Investment Research + Management @kryptonresearch
Joined July 2021
- Tweets10.6K
- Following362
- Followers1.1K
- Likes6K
Pinned Tweet
Krypton Research is a completely different investment firm.
I believe every investor needs thinking like this.
There are thousands of finance substack posting the same “deep dives” or “value picks”.
We take a genuinely differentiated approach and it pays off. $SPY
For ~$1 per day, Krypton Research delivers this to thousands of readers:
- Investment research rooted in abstraction, thinking about the future and finding investments that fit our outlooks… not just ones with pretty 10-Ks.
- Weekly market commentary on subjects that are deeper than the regular headlines or recaps posted by most accounts.
- Monthly “Abstractions” diving deep on topics that are usually complex and sometimes unrelated to direct investments, but mold how you think about the world… it’s all connected.
All of that and more in a subscription that’s paid off multiple times over by our research and investment picks.
Join today: kryptonresearch.co/portfolio
1
11
2,833
$MOD getting acquired for 2024 prices.
Great timing by Samsung to say the least.
Very important readthroughs across the data center space with this deal though… pay attention.
1
10
1,104
$CSCO With yet another massive quarter.
Networking is a "supercycle in its early stages" said the CEO.
Our top networking pick is a sleepy $4B company that is sitting squarely in the hottest upcycle trend in the world right now with industrial networks.
100%+ upside... yet no one wants to talk about it:
The next major theme: Physical AI
kryptonresearch.substack.com…
9 months ago we identified the CPU trade that is being discussed today and sending shares of $AMD $INTC and more skywards.
The “Agentic Era” underpinned that trade and we slammed the table on $ARM at $110/share.
Arm went +300% from there, now the public is on it.
The next era will be the “Physical AI” one.
What will be the CPU of that theme?
Simple.
Networks.
Every AI model and agent that currently lives in data centers and servers will need networks to travel over in order to interact with and manipulate the physical environment.
Krypton Research is rooted in deep abstraction, and we are early to this setup as usual.
Our top pick isn’t the consensus ideas like Cisco, Ciena or HPE.
Instead, we chose a sleepy $4B company that is transforming itself into a leader in the space, stacking contracts with hyperscalers already, and simply overlooked by the market today.
130%+ upside.
kryptonresearch.substack.com…
6
440
Hey look! Another $SMCI probe…
Sick stock, don’t get how anyone can own it, have to completely disregard the great financials here given the endless parade of investigations and scandals.
Despite the $NVDA fallout, the show will go on.
Our strategy at Krypton is always looking ahead in the investment cycle, and we were already bullish on the next generation of semiconductors and new architectures.
Huang out only accelerated that thesis:
$NVDA and $AVGO earnings have made it clear that compute demand is only growing.
However, it’s the nature of that demand that will change in the coming years.
AI will need to enter the physical world, and inference/physical AI workloads will require different chip designs.
Ones that can beat both the power wall and memory wall.
We made 150%+ investing in beating the power wall through $ARM.
Now our newest chips investment is on a company looking to break the most recent bottleneck in compute… memory.
The next semiconductor giant: kryptonresearch.substack.com…
4
222
In all seriousness this is bad news for Nvidia, real chance Jensen is forced out now.
Don’t think it’s a major long term hit, unless the China investigations turn anything up, but short term very negative.
$NVDA
7
497
Going forward in the AI trade it will become increasingly important to be invested in “best in breed” companies.
The early cycle of anything with any connection working won’t work.
Folks will call it a trash market, but all you need to do is some research before buying…
4
291
If the market likes $CRWV here after its earnings this much, just think about the value in $ORCL at these prices…
Full length investment research memo:
Our deepest research piece… and arguably the most consequential is now published. $ORCL
We take a look at Oracle, the company at the center of the largest infrastructure buildout in human history, its financials, and its potential as an investment:
kryptonresearch.substack.com…
4
23
2,091
Nvidia $NVDA from 4 months remains the clearest view on the company. Said I would not have been a buyer then, and not now.
The Jensen thing is not a huge deal, and likely blows over in coming weeks, but the structural thesis for Nvidia stock just isn't there anymore:
$NVDA Has been in the penalty box for over a year now, and while many will point to low valuation multiples and strong growth, its once again why financials don't always tell the full story and why abstraction matters.
The chips space is becoming more fragmented by the day, with custom silicon being designed by almost every hyperscaler and frontier lab at this point.
Markets are looking ahead towards this, but more importantly, to the rise of inference compute rather than what has so far been intensive model training compute.
Once we reach sufficient training on frontier models, not even AGI, two things will happen:
1. Recursive training: Models can increasingly train themselves, requiring less time and less overall compute.
2. Agentic/Physical/Edge AI emerges: Use cases will be specific and require much more CPU usage (hence $AMD $ARM and even $AMBA etc... rallying) that can tackle the main constraints these areas face:
Power and Memory.
We bought $ARM at $110 to address the power issue, and now we are adding heavily to the semi that is addressing the memory constrained (full thesis linked below).
As for Nvidia, this means that investors are not quick to reprice the stock higher as future earnings come into question as their customers are faced with more compute options (in house and otherwise) and a new use case.
Good to own, not sure if I'd be a buyer here though.
4
316
Starting to get the vibe these Iranian "leaders", of which there is a new one every week, are not piecing together that Qatari and Pakistani mediators doesn't mean they're negotiating with Qatar or Pakistan...
1
4
229
Hard to see $NVDA making new all time highs any time soon, maybe at all in this cycle, but this $165 spot is extremely important.
2
6
303
This current CPI report should be discounted given what has happened to oil prices in the last few days/weeks.
A resolution in Iran will only serve to send energy prices careening lower.
2
6
757
$MIR been a rockstar position for us as of late, and our full thesis is now public +30% after our original published piece.
Worth reading in entirety, especially with $VST and $CEG validating the nuclear thesis in recent days with strong business outlooks.
Free read:
Another Krypton Research thesis is now public for all readers!
We invested in Mirion Technologies $MIR less than 3 months ago as a hidden Nuclear energy gem.
Today that position has crested 30% in returns and the thesis is now public:
kryptonresearch.substack.com…
4
373