Nasty out there. Chillin with a lot of cash and Vicor and other positions are all too small to bother me. Earnings next tuesday premarket for $VICR.
I think the details on that license will be telling. If its over $20m a quarter going forward, Vicor should fly IMHO. There's a chance a lot of the first payment is retroactive/catchup payments, but I'm leaning towards this one being much larger than the one in signed in May. That license was $5m a quarter in year 1 and $10m a quarter in year 2. This latest one caused them to increase guidance by 20% more than the original 10% they guided for on the last call, roughly $28m over the original 10% sequential increase.
Other details I want to hear about:
-Timing to get fab1 to 1.5b run rate, they said its still on track for roughly 800m run rate by year end (implying 66m in december)-but they are going to get it up to "at least 1.5b" by putting some of the simpler steps in an adjacent building and buying a new 3d interconnect line.
Whenever they use the term "at least," as evidenced by that first guidance update, you know theres room for upside-that upside could come from alternate sourcing some lower margin products and making room for more VPD at fab1, the operating leverage that would bring if thats how it goes is huge.
-Fab 2(and more), they bought an existing building and land for a lot more fabs. Curious if they think they can get that fab up and running by end of 27/early 28. Would be extremely bullish if so as that is when gen2 VPD ramps are happening.
-Gen2 VPD is now being sampled/qualified at select/key customers, commentary/feedback from those customers will be great to hear.
I do not think they'll comment much on the new lawsuits, other than maybe implying the latest license for VPD for a Leading AI OEM may have come from the strength of their case.
Macro is ugly, I'm not bullish on tech/semis just yet. I do think Vicor is on track to becoming a "mini ARM" for power delivery at the point of load. Willing to say in even if it drops to 160 again and just buy more assuming no bad news.
Show more
Israeli Finance Minister Bezalel Smotrich:
I'm putting an end to the idea of a Palestinian state by strengthening our presence—with the approval of the Americans, the Prime Minister, and Cabinet decisions.
We've established 103 communities, 160 farms, legalized about one million dunams of land, all with legal authority.
That's what drives people crazy—not the other things.
I'm investing billions, building roads, advancing land registration, and changing the reality on the ground.
Show more
Jefferies raises its target price on H&M to SEK 160 from SEK 154.
+ $nbis secured 3.5B from Citi @ SOFR + 2.75-3.50%
+ BofA maintains 310 $nbis price target
fundamentals only improving, despite the broader sell-off in the index
once volatility comes down and the war deescalates, $nbis will take back off to new highs
I’m hoping we can see a larger move below 160 with $vix well over 20. yields already rolling over, just need oil to stop and things will reverse drastically
Show more
Open-weight AI is 6–7 months behind the closed frontier but frankly your workload doesn't care. Did your problems get 6–7 months harder?
Epoch measured the capability gap specifically for cybersecurity.
🟣 GLM-5.3 → 160 Cyber ECI
🟦 GPT-5.4 from March → 157
🟦 today's closed frontier → ~171
So today's best open-weight cyber model is roughly where the closed frontier was 6–7 months ago.
That sounds like bad news for Local AI.
I'm not sure it is.
Because there's a more fundamental question - Did your problems get 6–7 months harder?
If yesterday's frontier model can already
💻 build the app you need
📄 search your documents
🔧 fix your code
📊 analyze your spreadsheet
🤖 automate your workflow
then today's smarter frontier model doesn't suddenly make yesterday's model incapable.
This is the part of AI benchmarks I think we're beginning to miss.
The frontier keeps moving.
But the capability threshold required to solve problems nearly all of us encounter often doesn't.
Once a local model crosses that threshold, the competition changes.
Most people don't actually want to code.
They want the thing the code produces.
If a local model can write the tool, run it and solve the problem, it doesn't necessarily matter that Opus can solve a substantially harder problem.
And today's other OpenRouter number makes this point 👉 Open weights → 72% of tokens / 27% of spend.
⚠️ Epoch measured cybersecurity capability, not general everyday usefulness, and GLM-5.3's Cyber ECI currently comes from only three benchmark results. Epoch AI
Link in ALT
Show more
Five years ago today, an ISIS-K suicide bomber attacked Abbey Gate during the evacuation from Kabul, murdering 13 American service members and approximately 160 Afghan civilians.
We remember our fallen 13:
Staff Sgt. Darin T. Hoover, USMC
Sgt. Johanny Rosario Pichardo, USMC
Sgt. Nicole L. Gee, USMC
Cpl. Hunter Lopez, USMC
Cpl. Daegan W. Page, USMC
Cpl. Humberto A. Sanchez, USMC
Lance Cpl. David L. Espinoza, USMC
Lance Cpl. Jared M. Schmitz, USMC
Lance Cpl. Rylee J. McCollum, USMC
Lance Cpl. Dylan R. Merola, USMC
Lance Cpl. Kareem M. Nikoui, USMC
Hospital Corpsman 3rd Class Maxton W. Soviak, USN
Staff Sgt. Ryan C. Knauss, U.S. Army
America must always remember the names of these heroes. But remembrance also demands accountability.
Last year, under President Trump’s leadership, this FBI—working alongside DOJ, CIA, and our international partners—apprehended and brought back Mohammad Sharifullah, an ISIS-K operative and key player in the attack. A federal jury has since convicted him of participating in a nine-year conspiracy to provide material support to ISIS-K.
Five years later, our promise remains unchanged:
We will remember the fallen.
We will stand with their families.
We will hunt those responsible—no matter where they hide or how long it takes.
-DKP 🇺🇸
Show more
Japan has spent more defending its currency this year than at any point in its history.
It just lost the level again.
USDJPY moved nearly 1% in minutes during US hours today. From above 160 down to the 158.70s.
Lowest since August 24. Currencies that size don’t do that on their own…
If it was Japan, that’s the Ministry of Finance, not the BOJ. And we won’t know until month-end data.
They’ve already tried twice this year.
¥11.73 trillion across April and May. Roughly $73 billion, defending this exact level. Nearly DOUBLE the largest intervention in Japanese history.
Then a joint operation with the US in late July.
The yen has given back more than half those gains since.
The problem is the rate gap. Japan is at 1%. The Fed is at 3.50-3.75% with a 70% chance of a hike this month.
You can’t intervene your way out of that. You can only slow it down.
Which is why Bessent has been pushing Ueda to hike instead.
BOJ meets September 17-18. That’s the one that actually matters.
If you want to know where I’m deploying capital, turn notifications on. A lot of people will regret not doing it.
Show more