
  SPEAKER_00: Marvel technology laid out its most ambitious targets yet at its investor day on Tuesday.
• 4:42
  SPEAKER_00: The chipmaker set a fiscal 2031 revenue target of$70 to$90 billion, far above the roughly$47 billion analysts had expected, and raised its fiscal 2028 goal to about$20 billion.
• 4:57
  SPEAKER_00: Marvel said demand for custom AI processors and the high-speed interconnects that link data center chips together is driving the outlook.
• 5:05
  SPEAKER_00: The company expects its interconnect business to supply a large share of that growth.
• 5:09
  SPEAKER_00: Shares jumped nearly 10% during Tuesday's session.
• 5:12
  SPEAKER_00: The forecast is the latest signal that suppliers beyond the biggest chip designers expect the data center spending boom to continue for years.
• 5:20
  SPEAKER_02: Over to Google.
• 5:21
  SPEAKER_02: Google on Tuesday released NanoBanana 2.1, an updated version of its image generation and editing model.
• 5:30
  SPEAKER_02: The company says the new model improves visual design, mask-based editing, and keeping subjects consistent across images, and it can work from up to 14 reference images and produce pictures in 4K resolution.
• 5:46
  SPEAKER_02: Google also cuts standard image output prices in its developer platform roughly in half compared with the previous version.
• 5:54
  SPEAKER_02: The model is rolling out across the Gemini app, AI mode in Google Search, AI Studio, and Google's advertising tools.
• 6:03
  SPEAKER_02: Lower prices matter for businesses that generate large volumes of marketing images, and the release keeps pressure on rival image tools as the cost of visual content creation keeps falling.
• 6:15
  SPEAKER_04: A partnership deepens.
• 6:17
  SPEAKER_04: Atlassian and OpenAI are expanding their partnership, bringing OpenAI's newest frontier models, including GPT-6 Astra, across Atlassian's platform and its Rovo AI assistant.
• 6:29
  SPEAKER_04: The companies announced the expanded agreement Tuesday.
• 6:32
  SPEAKER_04: It builds on a relationship that began in 2023.
• 6:36
  SPEAKER_04: The goal is to let AI agents act on the knowledge already stored in tools like Jira and Confluence, turning plans, tickets, and documents into completed work.
• 6:46
  SPEAKER_04: VentureBeat reported that Atlassian made a spending commitment as part of the deal.
• 6:51
  SPEAKER_04: Though OpenAI's announcement did not disclose financial terms, Atlassian says its platform will remain open to models from multiple providers.
• 6:59
  SPEAKER_04: For enterprise software makers, the deal reflects a broader shift from chat assistance toward agents embedded directly in everyday workflows.
• 7:07
  SPEAKER_04: In Europe, a notable acquisition.
• 7:09
  SPEAKER_01: SAP has agreed to acquire TechWolf, a company based in Ghent, Belgium, that uses AI to map the skills and tasks of a company's workforce.
• 7:18
  SPEAKER_01: The two companies announced the deal on Tuesday, and the terms were not disclosed.
• 7:24
  SPEAKER_01: TechWolf analyzes data from the tools employees already use to build a continuously updated picture of who does what inside an organization.
• 7:33
  SPEAKER_01: SAP says the deal will give its customers evidence-based insight into how work is changing as AI takes on more tasks.
• 7:43
  SPEAKER_01: TechWolf will continue to operate with its own brand and team after the deal closes, according to reports.
• 7:49
  SPEAKER_01: The deal is still subject to customary closing conditions.
• 7:53
  SPEAKER_05: Now, Workforce News.
• 7:54
  SPEAKER_03: HubSpot is cutting about 660 jobs, roughly 7% of its global workforce.
• 8:00
  SPEAKER_03: The company disclosed Tuesday in a regulatory filing.
• 8:04
  SPEAKER_03: Chief Executive Yamini Rangan told employees the company is shifting from building software toward delivering outcomes for customers with AI.
• 8:12
  SPEAKER_03: A change, she said, is transforming product, pricing, and customer service.
• 8:16
  SPEAKER_03: Rangan also wrote that the cuts were not driven by AI-related efficiencies.
• 8:21
  SPEAKER_03: HubSpot expects$65 to$75 million in related charges, mostly in the fourth quarter, and reaffirmed its financial guidance for the third quarter and full year.
• 8:32
  SPEAKER_03: The move adds HubSpot to a growing list of software companies reorganizing their workforces around AI products, even as they say automation is not the direct cause.
• 8:42
  SPEAKER_05: Next, a chip push in Seoul.
• 8:44
  SPEAKER_00: AMD Chief Executive Lisa Su met South Korean officials and 14 Korean AI companies in Seoul on Wednesday.
• 8:51
  SPEAKER_00: The aim is to build combined computing infrastructure that pairs AMD processors and graphics chips with domestically made AI chips.
• 8:59
  SPEAKER_00: South Korea's science ministry said the effort would be tied to domestic demonstration projects and could produce a model that Korean chip companies can take to global markets.
• 9:08
  SPEAKER_00: The participants span chip design, memory, networking, cloud, and AI model development.
• 9:14
  SPEAKER_00: The meeting follows a cooperation agreement reached during a summit in San Francisco in July.
• 9:19
  SPEAKER_00: Sue has said demand for AI chips will last for years and that AMD is securing manufacturing capacity three to five years ahead.
• 9:26
  SPEAKER_05: Finally, a startup funding round.
• 9:29
  SPEAKER_02: Vinci, a Palo Alto startup that uses AI to simulate the physics of chips and other hardware, has raised$250 million in a Series B round at a valuation of$1.5 billion.
• 9:44
  SPEAKER_02: The company announced the round Tuesday.
• 9:47
  SPEAKER_02: It was co-led by Advent, TemSec, and Zora.
• 9:51
  SPEAKER_02: Vinci Software runs high-fidelity physics simulations that engineers use to test chip and hardware designs before they are built.
• 10:01
  SPEAKER_02: The company says it will use the money to expand its suite of simulation software.
• 10:07
  SPEAKER_02: The raise comes less than a year after Vinci emerged from stealth, and it reflects investor appetite for AI tools aimed at the engineering bottlenecks behind the chip boom itself.
• 10:23
  SPEAKER_05: For DX today, stay curious.