Palo Alto Networks to Acquire Applied Compute for $4.5B
investors.paloaltonetworks.comIt seems like PANW has cloud and enterprise infrastructure ambitions. It is no longer a cybersecurity company.
They still are. They are trying to maintain their position as a premium cyber company that can deal with emerging threats.
In particular, they were converting licenses over to what they called Wildfire AI - https://docs.paloaltonetworks.com/advanced-wildfire - as a value-add sub on top of their hardware. I don't know anything about Applied Compute but I don't think they're trying to run commodity inference for people. It's probably going to be edge models and more analytic subscriptions.
VC at it's best is incredible.
What does this mean?
Perhaps it means Variable Change?
When you agree to a price, you also agree to a change amount. The final amount paid gets the headline, but the true price and the change given back is kept on the down low. Where the change ends up going can also be disguised. It's like getting a massive kickback.
It's sort of like buying your own books to get on a best-seller list. The true sales amount is hidden, and the best-seller status garners even more sales.
I meant Venture Capital! Investors will be sharing a massive ~$1.85b of proceeds on $510m invested. Benchmark for instance got in at the seed and will return 33x the invested money. This is impressive for a startup under 2 years.
Right - and in 2 years of runway, that's unprecedented - comparable to Meta + Manus.
Variable Change itself would be difficult if a fund's LPs raise ire over their missing money. Not saying it's impossible, and the "intention to deceive" would be a worrisome component, yet Gemini says Variable Change is a "sharp analogy, and it actually mirrors several real-world mechanics in how private valuations and venture rounds are engineered." https://share.gemini.google/Eof8MCVp1svt
Wow. $4.5B to serve fine-tuned open-weight Chinese models. This is a ~2-year-old startup with < 100 employees.
At its peak, pets.com was valued at nearly a billion in 2026 dollars, selling dog food online.
The pets.com thing doesn’t seem that crazy looking back. Currently chewy.com has an 8B market cap. The concept is clearly valuable they just weren’t able to execute on it.
True. I went through their offering and it is about providing forward-deployed engineering teams and a full-stack reinforcement learning (RL) platform for fine-tuned models. What if Chinese companies stop or slow down releasing open-weight models? I do not see American companies doing it regularly, i.e., Meta and OpenAI. There is so much reliance on companies investing in Chinese open models and enterprises willing to deploy them on their on-prem VPCs.
Note: I am questioning the business value and not the talent.
I think the endgame of the open weight models is to slow down or stop releasing them and go to a licensing type of business.
Even then there's still value to be had if all parties behave and prices and performance are competitive.
I read as that's what pets.com seeked to do and got super confused before realizing you were talking about the subject of this thread
Same. Also, there are U.S. open weight model providers like Beam, Thinking Machines, NVIDIA, etc.
Do the already released ones just stop working?
The question is: can you repeat this and capture their market easily yourself? The former maybe easier than the latter...
All 3 founders went to Stanford and worked at OpenAI. It's a big club and you ain't in it.
My cousin went to Stanford, got a job at google (pre-IPO) and hasn't had to work a regular job for more than a decade. I went to a state university, joined a few now-failed startups and worked for Wikipedia for 7 years. Now I'm living on my 401k, essentially unemployable.
And some Saudi Prince was born a hundred thousand times richer than you or your cousin will ever be. We play the hand's we’re dealt and presumably try to leave the world a little better than when we found it. At least we used to try and do that. Not anymore.
My friend and I played a round of golf with a random high school senior last week who spoke about how he was going to go to X college and get Y job to make Z dollars. For the first time in my life I truly felt old because I gave him the speech about how going to college in the US is more about where you go and who you meet than what you major in and your grades (as far as financial success in certain fields is concerned). Nobody explained this to me when I was younger. I heard a lot of "it's just a piece of paper you have to get out of the way" or "they're the best years of your life" or some variation thereof but never "you're going to get passed on for a job you're perfect for because you didn't go to Stanford/have a buddy/know a guy who knows a guy". I don't lose sleep over any of this because I couldn't have afforded to attend a name-brand school but I do wish I'd played things slightly differently.
As an aside: if you worked for Wikipedia I'd imagine that you're eminently employable.
Wikipedia seems like a pretty solid job for the resume
Don't let the resentment win
I have no resentment. I chose my path and I'm pretty happy with my life. I was just kind of agreeing that going to Stanford might affect one's financial prospects.
Access is a big part of what makes people successful, i don't think it's up for debate. Same person with talent can be born in a poor 3rd world country to a farmer and his or her odds of making it out will be absymbal. Even being born in America and having access to SV is a privileged position so should i feel bad that i have access to American tech ecosystem? My point is it's pointless to engage in this sort of exercise - Stanford grads do get better opportunities and that's not up for debate but a bunch of anthropic founders didn't go to stanford and still ended up doing well. If you can found a company with a rare set of skills that makes money, i'm sure you'll find funding even if you went to podunk university but if you want to use status and prestige to cash out and rip people off than yea going to stanford or harvard will give you a huge leg up.
Look up at the millionaire employees on paper from Anthropic.
"The Best AI is Built, Not Bought." - Applied Compute
Respect to those kids for kicking ass though.