- > Then in maybe one of the best rug pulls of all time, in July they quietly changed their valuation to $500 million. A 75% cut in four months. I’ve never seen anything like that since the 2008 financial crisis.
Not sure where the author is getting their information from but there is seemingly little correlation between the investment rounds quoted in this post and other online sources. No mention for example of the Series E that valued Groq at $6.9bn.
- > About a year ago, Groq announced a $1.5 billion infrastructure investment deal with Saudi Arabia. They also secured a $750 million Series D funding round.... Then in maybe one of the best rug pulls of all time, in July they quietly changed their revenue projections to $500 million. A 75% cut in four months. I’ve never seen anything like that since the 2008 financial crisis.
Not following the core argument here. Author seems to be comparing valuation in funding rounds to revenue projections. Revenue projection was revised downward, valuation was not.
Good point about not running the proprietary models, but that doesn't preclude strategic fit with Nvidia.
- Ya well, startups are just low risk R&D facilities in service of big tech now https://centreforaileadership.org/resources/opinion_startups...
- Groq is not a publicly traded company and has no legal reporting requirements. Sure, their projected revenue numbers they gave to investors dropped from $2b in February to $500m in July, but a later funding round in September showed it wasn't significant to how insiders saw the company. Contrary to what this article would imply, their valuation more than doubled from $2.8b last year to $6.9b this year after Groq's latest round of investment in September (after their revenue adjustment). Considering they increased revenue from $90m to $500m and got a $1.5b commitment from Saudi Arabia, I really don't see this being 'hype'.
src: https://www.reuters.com/business/groq-more-than-doubles-valu...
- It's a shame. Groq was really great. Nvidia is stifling innovation here. I don't understand how market regulators allow this.
- > To visualize $1.5 billion: if you cashed that check out in $100 bills and stacked them one on top of another, it would reach a five story building. For ordinary plebeians like us, at the average US salary of around $75K, you’d need to work 20,000 years to earn that.
No, we don't need to visualize that.
- This article was good, but the blaming the us-east-1 outage on layoffs doesn't seem accurate.
- Since this article mentions SRAM, which not everyone knows a lot about, this could be helpful for anyone interested: https://people.freebsd.org/~lstewart/articles/cpumemory.pdf
- Why do you think? Not because of any output of the company, of course.
But because buying it helps perpetuate the hype and money cycle of the 'AI' trend for awhile longer. It may not look like it directly, but a purchase like this keeps Nvidia's stock up in the future, which is all investors care about.
- A) What’s 20b comparing to the extravagant current valuation of Nvidia at 4.64t?
https://claude.ai/public/artifacts/8c395eb5-8d22-431f-b6ba-0...
B) All info the OP(= author) knows is known to the professionals dealing with the due diligence. They decided to do so while looking at data which is not available the public. So assuming they know some things why we don’t know is not a far fetched idea.
- If your goal isn't four times greater than what you're likely to achieve, than you might as well not have goals.
— A motivational speaker, probably
Revenue targets are meaningless, especially in hyped fields.
- I often tell startups if you have amazing revenue, that your startup might get a lower acquisition offer price because your acquirer doesn’t care about your revenue. They might just care about your tech.
- You are missing the point that it is a strategic acquisition to kickstart a new vertical that they have struggled with: serving inference. They have tried to organically grow this and do weird things like inference within their other customers’ clouds.
It certainly isn’t a “panic” as nvidia is so flush with cash. This is a minuscule amount of money for them.
- Nothing against the author, the content is good. I just wish this wasn't fluffed out with AI.
- Chamath is involved, this is the only thing that you need to known. (it's probably a scamish company that will eventually belly up)
- Yes, this was a defensive move from Nvidia.
My understanding is Groq failed to deploy their second-gen chips on time, which caused their stock to deflate.
Groq's primary advantage over Cerebras and SambaNova, as I see it, is they don't fabricate on TSMC. That's attractive to Nvidia, who doesn't want to give up any of their datacenter GPU allocation.
- Kind of weird to base the bubble on how tall a stack of $100 bills is...
- What’s more concerning is the growing trend of big tech companies “acquiring” a startup’s leadership team and IP and screwing over all the employees holding equity.
- That probably explains why the Groq board accepted the deal.
- Killing your competition is priceless though
- It's a throwaway company used to tie up and defer account receivables for stock reporting, no?
- It’s amusing to see repeated “it’s a bubble” takes.
One of them will surely be right eventually!
- The bubble take is tired. This was regulatory arbitrage: IP licensing instead of acquisition to dodge CFIUS/antitrust. The $13B premium to avoid years of hold up while enriching Chamath and giving Trump's AI Czar a Christmas present. So many other things at play here than just "AI bubble so big it will boom".
Here's my take on what actually happened: https://ossa-ma.github.io/blog/groq
- In three years China will be making far cheaper clones of Nvidia chips
Current administration just handed it to them for a bribe
https://www.bbc.com/news/articles/ckg9q635q6po
They'll tear them down and x-ray/electron-microscope it and they've gotten exceeding great at cloning chips
- They understand the value of the tech
- And it was a no brainer
- > At that time, the company was valued at $2 billion. Hello bubble. Then in maybe one of the best rug pulls of all time, in July they quietly changed their revenue projections to $500 million. A 75% cut in four months.
I feel like I'm missing something here…
- The writing style here is so belittling, and frankly stupid.
E.g. "billion is so big!", uh, I've heard of a billion before, and then comparing the value of a company to a single person's salary, as if that was very relevant.
- Author seems to be conflating valuation and projected revenue
- It was probably less about revenue targets and more about pre-emptively removing potential future competition while it is still relatively cheap.
- This is missing the point. Even at a clipped revenue projection - $500m at $2b is a 4x multiple on revenue.
A 4x for an AI cloud+infra play that targets speed and cost? Where do I send the check?
If NVIDIA believes it can take this and scale it, $20b is a no brainer.
- The American Dream /s