Ming-yen Ho (@homervyn7) profile picture

IO & political economy of high tech manufacturing. Semiconductors/industrial policy/East Asia. PhD candidate @BerkeleyHaas, on the job market.

Berkeley, CA
Joined July 2016
Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)Photo by Ming-yen Ho (@homervyn7)
Ming-yen Ho retweeted
Datasection (3905.T) just cut FY revenue guidance by 89%, from ¥162B to ¥17B, and the stock went limit-down. This is the "Japanese neocloud" that the FT reported is effectively a Tencent datacenter: Chinese offtake, Nvidia Blackwell GPUs, Japanese flag. (1/8)🧵
18
59
45
356
115,915
Ming-yen Ho retweeted
China’s technology ministry (MIIT) approved its first quality-control management system for semiconductor equipment and components, saying ISO 9001 is not strict enough, media report. (ISO 9001 is from the Swiss-based International Organization for Standardization, an NGO made up of the standards setting bodies of around 170 nations.) #Semiconductors news.cn/20261117/2fb762588ec…
2
3
18
5,290
Ming-yen Ho retweeted
New US export rules end fast tracking for chip foundry and approved-chip packagers (OSAT) verification of advanced AI chips when the chip designer is based in a controlled country or in Singapore, Malaysia, Indonesia, Thailand or Brazil, starting Jan 1, 2027, which will impact six of the 24 approved OSATs, media report, Korea’s Samsung, Doosan Tesna, LB Semicon, Nepes, SFA Semicon and Steco. #semiconductors etnews.com/20261117000325
1
8
1
48
6,759
Ming-yen Ho retweeted
Over the past week, I’ve been overwhelmed by the love and support of friends, industry colleagues, and people I've never met. Thank you to everyone who has shown kindness to my family and me. It means the world to me. I came to the United States 53 years ago, and becoming an American citizen is one of the greatest gifts and proudest moments of my life. NVIDIA has been my life’s work. Building it into the company it is today alongside my colleagues has been a privilege and an honor. Nowhere but in America could a company like NVIDIA be possible. I love our country, and I love NVIDIA. I did not live up to the standards expected of everyone at NVIDIA, and I have made serious mistakes in my personal life. I created problems for the Board that I deeply regret. I own my mistakes completely, and I am committed to cooperating with the Board’s independent investigation. Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO. Thank you again to everyone who has shared their good wishes. I am profoundly humbled by your support. To my family: I love you, and I will spend the rest of my life being worthy of yours. Jensen
58,732
119,573
24,171
1,143,085
54,161,020
Ming-yen Ho retweeted
A while back, I was told Ascend-950 series was really good, better than hopper. Here is some China Mobile result measuring 950DT vs 910 series & H200 for inference on GLM-5.2. Result showed 950DT was 11x 910B & 3x 910C in tps. About 14% lower than H200. So despite the lower TFLOPS & higher cost (1.3x of 910C), it actually far surpassed the latter in actual inference performance. This shows that a 64-card cabinet w/ high speed interconnect & larger + higher bandwidth HBM is far more costly & more performant than the compute unit performance. When factoring in rental cost, 950DT economically wins against H200. That's why Chinese AI labs want B300, not H200. Even 950DTs have more favorable economics than H200.
20
54
10
354
36,941
Jensen Huang was ousted as Nvidia CEO this week, in part due to his crusade in DC against export controls. In joint work with @LuenTse , we study how did U.S. industrial and geoeconomic policies impact firm lobbying.
1
5
10
656
Both P(lobby) and lobbying $$ | already lobbying before 2019) increased. CHIPS Act reward is individualistic: firms start lobbying to get a share of the pie, instead of freeriding to let big firms get a public benefit.
1
30
ASML and TSMC, the giants of the industry, started lobbying after 2019 possibly due to both Huawei export controls and CHIPS. Nvidia another entrant in 2022. Given recent events, their political presence in DC will not go away anytime soon. Paper: homervyn.github.io/files/Ho_….
45
Ming-yen Ho retweeted
🎓 Nina Nikiforova (@nikiforovannina) is on the job market! 🌾 Pol. Econ., IO JMP: 🗳️ What do populist voters want? ⚙️ Nonparametric identification & estimation of voter preferences. 🇩🇪 AfD vote: first anti-elite, now right-wing. 📌 Visited Berkeley 🔗 nikiforovanina.github.io
🎓Meet our @unibocconi econ job market candidates! You know how good our master students are? Wait until you meet our PhDs! 🚀 This year's cohort is EX-CEP-TION-AL, and I mean it. 👉 All 10, from macro to behavioral, political economy, IO and trade: unibocconi.it/en/programs/ph…
19
6
105
36,545
Joint Statement from Department of Commerce and Department of Justice on Megaspeed Investigation.
893
1,678
806
7,308
5,517,234
Ming-yen Ho retweeted
DeepSeek is now restricting users due to “insufficient compute capacity”. Prompting rationing over inference usage and tighter restrictions (input limit, conversation history) per Cailian Believe it or not, compute shortages are a bullish signal for AI infra demand
176
108
43
1,595
332,320
I'm on the academic job market! I study the IO and political economy of high-tech manufacturing. My JMP asks: in a dynamic industry where winners are hard to spot, can governments be good VCs? Should the investments be centralized, or delegated to local governments? 🧵
1
29
2
103
7,211
I model VC choice, firm growth and import competition. Firm–VC alumni ties and VCs' different choice sets with observed budgets identify the choices. 2023 domestic chip sales would have been 11.6% lower had private VCs made the government deals, 4.0% lower had Beijing made the local deals.
1
1
184
Comments welcome! Paper: homervyn.github.io/files/Ho_…. My interest in China's chip policy began when TSMC managers told me the officials who helped build their Nanjing fab were soon promoted. Now that mature fab needs yearly US licenses to import chipmaking tools. Next on my agenda: the equilibrium effects of export controls.
1
3
209