Powered by LumidaWealth.com
Lumida News
  • Home
  • EarningsNEW
  • News
    • Alt Assets
    • Crypto
    • Equities
    • Macro
    • Markets
    • Real Estate
  • Lifestyle
    • Family Office
    • Health and Longevity
  • Themes
    • Aging & Longevity
    • AI
    • CRE
    • Digital Assets
    • Legacy Brands
    • Nuclear Renaissance
    • Private Credit
  • About Us
No Result
View All Result
Lumida News
  • Home
  • EarningsNEW
  • News
    • Alt Assets
    • Crypto
    • Equities
    • Macro
    • Markets
    • Real Estate
  • Lifestyle
    • Family Office
    • Health and Longevity
  • Themes
    • Aging & Longevity
    • AI
    • CRE
    • Digital Assets
    • Legacy Brands
    • Nuclear Renaissance
    • Private Credit
  • About Us
No Result
View All Result
Lumida News
No Result
View All Result
  • Lumida Wealth
  • Lumida Ledger
  • LUMIDA ETF
  • About Us
Home Themes AI

The Market Misread Kimi K3 the Same Way It Misread DeepSeek — Here’s What the Semiconductor Selloff Got Wrong

by Team Lumida
July 20, 2026
in AI
Reading Time: 5 mins read
A A
0
A person holding a cell phone in their hand

Photo by Solen Feyissa on Unsplash

Share on TelegramShare on TwitterShare on FacebookShare on LinkedinShare on Whatsapp
  • Moonshot AI’s Kimi K3 triggered a sharp selloff in semiconductor stocks on its release last Friday — the same pattern seen when DeepSeek’s R1 debuted in early 2025 and wiped nearly $600 billion from Nvidia’s market cap in a single session — but Bloomberg analysis argues the comparison fundamentally misreads what Kimi K3 actually does; DeepSeek’s breakthrough was making AI cheaper to both train AND run, reducing the total compute and memory infrastructure required to operate frontier-class AI; Kimi K3 improves computational efficiency through a high sparsity ratio (fewer parameters are activated per task relative to the model’s total 2.8 trillion parameter count), but it does not reduce memory requirements — it dramatically increases them; those 2.8 trillion parameters must all be stored in memory regardless of how efficiently the active computation is handled.
  • Kimi K3 is China’s largest AI model ever, with 2.8 trillion parameters and the highest sparsity ratio yet recorded among disclosed AI models, making it a genuine frontier achievement in computational efficiency; but even after compressing the model using lower-precision data formats (a standard technique for reducing memory footprint), Kimi K3 occupies approximately 1.4 terabytes of memory — a figure that is not meaningfully reducible without degrading model capability; deploying Kimi K3 at scale therefore requires clusters of memory-intensive AI processors such as Nvidia’s Blackwell GB300 NVL72 systems, which are among the most expensive and hardware-intensive AI accelerators currently available; the model’s GPU demand has already been validated in the market — Moonshot AI’s Kimi platform announced it was temporarily pausing new subscriptions because demand in the 48 hours after launch “pushed close to the limits of our current capacity.”
  • Alibaba compounded the Chinese AI moment over the weekend by unveiling its Qwen 3.8-Max preview, a 2.4 trillion-parameter model that the company says performs alongside leading frontier models — though Alibaba has not yet published benchmark results; as Chinese AI developers increasingly tailor their models to domestic chips, China’s domestic hardware ecosystem stands to benefit from the inference demand their models generate; the dual release of Kimi K3 and Qwen 3.8-Max sent China’s tech-heavy ChiNext Index up as much as 3.6% on Monday, reflecting investor optimism that Chinese AI companies are now building at the frontier tier, with the domestic hardware and cloud infrastructure ecosystem positioned to capture the resulting inference revenue.
  • The broader strategic implication cuts both ways: competition in advanced AI is unambiguously broadening beyond US frontier labs like OpenAI and Anthropic, which increases competitive pressure on US model providers’ pricing power and erodes the premium they command for frontier-tier capabilities; but the memory and hardware requirements of models at this parameter scale — whether built in the US or China — continue to support investment in the infrastructure layer, particularly high-bandwidth memory (SK Hynix is the dominant supplier), advanced packaging and silicon (TSMC), and GPU systems (Nvidia); the Kimi K3 misread parallels the DeepSeek misread in that a Chinese efficiency breakthrough initially reads as bad for semiconductor demand when the actual analysis supports the opposite conclusion.

What Happened?

Moonshot AI released Kimi K3, China’s largest AI model with 2.8 trillion parameters, triggering a semiconductor stock selloff similar to DeepSeek’s debut in early 2025. Bloomberg analysis argues the selloff misread the model’s implications: unlike DeepSeek, which reduced both compute and memory demands, Kimi K3’s compute efficiency improvement is offset by dramatically higher memory requirements — approximately 1.4 terabytes compressed — requiring deployment on Nvidia’s most memory-intensive Blackwell systems. Moonshot also announced it is pausing new subscriptions due to GPU capacity constraints. Alibaba simultaneously unveiled Qwen 3.8-Max, a 2.4 trillion-parameter model.

Why It Matters?

The market’s reflexive reaction to Chinese AI efficiency breakthroughs — sell semiconductors — is being tested against a more nuanced reality: models at the 2-3 trillion parameter scale are memory-bound, not compute-bound, meaning that efficiency gains at the compute layer do not reduce hardware spending if memory requirements grow proportionally. If Kimi K3 and Qwen 3.8-Max represent the direction of frontier AI development, the implication is sustained and possibly growing demand for high-bandwidth memory and memory-rich accelerators, even as inference compute per query becomes more efficient. The two Chinese releases also validate that the frontier AI race has genuinely gone global, adding competitive pressure on US labs’ pricing models while simultaneously expanding the total addressable market for AI infrastructure globally.

What’s Next?

Watch July 27, when Moonshot plans to release Kimi K3’s model weights publicly — open-sourcing a 2.8 trillion-parameter frontier model would be a significant event, allowing enterprises to run the model themselves but requiring substantial hardware investment to do so at scale; watch Alibaba’s Qwen 3.8-Max benchmark results when they are published, which will be the first independent validation of whether the model actually competes with GPT-4o and Claude at the frontier tier; watch SK Hynix and Nvidia’s order books and guidance commentary for any signal that the Kimi K3-style memory-heavy model architecture is translating into incremental hardware demand; and watch whether the semiconductor market corrects the Kimi K3 misread as the memory-requirement analysis becomes more widely understood among investors.

Source: Bloomberg

Previous Post

US Gasoline Prices Cross $4 a Gallon Again as Iran Conflict Hits Crude and Russian Refinery Outages Tighten Supply

Next Post

The Pentagon Is Starting to Buy From Defense-Tech Startups — But It’s Not Replacing the Old Guard

Recommended For You

Trump Says the US Could Let Nvidia Sell Blackwell Chips to China in ‘Two Years or Three Years,’ Calling the H200 ‘Semi-Obsolete’ as Congress Moves to Bar Sales for 18 Months

by Team Lumida
1 day ago

Trump said Nvidia could sell Blackwell chips to China in two or three years, as senators push the AI Overwatch Act to bar such sales for at least...

Read more

Intel Signs Its First Outside Customer for 18A Chips and Names Microsoft as Its First Public Customer for EMIB Packaging, After Losing More Than Half Its Value Since September

by Team Lumida
1 day ago
Intel Surges 5%, SK Hynix Up 3% on Reports of U.S. Memory Chip Manufacturing Deal to Serve Nvidia and AI Demand

Intel CEO Lip-Bu Tan said the company signed its first outside customer for 18A chips and named Microsoft as its first public EMIB packaging customer.

Read more

China’s Military Is ‘Actively Soliciting’ Nvidia’s Most Advanced Chips for AI-Enabled Military Capabilities, a Congressional Commission Says in Its 2026 Report

by Team Lumida
2 days ago

A congressional commission said China’s military is seeking advanced Nvidia chips and evading export controls, including via data centers in Malaysia.

Read more

Bipartisan Push to Attach AI Chip Export Limits to the December NDAA Gains Steam After the Huang Affair, With Schumer, Jeffries and Speaker Johnson Backing the AI Overwatch, MATCH and Chip Security Acts

by Team Lumida
2 days ago

Lawmakers from both parties are pushing three chip export bills into the December defense bill after Jensen Huang’s resignation, but time is short.

Read more

Banks Pause a $480 Million Nvidia GPU Loan to Aolani Cloud Tied to ByteDance After New US Licensing Rules for Malaysia and Singapore Data Centers

by Team Lumida
3 days ago
ByteDance Is Buying Custom Chips From Taiwan Because It Cannot Buy Nvidia’s, Which US Rules Permit

Banks paused or reconsidered a $480 million Aolani Cloud loan for Nvidia GPUs tied to ByteDance after new US licensing rules, Bloomberg reported.

Read more

Commerce and Justice Say They Began Investigating Nvidia Partner Megaspeed in 2024, as Critics Ask Why It Kept Getting Chips After Nvidia’s ‘No Sell’ Listing

by Team Lumida
3 days ago

Commerce and Justice said they opened export-control probes into Megaspeed in 2024 and inspected it in June 2025, as critics question how it kept buying chips.

Read more

Nvidia Warns Chip Buyers That the Remote Access Security Act Could Expose Them to US ‘Legal Action’ and Caps Sales to Clouds Renting Over 50% of GPUs to Chinese Firms

by Team Lumida
3 days ago

Nvidia cautioned customers that the proposed Remote Access Security Act could bring US legal action over Chinese cloud access to its chips, CNBC reported.

Read more

Anthropic’s IPO Could Hand 300-Plus Investors a Windfall, With a $1.33 Trillion Valuation About 38% Above Its $965 Billion May Round

by Team Lumida
4 days ago

Anthropic’s IPO, expected to value it at $1.33 trillion or more, is set to reward over 300 investors, from Spark and Menlo to Amazon and Alphabet.

Read more

Commerce Moves to Close the Offshore Data-Center Loophole for Chinese AI Firms and Requires Licenses for Chip Exports to Five Countries, Its First Policy Response to the Nvidia Scandal

by Team Lumida
4 days ago

Commerce issued a rule curbing Chinese firms’ offshore access to Nvidia compute and requiring chip-export licenses for five higher-risk countries.

Read more

Trump Administration Plans to Name Nvidia, OpenAI and More Than Two Dozen AI Firms ‘National Champions’ as the BAI AI ETF Sits 33% Below Its June Record

by Team Lumida
4 days ago
The AI Capex Machine Keeps Roaring — Nvidia’s Manufacturing Partner Hon Hai Crushes Estimates as Server Demand Drowns Out Recession Fears

The Trump administration plans to name more than two dozen AI firms, including Nvidia and OpenAI, US national champions, the WSJ reported.

Read more
Next Post
Pentagon Expands Chinese Military Company List, Adding Tech and Industrial Giants

The Pentagon Is Starting to Buy From Defense-Tech Startups — But It's Not Replacing the Old Guard

stacked gold bullion bars

Gold Is Down 22% Since the Iran War Started

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Related News

skyline photography of nuclear plant cooling tower blowing smokes under white and orange sky at daytime

Uranium Giants Unite: What Paladin’s Acquisition of Fission Means for Investors

June 24, 2024
Oracle’s Q4 earnings missed expectations but stock jumped ~11% after new cloud deals

Oracle Tries to Steady AI Data-Center Narrative as Blue Owl Drops Out of Michigan Financing

December 17, 2025
Family Offices Eye AI: A $600 Billion Opportunity

Family Offices Eye AI: A $600 Billion Opportunity

May 30, 2024

Subscribe to Lumida Ledger

Browse by Category

  • Lifestyle
    • Family Office
    • Health and Longevity
    • Legacy
    • Next Gen Wealth
    • Trust, Tax, and Estate
  • News
    • Alt Assets
    • Crypto
    • Equities
    • Latest
    • Macro
    • Markets
    • Real Estate
  • Opinions
    • Investing Philosophy
    • Op-Ed
  • Research
    • Trackers
  • Themes
    • Aging & Longevity
    • AI
    • Biotech
    • CRE
    • Cybersecurity
    • Digital Assets
    • Legacy Brands
    • Nuclear Renaissance
    • Private Credit
    • Software
Facebook Twitter Instagram Youtube TikTok LinkedIn
Lumida News

Premium insights to help you invest beyond the ordinary. Lumida Wealth Management LLC (‘Lumida”) is an SEC registered investment adviser

CATEGORIES

  • Aging & Longevity
  • AI
  • Alt Assets
  • Biotech
  • CRE
  • Crypto
  • Cybersecurity
  • Digital Assets
  • Equities
  • Family Office
  • Health and Longevity
  • Investing Philosophy
  • Latest
  • Legacy
  • Legacy Brands
  • Lifestyle
  • Macro
  • Markets
  • News
  • Next Gen Wealth
  • Nuclear Renaissance
  • Op-Ed
  • Private Credit
  • Real Estate
  • Software
  • Themes
  • Trackers
  • Trust, Tax, and Estate

BROWSE BY TAG

AI AI chips Amazon Apple Artificial Intelligence Banking Bank of America China Commercial Real Estate CPI data centers Donald Trump EARNINGS ELON MUSK Federal Reserve financial services generative AI Goldman Sachs Google India Inflation Interest Rates Investment Strategy Japan Jerome Powell JPMorgan Markets Meta Microsoft Nasdaq Novo Nordisk Nuclear Energy Nvidia oil OpenAI private equity S&P 500 SEC Semiconductor stock market stock surge Tech Stocks tesla Trump Whale Watch

© 2025 Lumida Wealth Management LLC is an SEC registered investment adviser. Privacy Policy. Cookies Policy.
Disclaimer Important Information This site is for informational purposes only. Information presented on this site does not constitute as investment advice.

Lumida Wealth Management LLC (‘Lumida”) is an SEC registered investment adviser. SEC registration does not constitute an endorsement of the firm by the Commission nor does it indicate that the adviser has attained a particular level of skill or ability.

Lumida's website (referred to herein as the "Website") is limited to the dissemination of general information pertaining to its advisory services, together with access to additional investment-related information, publications, and links. Accordingly, the publication of the Website on the Internet should not be construed by any client and/or prospective client Lumida’s solicitation to effect, or attempt to effect transactions in securities, or the rendering of personalized investment advice for compensation, over the Internet.

Any subsequent, direct communication by Lumida with a prospective client will be conducted by a representative that is either registered or qualifies for an exemption or exclusion from registration in the state where the prospective client resides.

‍Lead Capture Forms: By submitting your contact information in the forms on this site, you are not obligated to invest in Lumida's product or services.
‍Address: Lumida Wealth Management, 25 W 39th Street Suite 700, New York, NY 10018

No Result
View All Result
  • Home
  • Earnings
  • News
    • Alt Assets
    • Crypto
    • Equities
    • Macro
    • Markets
    • Real Estate
  • Lifestyle
    • Family Office
    • Health and Longevity
  • Themes
    • Aging & Longevity
    • AI
    • CRE
    • Digital Assets
    • Legacy Brands
    • Nuclear Renaissance
    • Private Credit
  • About Us

© 2025 Lumida Wealth Management LLC is an SEC registered investment adviser. Privacy Policy. Cookies Policy.
Disclaimer Important Information This site is for informational purposes only. Information presented on this site does not constitute as investment advice.

Lumida Wealth Management LLC (‘Lumida”) is an SEC registered investment adviser. SEC registration does not constitute an endorsement of the firm by the Commission nor does it indicate that the adviser has attained a particular level of skill or ability.

Lumida's website (referred to herein as the "Website") is limited to the dissemination of general information pertaining to its advisory services, together with access to additional investment-related information, publications, and links. Accordingly, the publication of the Website on the Internet should not be construed by any client and/or prospective client Lumida’s solicitation to effect, or attempt to effect transactions in securities, or the rendering of personalized investment advice for compensation, over the Internet.

Any subsequent, direct communication by Lumida with a prospective client will be conducted by a representative that is either registered or qualifies for an exemption or exclusion from registration in the state where the prospective client resides.

‍Lead Capture Forms: By submitting your contact information in the forms on this site, you are not obligated to invest in Lumida's product or services.
‍Address: Lumida Wealth Management, 25 W 39th Street Suite 700, New York, NY 10018