Already own a few.NVDA could be a buy real soon if not already
See above.Already own a few.
I’m curious on what you’re basing this recommendation. Care to elaborate?
Fungible refers to items that tend to be interchangeable and indentical in value. AI GPUs are not interchangeable, even within a given brand. They can also have specific uses.Complete engineering answer.
A commodity is simply something that is fungible. You don't need a more complex definition. If you don't know what fungible means look it up, but its essentially something that can be replaced by another item to do the same function.
So if I can run my Chat GPT model on NVDA chips in whomever datacenter, or I can run my Chat model on AWS/GOOG/Meta whomever chip - those data centers are fungible, and it doesn't seem to matter who's chips they have.
Market hasn't figured it out yet. Buyers of such services soon will.
Buy the rumor sell the news more than likely.Some PLTR boost
Supposed to be up after the great news.. Down over $10.... today
I think Burry soiled it and once investors leave it's hard to get them back...
If I can run my LLM on a chip from any of the above mentioned companies to the same affect - how is it not fungible?Fungible refers to items that tend to be interchangeable and indentical in value. AI GPUs are not interchangeable, even within a given brand. They can also have specific uses.
High-end AI chips are currently treated as scarce, differentiated products. It is fair to say the industry is moving towards commodity components and that there are commodity components. That's part of the product life cycle.
Performance can vary wildly from instant to unusable depending on requirements.If I can run my LLM on a chip from any of the above mentioned companies to the same affect - how is it not fungible?
They both will run the same LLM There not Asics. There may be differences but there nuanced. There may be differences in speed or energy use which can be solved by price.Performance can vary wildly from instant to unusable depending on requirements.
My point is, the technology continues to evolve and can be specialized. Fungible inplies interchangeable.
AI chips are not interchangeable commodities. There are differences in chip architecture, software capabilities and intended use, and performance.
Will there be a surprise ROI in the near enough future to stop the Doomsday clock?I don't think anyone here questions AI's relative importance going forward.
The question is - so far the chip makers and datacenter owners are the ones being bid up due to "AI". However if we can now see that Chips are a commodity making data centers a commodity, or at least a very hard to differentiate non commodity of some type, then are we bidding up the wrong thing? No one has figured out to make much money on any of this stuff. So far its just spend. Now even the great tech companies are borrowing money when they use to be filthy stinking cash rich.
Follow on question if you agree to above, who will be the Google / Meta / Apple of 2001 in comparison?
I am clearly just speculating out loud and no one knows the future. However if a FLOPs (measure of compute) is cheaper at one datacenter over the other, no one will care which chip. Hence the moat the chip makers themselves claimed to have doesn't really exist. Of course if a chip maker can make and sell one cheaper, then they win - but again there goes there margin.Will there be a surprise ROI in the near enough future to stop the Doomsday clock?
What is your guess to the last question?
My uneducated guess would be Amazon, based solely on their real world footprint size.
There are far fewer players than there were in 2001 to pick from.
Also if it happens what will the AI bust be called? Maybe AI knows the answer![]()
Long bondsI need something to buy, long term in the sense of months and depending on the ride years.
Still in Walmart for over 5 years and thank GOD I didnt sell AMZN which I bought a while back (year?) around 220
However the other 1/3+ I sold at 720+ and that was META. .I was thinking I would get back in at this price 600 but my mind is not feeling the love right now, which means I should most likely buy but my brain wont let me.
AmazingI am clearly just speculating out loud and no one knows the future. However if a FLOPs (measure of compute) is cheaper at one datacenter over the other, no one will care which chip. Hence the moat the chip makers themselves claimed to have doesn't really exist. Of course if a chip maker can make and sell one cheaper, then they win - but again there goes there margin.
So assuming this is potentially accurate, there are some huge differences to 2001. For one, the big datacenter players - AWS, MSFT, GOOG have massive positive cash flows from there other businesses. There is also still a shortage of data centers so there margins are still pretty good - for now. I suspect they will at some point simply stop spending the horrific amount of money there spending on soon to be obsolete chips / datacenters. This is my guess - why play in a commodity business. This might actually cause those stocks to rise - better earnings.
In this scenario the chip makers themselves have the problem. Which again, the three above make chips as do others. Which to pick?
Comparing to 2001, the chip makers to me seem like the hard drive makers back then to me. Whomever could make the most storage for the least money was winning at that instant. Race to the bottom.
Obviously there is no way to time any of this - and they could all double from here. Which is why I consider myself a trader, not an investor. I am not married to any holding.
NVDA Already sold.Already own a few.
I’m curious on what you’re basing this recommendation. Care to elaborate?
Look at:I need something to buy, long term in the sense of months and depending on the ride years.
Still in Walmart for over 5 years and thank GOD I didnt sell AMZN which I bought a while back (year?) around 220
However the other 1/3+ I sold at 720+ and that was META. .I was thinking I would get back in at this price 600 but my mind is not feeling the love right now, which means I should most likely buy but my brain wont let me.
Good money if you're homelessNVDA Already sold.
Easy $2500.
****************************************************
Now I'm glad I sold SWKS yesterday!
Date
May-05-2026
Symbol SWKS
Symbol description
SKYWORKS SOLUTIONS INC
Type
Margin
Shares
-XX00.000
Price
$71.96
Today= >Skyworks Solutions $64.74
-7.82 (-10.78%)
++++++++++++++++++++++++++++++++++
Look at:
BINC
JAAA, PAAA, FAAA
BUXX, BILS, SGOV. GSST, etc those sort of ST bills funds
VNLA
For interest
______________________________________________________________________________
My loaned share money came in yesterday, over $700 for April. Easy money.
_____________________________________________________________________________
I bought FINS and exercised my DHY rights
Someone was looking for dividend stocks here not long ago. Paging whomever.In the last week, I bought GIS, KMB, CLX.