Tech’s trillion-dollar tumble and 7 other major R&D moves this week

Welcome to our weekly briefing on R&D headlines for January 27, 2025. In this edition, we cover the disruption from China’s DeepSeek reasoning AI model R1, which punches above its weight class. Ostensibly trained on a shoestring budget, it outperforms OpenAI’s o1 by some measures, according to DeepSeek’s technical report. The company also has publicly released its model weights and offers distilled versions of the model. In other news, he U.S. Navy’s potential outsourcing of frontline shipbuilding, Hyundai’s European R&D expansion, Renault’s new EV center in Shanghai, AbbVie’s latest acquisition, South Korea’s ambitious space program funding, and more.

AI & Technology

DeepSeek’s disruption 

As U.S. markets opened for trading on Monday, tech’s megacap companies lost about $1 trillion in market cap at the start of the day, deepening a downturn that sent the Nasdaq into correction territory last week. Investors pointed to a combination of factors, including recession fears and a surprise challenge from Chinese AI upstart DeepSeek, whose newly released opensourced model threatened to undercut the existing U.S. AI ecosystem.

The model is free for the public and costs about 95% less via API use than OpenAI’s o1.

On January 27, Nvidia shed more than $300 billion in market cap at the opening bell, though it quickly regained about half of its loss. Apple and Amazon also opened with substantial dips of $224 billion and $109 billion, respectively, though both regained some ground by the close.

Add those losses to sharp declines in Meta, Microsoft, Alphabet, and Tesla, and the seven most valuable U.S. tech companies collectively shed nearly $1 trillion in early trading.

Tech workers across social media lit up this week after Chinese startup DeepSeek released its latest AI model. Users of the anonymous tech site Blind voiced both enthusiasm and suspicion. Some engineers claimed that DeepSeek’s cost and performance figures “don’t add up,” while others argued it could mark a turning point in affordable AI development.

A forum poll with over 1,200 participants showed tech employees almost evenly split: roughly a third believed DeepSeek was “lying” in its reporting about training costs and methodology, another third were convinced it was legitimate, and the remainder were unsure. A separate, smaller poll had most respondents predicting an “AI bubble” could soon burst.

In user comments, a few Microsoft and Apple staffers mocked the hype, insisting the West still holds a technological edge. Several NVIDIA employees countered that DeepSeek is actually increasing long-term demand for advanced chips. They reasoned that “if the same results are achieved with a third of the compute, imagine how well the model would scale with more.” Still others claimed DeepSeek is sitting on tens of thousands of NVIDIA H100 GPUs—contradicting suggestions that the startup avoided high-end hardware and used NVIDIA’s H100 hardware, which launched more than four years ago.

Conversations also veered into geopolitics. Some posters hurled barbs at perceived “Chinese propaganda,” while others lobbed accusations of “American exceptionalism.” The back-and-forth underscored deeper tensions in the global race to build and monetize powerful AI.

Major news outlets have covered similar debates. CBS News and Reuters reported that DeepSeek’s free AI app overtook ChatGPT in downloads on Apple’s App Store, and that the company spent as little as $6 million on training—far below what U.S. competitors typically pay. Their coverage noted the fierce market reaction: NVIDIA’s share price briefly dipped more than 10%, while other chipmakers and AI-oriented stocks followed suit. Analysts called it a “Sputnik moment,” referencing fears that DeepSeek’s low-cost approach could upend the AI status quo.

“The recent market turbulence in AI-related stocks comes amid a broader debate about AI development and competition. Yann LeCun, Meta’s Chief AI Scientist, offers a nuanced view on the perceived AI race between nations. ‘The correct reading is: Open source models are surpassing proprietary ones,’ LeCun argues, pushing back against claims of Chinese AI supremacy. He points to DeepSeek’s success as evidence of open research’s power, noting how they ‘profited from open research and open source’ including Meta’s PyTorch and Llama. LeCun’s emphasis on collaborative progress rather than national competition provides an interesting counterpoint to market fears. This perspective suggests that the day’s sharp declines in AI-focused companies like NVIDIA (-17.29%) and Oracle (-14.66%) might reflect broader market dynamics rather than fundamental shifts in technological leadership. Moreover, LeCun’s advocacy for open-source development aligns with a growing trend in the AI community, where technological advances often build upon publicly shared research and tools.”

Quantum computing firm Quantinuum heads to New Mexico

Quantum computing firm Quantinuum, Broomfield, Colorado, announced last week the selection of its R&D center in New Mexico. The center will focus on the development of photonics technologies and is expected to open later in 2025. The company was founded in 2021 when Honeywell spun out its Quantum Solutions division and merged it with UK-based Cambridge Quantum Computing. Honeywell owns a 54% stake in the new company, while IBM is also an investor.

Manufacturing & Industrial

U.S. Navy’s potential outsourcing of frontline shipbuilding

The U.S. Navy is likely to outsource the building of its frontline naval vessels in the near future to South Korean conglomerate Hanwha Group, Seoul, due to the extreme shortage of shipbuilding capacity in the U.S. This forecast was made following President Trump’s post-election mention of Korean cooperation in the maintenance, repair and overhaul (MRO) of naval vessels last November. Hanwha completed its $100 million acquisition of the Philly Shipyard in Philadelphia, Pennsylvania, in December.

The U.S. Navy is currently restricted from constructing its ships or any major hull component in a foreign shipyard (per USC 8679). However, President Trump could authorize an exception if it is in the national security interest to do so, which analysts clearly state it is with the current massive Chinese production capability and the very weak by comparison U.S. production capability. According to the Alliance for American Manufacturing and the Korean Ministry of Trade, Industry and Energy, China currently has the capability to build 23 million tons of ships per year, while the U.S. can only produce about 100,000 tons.

Hyundai’s European R&D expansion

Hyundai Motor Company, Seoul, South Korea, announced last week that it is expanding its R&D facilities in Europe with the opening in March 2025 of its Hyundai Motor Europe Technical Center at the Nürburgring in Germany. The company’s latest expansion of its existing facility includes new workshop areas, specialized laboratories, and high-voltage electric vehicle charging capabilities. Hyundai also mentioned that this expansion is part of an upcoming larger facility located in the Rhine-Main region, which will house the company’s large four-wheel noise, vibration, and harshness dynamometer along with advanced chassis and powertrain dynos to simulate vehicle performance without the need for on-road testing.

Renault’s new EV center in Shanghai

Renault Group, Boulogne-Billancourt, France, announced last week that it has begun operating its new dedicated electric vehicle (EV) R&D center in Shanghai, China. The center has 150 staff and is focused on developing battery EVs for the European marketplace. One of the projects the center is working on is the Twingo E-tech prototype, which is scheduled to be launched in 2026 and priced under $20,600. The center has teams dedicated to developing low-cost, high-tech, and responsive supply chains for future EVs. Renault has recently strengthened its ties with CATL, China’s largest EV battery manufacturer.

Li Auto inaugurates first overseas R&D center in Munich

Chinese new energy vehicle manufacturer Li Auto, Beijing, China, announced last week that it inaugurated its first overseas R&D center in Munich, Germany. The Munich center will adopt a localized and collaborative approach to R&D, working alongside Li Auto’s Chinese team to drive development of next-generation technologies, including those related to vehicle design, power semiconductors, smart chassis systems, and electric drivetrains. The center will also work to understand the European market and develop capabilities aligned with European automotive regulations and consumer preferences. This R&D center marks the beginning of Li Auto’s global strategy and will serve as a cornerstone of the company’s AI core strategy.

Healthcare & Pharmaceuticals

AbbVie’s latest acquisition

AbbVie, North Chicago, Illinois, announced last week the acquisition of Nimble Therapeutics, Madison, Wisconsin. The acquisition is expected to enhance AbbVie’s immunology portfolio by incorporating Nimble’s proprietary technology designed for the rapid discovery and optimization of oral peptide candidates. This approach should be particularly applicable for combating targets within autoimmune diseases.

Additional developments

South Korea’s ambitious space program funding

South Korea’s national space agency announced last week that it will invest more than $560 million in R&D projects in 2025 as part of the country’s efforts to become one of five global leaders in the aerospace sector. This investment is a 43.3% increase from its 2024 investment, according to the Korean Aerospace Administration. About a third of this investment will be funneled into R&D projects by the Korean Aerospace Research Institute and the Korean Astronomy and Space Science Institute. Another third will be injected into advancing the country’s space rocket technologies. The final third of the investment will be used to develop a next-generation space rocket and reusable rocket technologies. The Korean space agency also plans to invest in a joint project with the U.S. NASA to launch a Lunar Space Environment Monitor and continue its independent lunar spacecraft development.

R&D World Index Update

The R&D World Index (RDWI) for the week ending January 24, 2025 closed at 3,943.31 for the 25 companies in the RDWI. The Index was up 4.03% (or 152.96 basis points). Seventeen RDWI members gained value last week from 1.49% (Honda Motors) to 14.02% (Oracle). Seven RDWI members lost value last week from -0.14% (Johnson & Johnson) to -3.13% (Apple). One RDW Index member was unchanged for the week (IBM, 224.80).

R&D World’s R&D Index is a weekly snapshot of the top international companies involved in R&D, based on Schonfeld & Associates’ June 2020 R&D Ratios & Budgets. These 25 companies span pharmaceuticals (10), automotive (6), and ICT (9), who collectively invested nearly $260 billion in R&D in 2019—about 10% of all global R&D spending by government, industry, and academia combined, according to R&D World’s 2021 Global R&D Funding Forecast.

R&D World Index Week Ending January 24, 2025

		Ticker	Exchange	2020 R&D Billions $	01/17/25	01/24/25	1/17/25 to 1/24/25	1/1/22 to 1/24/25
1	Alphabet/Google	GOOGL	NASDAQ	27.303	197.55	201.90	2.20%	39.39%
2	Microsoft	MSFT	NASDAQ	17.198	429.03	444.06	3.50%	32.03%
3	Volkswagen AG	VWAGY	OTC	17.259	9.77	10.37	6.14%	-64.49%
4	Apple	AAPL	NASDAQ	18.667	229.98	222.78	-3.13%	25.46%
5	Facebook/Meta	META	NASDAQ	13.874	612.77	647.49	5.67%	92.50%
6	Intel	INTC	NASDAQ	14.503	21.49	20.83	-3.07%	-59.55%
7	Johnson & Johnson	JNJ	NYSE	13.750	147.03	146.82	-0.14%	-14.18%
8	Roche Holdings AG	RHHBY	OTC	14.143	36.35	38.05	4.68%	-26.39%
9	Merck & Co.	MRK	NYSE	11.381	97.92	95.55	-2.42%	24.67%
10	Novartis	NVS	NYSE	9.387	97.53	99.97	2.50%	14.29%
11	Toyota	TM	NYSE	10.724	180.78	186.61	3.22%	0.71%
12	Pfizer	PFE	NYSE	8.336	26.30	26.09	-0.80%	-55.82%
13	Alibaba	BABA	NYSE	6.006	85.12	89.14	4.72%	-24.88%
14	GM	GM	NYSE	6.727	50.97	53.91	5.77%	-8.05%
15	Honda	HMC	NYSE	8.806	28.24	28.66	1.49%	0.74%
16	Ford	F	NYSE	9.340	10.18	10.12	-0.59%	-51.28%
17	AbbVie	ABBV	NYSE	13.949	171.56	170.30	-0.73%	25.78%
18	Sanofi SA	SNY	NYSE	7.750	50.74	52.48	3.43%	4.75%
19	Cisco	CSCO	NASDAQ	6.331	60.23	62.23	3.32%	-1.80%
20	Bristol-Myers Squibb	BMY	NYSE	7.130	56.29	59.67	6.00%	-4.30%
21	Astra Zeneca PLC	AZN	NYSE	6.056	66.60	69.06	3.69%	18.56%
22	IBM	IBM	NYSE	5.368	224.79	224.80	0.00%	68.19%
23	Oracle	ORCL	NYSE	6.928	161.03	183.60	14.02%	110.53%
24	Eli Lilly & Co.	LLY	NYSE	8.606	725.22	785.41	8.30%	184.34%
25	Stellantis NV	STLA	NYSE	5.309	12.88	13.41	4.11%	-31.37%
	Total			269.522	3790.35	3943.31	4.03%	39.95%
	ICT				2021.99	2096.83	3.70%	44.65%
	Automotive				292.82	303.08	3.50%	-11.35%
	Biopharmaceutical				1475.54	1543.40	4.60%	50.10%

:Editor’s note: Brian Buntz wrote the Deep Seek story.