Bernard Moon (@bernardmoon) header
Bernard Moon (@bernardmoon) profile picture
Silicon Valley
Joined April 2007
Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)Photo by Bernard Moon (@bernardmoon)
My interview on Bloomberg TV @technology discussing South Korea's #tech / #startup landscape on April 12, 2017. Starts from 33:18 to 39:18. youtube.com/watch?time_conti… #VentureCapital #Startups
44
50
3
365
Bernard Moon retweeted
If you're reading this, please help spread the word by sharing this article. Nearly six years after my daughter Hanako’s life was taken, I continue to seek justice.   Next Tuesday, November 17, may be the final court hearing in this case.  The Court will decide whether to accept the proposed plea agreement.  If it is rejected, Troy McAlister’s trial is scheduled to begin on Friday, November 20.   I do not believe this plea agreement serves justice, accountability, transparency, or the public interest.  I believe justice requires that the facts and evidence be heard in open court and the case be decided by a jury.   I would be deeply grateful for your support at Tuesday’s public hearing:   Public Hearing on 11/17/2026 at 1:30pm, Pacific Standard Time San Francisco Superior Court, Department 26 sftc-org.zoomgov.com/j/16069… Meeting ID: 160 6991 3087 Password: 443439   For more information about this case, please see Susan Dyer Reynolds’ reporting at thevoicesf.org/, and search “Hanako.”   I am deeply thankful for your support over these many years, and for continuing to stand with me for Hanako, Elizabeth Platt, and for justice.   #JusticeForHanako #SanFrancisco #Justice #USA @TheVOSF @SusanDReynolds
101
536
38
3,201
154,917
“Physics is the law. Everything else is a recommendation.” @elonmusk told me that at DealBook in 2023. Today, SpaceX caught Starship after 14 orbits of Earth. Plenty doubted it was possible. Just extraordinary. @SpaceX.
431
1,931
189
16,664
1,014,300
Sometimes I crack myself up when I write, especially these rapid fire takes since I set a guideline not to spend more than 15 mins in writing while prepping these newsletters. The SparkLabs Signal AI signal from the frontier markets to Silicon Valley sparklabsai.substack.com/p/t… Top Conspiracy Theories on Jensen Huang’s Ouster U.S. CONGRESSIONAL LEADERS: "EXPORT CONTROL THIS MOTHER F*@$er!” A couple years ago, Jensen was supposedly heard saying this to a toilet in the Senate restrooms while prepping for his meeting with various U.S. Senators and Congresspeople, but in the various voices from The Usual Suspects. Often repeating this phrase in Fenster’s (Benicio del Toro), McManus’ (Stephen Baldwin), and Verbal’s (Kevin Spacey) voices. The outcome of those closed door meetings was a pissed off establishment of D.C. and two years of planning a perfectly executed scheme. The planting of Alice Huang, who is really Cindy Wu from Lincoln, Nebraska, was easy since she was a sleeper agent developed a decade before by the CIA. LORI HUANG: 40 YEARS, 40 YEARS... HE TOOK YOU FOR 40 YEARS Once Jensen’s wife heard about the rumors of Alice Huang, it was downward spiral. The wonderful wife who supported Jensen during Nvidia’s startup days, raising two wonderful children, and supporting him during the ups and downs of building a trillion dollar became laser focused on revenge. Hiring the private investigators that tracked Jensen & Alice and Jensen & Kimberly with hundreds of photos and documents, which was eventually leaked to the press. CHINA'S POLITIBURO: KILL AMERICA'S LEADERSHIP IN AI Tired of their AI labs using model distillation from OpenAI and Anthropic and with these companies clamping down on this practice, they knew there would soon be an increased separation of performance between the U.S. models and Chinese. SO how best to stop the U.S. leadership in AI but to cut the head of the dragon, Jensen Huang! He powers them, he funds, and he drinks tea with them. So he simply had to be shutdown and outed by their top super spy, Alice Huang. #substack #AI #newsletter
8
15
298
Bernard Moon retweeted
This letter was sent to Nvidia's board today. A first for @atreidesmgmt Absolutely share @altcap belief that this letter outlines minimum acceptable standards.
606
1,542
209
13,132
1,910,211
🎙️On the Record #17 | New York-based VC MetaProp eyes more investments in Asia after Japan’s Toggle IPO signalsventures.substack.com…
6
248
Sad about Jensen. Shakespearean. But as de Gaulle said, “The graveyards are filled with indispensable men.”
301
1,811
211
18,237
1,333,502
Sorry this graphic by Gemini was so good, I had to post another newsletter promo... possibly my last one 🙄 I had too much fun asking Gemini to replace Dr. Evil & Mini-Me w Nvidia chips in the original poster 😂 Anthropic’s IPO Roadshow! And Jensen… Before we share about Anthropic’s roadshow, we have to share TBPN’s latest on this situation. Their listed key accusations word for word... sparklabsai.substack.com/p/t… #substack #AI #newsletter #venturecapital
🤖 Made with AI
10
1
15
299
Bernard Moon retweeted
Congratulations to our friends at @googlecloud and @AminVahdat! We’re proud to partner with you, with @AMD EPYC serving as the host processor for TPU 8tx.
Thrilled to share a big milestone from our AI & Infrastructure team at @Google. Today we’re expanding our 8th generation TPU portfolio with: • TPU 8i Max: Built for massive context windows and the highest compute density per rack • TPU 8tx: Purpose-built for secure, confidential inference This marks the first time we’ve introduced four distinct TPU systems in a single generation, in one year. For developers, hardware is only as good as the software stack. And we want you to be able to build with TPUs as easily and efficiently as we do at Google. You can now: ↳ Run native PyTorch & Hugging Face models via TorchTPU (zero code rewrites) ↳ Cut KV transfer latency >90% with open-source TPU Sync ↳ Eliminate idle compute waste with Agent Compute Engine Read more about our newest TPUs on our blog: bit.ly/3VfETPR I look forward to connecting with builders, researchers, engineers and organizations pushing the boundaries of AI and computing!
79
148
22
2,161
123,414
Bernard Moon retweeted
Larry Page has made many accurate predictions and this is one of them
Larry Page reportedly has a number for how much computing will end up in space: five nines. As in 99.999%. @patrick_oshag heard it secondhand: "Someone asked Larry Page how much of compute will be orbital, and he said five-nines." @martin_casado's reason it might happen: "It's just basically regulatory." He thinks there are 500 proposed bills to stop data center building in the US, Texas included. From @InvestLikeBest.
1,598
2,925
210
24,726
3,752,961
Excellent insights into the rapidly evolving AI landscape. Well worth subscribing!
Launch of our new AI newsletter! 2nd Post 🔥 Check it out once, read it twice and then SUBSCRIBE! (free is the same as paid, so insider tip is don't pay us) 👍 The SparkLabs Signal sparklabsai.substack.com/p/t… 600,000 Nvidia Stockpile? The All Chips Post WHERE IS JIMMY CONWAY, HENRY HILL and THE OTHER GOODFELLAS? China's Compute Squeeze, from Three Directions at Once Reuters reported that about 600,000 Nvidia H200 chips are stockpiled in Hong Kong because Beijing has withheld approval for large-scale imports — small batches have entered, with several major internet firms getting ~10,000 each, but no large-volume approval — while an eight-card H200 server now sells in mainland China for over 6 million yuan (~$900,000), about 3× its price a year ago. If this was NYC, the GoodFellas would have rob this stockpile last week… no offense meant to the Triads. The same day, CNBC reported Moonshot AI has discussed renting Blackwell clusters of 8,000+ GPUs from European neoclouds, since Blackwell falls under US export controls and current law gives regulators little authority over overseas cloud access. And Bloomberg reported Nov 4 that Chinese buyers who ordered ~6 million RTX Pro 5500 cards expect to receive only tens of thousands — roughly 1% — over the next four months, as Nvidia and TSMC allocate capacity to next-gen products. Why it matters: The constraint on Chinese AI is no longer just Washington's rules — Beijing is now the gatekeeper on H200 imports, and the workaround (renting abroad) is already drawing legislative attention via the Remote Access Security Act. Nvidia's own forecast assumes zero China data-center revenue; it reports Q3 results Nov 17. (source: Reuters, Nov 5, 2026 & CNBC/AI Stock Wire, Nov 5, 2026) There is plenty of money chasing AI capacity. The harder question is how quickly that capacity can actually be built, delivered and used... (rest is here, subscribe!) sparklabsai.substack.com/p/t… #substack #newsletter #AI #technews #startups #venturecapital #founders
2
2
606
Playing pickup basketball from my 20s to 50s, I encountered the same archetype player when I play post defense against some buff low-post players (18+ inch biceps) who are whiny 🤔 20s I was 185 lbs, 6'1" but could squat 500 lbs. All core and legs no upper body (bench max 175) so arms like twigs. When I would push these guys out of the paint or now just hold them, they all similarly complain that I'm using my arms or to stop hooking them on D, which I wasn't doing. "You think my twigs could hook your frame?" and I would laugh. 40s I was a fat 240 (3 kids) and now 50s at 220 lbs, fat, slow and squat 400, but these types say the same thing 😐 Odd discovery 🙄🫤🤔
2
9
590
Recent investment from our SparkLabs DeepTech Fund (sparklabsdeeptech.co.kr/en) "Welding robot startup albeeware raises $660,000 seed round from SparkLabs, SOSV" app.dealroom.co/news/note/we… #robots #AI #physicalAI #startups #venturecapital
3
8
242
Engineers are preparing for the arrival of quantum computers that crack online security. But with AI’s math breakthroughs, rampant code-breaking could come far sooner, writes Christopher @Mims wsj.com/tech/ai/ai-math-onli… via @WSJ
1
1
5
202
Launch of our new AI newsletter! 2nd Post 🔥 Check it out once, read it twice and then SUBSCRIBE! (free is the same as paid, so insider tip is don't pay us) 👍 The SparkLabs Signal sparklabsai.substack.com/p/t… 600,000 Nvidia Stockpile? The All Chips Post WHERE IS JIMMY CONWAY, HENRY HILL and THE OTHER GOODFELLAS? China's Compute Squeeze, from Three Directions at Once Reuters reported that about 600,000 Nvidia H200 chips are stockpiled in Hong Kong because Beijing has withheld approval for large-scale imports — small batches have entered, with several major internet firms getting ~10,000 each, but no large-volume approval — while an eight-card H200 server now sells in mainland China for over 6 million yuan (~$900,000), about 3× its price a year ago. If this was NYC, the GoodFellas would have rob this stockpile last week… no offense meant to the Triads. The same day, CNBC reported Moonshot AI has discussed renting Blackwell clusters of 8,000+ GPUs from European neoclouds, since Blackwell falls under US export controls and current law gives regulators little authority over overseas cloud access. And Bloomberg reported Nov 4 that Chinese buyers who ordered ~6 million RTX Pro 5500 cards expect to receive only tens of thousands — roughly 1% — over the next four months, as Nvidia and TSMC allocate capacity to next-gen products. Why it matters: The constraint on Chinese AI is no longer just Washington's rules — Beijing is now the gatekeeper on H200 imports, and the workaround (renting abroad) is already drawing legislative attention via the Remote Access Security Act. Nvidia's own forecast assumes zero China data-center revenue; it reports Q3 results Nov 17. (source: Reuters, Nov 5, 2026 & CNBC/AI Stock Wire, Nov 5, 2026) There is plenty of money chasing AI capacity. The harder question is how quickly that capacity can actually be built, delivered and used... (rest is here, subscribe!) sparklabsai.substack.com/p/t… #substack #newsletter #AI #technews #startups #venturecapital #founders
4
1
1
6
823