Watch the mitral valve carefully during systole.
What is the most likely diagnosis?
A. Mitral stenosis
B. Mitral valve prolapse
C. Mitral annular calcification
D. Flail mitral leaflet
I don’t have differentiated alpha on the memory bubble debate, but one important perspective:
Sell side reports projecting the memory supply/demand imbalance to peak in the next few quarters are number crunchers modeling today’s known AI use cases.
The problem is AI’s biggest use cases may not exist yet.
I’d rather bet on AI creating unforeseen demand than on a spreadsheet correctly predicting its known memory needs two years from now. $SNDK $MU
Everyone panic selling $MU at $812 while ignoring: $50B guided Q4 revenue. $31 EPS. 16 take-or-pay Strategic Customer Agreements just signed with Qualcomm, Visteon & HARMAN. Simply Wall St: 39% undervalued. Late Sept earnings. 3rd dip of 2026. Both prior ones were the trade, NFA.
At $822, you're at 5.5x forward earnings on a company guiding $50B revenue next quarter. The math hasn't changed. The fear has gotten louder. Same movie, third act. Not financial advice.
$MU This is the third -30%+ drawdown in this bull run. March: $471 → $321 (-32%). Now: $1,255 → $806 (-36%). Both prior drawdowns were the opportunity. The thesis - 16 SCAs, HBM4 sold out, $50B Q4 guide, September 22 earnings - is unchanged.
$MU - Updated target: $1,500-2,000. Next earnings September 22. If Q4 delivers $50B and $31 EPS and beats by 20% again, you're looking at a $140+ annualized EPS run rate entering FY2027. At that point $2,000/share isn't a bull case. It's arithmetic. Not financial advice.
Scientifically it shows that babies/children learn emotions like fear by watching how adults react in these situations. So message can be on these same lines.
SanDisk $SNDK CEO on signing extremely significant long-term memory commitments from data center customers:
"We have customers putting up billions of dollars in collateral through various financial instruments that will survive for the life of these contracts. If they don't meet their obligations for consistent purchasing every quarter, that financial commitment immediately comes to us as compensation.
Honestly, I never expect to collect those because our customers are extremely serious about needing this product. In the normal case, we sign an agreement and, within weeks, we’re having a conversation about how to increase the amount of product we can provide. It is a very dynamic market."
I've known Adam since early 2014, and without a doubt, this is the best public representation of how he thinks and operates.
Yes, I'm an $APP shareholder and a fan of Adam personally, but I think anyone running a business in today's tech landscape can learn a lot from him as a leader.
nice interview @HarryStebbings
https://t.co/g2H6zf7tlE
Actors are so cooked.
This is the most amazing AI I’ve seen yet and it’s even getting the emotions fairly accurate.
Great work from Taylor Chien. I might even go watch this movie.
$MU -4% $SNDK -6% on a research paper. TurboQuant targets inference KV cache only - irrelevant to HBM training demand. Morgan Stanley: "neutral to slightly positive" for memory hardware. Jevons Paradox: cheaper AI = more AI deployed = more chips needed. Buy the overreaction imo.
OUR NEW VIDEO MUST watch it over and over
Get smarter watch this over and over Those that are doing great have watch it all the time IT stunningly good
https://t.co/SIxHVokyGY