SpaceX is arranging a $40 billion financing package through Apollo Global Management to secure AI chips while keeping the debt off its balance sheet.
$SPCX
50
Samsungβs HBM comeback :
Q3 HBM4 sales more than tripled quarter over quarter, and Samsung said HBM4 reached 60% of its HBM sales.
Now itβs reportedly developing custom HBM4E for Metaβs AI chips, potentially pairing Samsung memory with a 2nm base die.
Samsung also pulled forward its new HBM packaging plantβs planned start to June 2028.
$DRAM $MU
4
130
Yo @strictrope this your twin ??
109
LPS provides will be big winners in the future π―π― $IREN
Also borrowing this screenshot from @_Sgr_A_Star, and adding some color from last week's $IREN IR notes:
*Hyperscalers are now much more willing to prepay a significant percentage of the TCV compared to last year (e.g. $MSFT at 20%).*
Sweetwater 1 has 200MW IT under construction, 150~170MW of VR200s on order, and 9-12 months until the first data hall is commissioned.
The iron is hot, time for @IREN_Ltd to lock in an anchor tenant, ahead of the inevitable GPU debt for this deployment.
No matter what you read from the shady desks such as SemiAnalysis/Citrini et al, nothing will change the fact that $IREN has secured its H2 CY27 VR200 allocation, and with secure, permitted power, and a blazing fast construction process underway, Iren will land an impressive IG deal, including funding, and the market will change its mind overnight.
Did you all really forget that 2027 power is ULTRA scarce? Let alone 2027 LPS?
I'll be there, with more shares thanks to the discount of an uninformed market.
Long $IREN
1
11
638
TrendForce:
NAND prices are expected to start falling in Q2 2027 as Chinaβs YMTC aggressively expands production capacity. DRAM prices, by contrast, are expected to hold up through the end of 2027.
Conventional DRAM average selling prices (ASPs) are expected to rise 10β15% QoQ in Q4 2026. TrendForce forecasts more gradual increases in 2027, with ASPs up 5β10% QoQ in Q1, 3β8% in Q2, and 0β5% in both Q3 and Q4.
TrendForce expects HBM4 pricing to rise from at least $2.05 per Gb in 2026 to $3.00β$5.50 per Gb in 2027, and projects a year-over-year increase of more than 100%. Pricing discussions for next-generation HBM4E are at $5 per Gb or higher.
zdnet.co.kr/view/?no=2026111β¦
2
219
I hope everyone enjoys tomorrow !! $SPCX
Rumor has it @elonmusk has a little @TomorrowXSummit FOMO. π
See you tomorrow? π
1
3
256
$NVDA | Nvidia introduced Alibaba to Alice Huang for a workaround on U.S. export controls.
Nvidiaβs previously unreported role in connecting Alibaba and Megaspeed shows that they actively helped Chinese companies obtain Nvidia chips to bypass U.S. export limits. Those rules restricted direct sales to China but did not block overseas access by Chinese firms.
1
161
Brad Gerstner says 26% of Anthropicβs compute budget went to serving paying customers in the first 9 months of 2026.
That took about 500 MW of its 2 GW. Most of the rest went to training.
He estimates another 500 MW could support the extra $100B in revenue forecast for 2027. $ANT
1
94
Biggest line from $NVDA earnings today:
Supply commitments went $119B β $279B β $390B in two quarters because of MULTI-YEAR procurement of memory supply. $MU $SKHY
βWe've partnered with our extensive network of suppliers to secure the critical components needed to meet demand for the next several years. Our supply commitments increased from $279 billion last quarter to $390 billion, primarily related to the multi-year procurement of memory supply.β
Bullish!
5
340
Land, Power, and Shell are becoming just as important to the AI buildout as the chips.
Kress says hyperscalers are feverishly looking for more of NVIDIAβs available compute. She also called LPS availability critical to building data centers and a top priority for NVIDIAβs support.
The demand is there. More compute still needs a powered, finished site before it can go live.
The carousel breaks down the results, outlook, and commitments behind the buildout. $NVDA $IREN $NBIS
1
2
225
Texas issues air permit for massive SpaceX data center.
The permit approval arrives only weeks before Democratic Gov.-elect Gina Hinojosa assumes office, having pledged to halt new data center development. $SPCX
4
329
βJensen resigned from his position with NVIDIA, and the board accepted his resignation. The board has determined that they will not reinstate Jensen. And that is our statement and what the board has decided to do.β
Tough situation, obviously itβs time to find a new board preferably one that doesnβt solely have Taco Bell experience. $NVDA
1
3
397
If this is true, shareholders will wait for a decent pump and sell most of their $NVDA and move on to something like $GOOGL or $AMZN atleast from what Iβve heard.
The most important part of the call was this.
The board is steadfast about not reinstating Jensen back as CEO.
This will go down as the worst call in the history of calls.
Employees will leave. Shareholders will sell. Customers will switch.
I've already signed the petition, I'm not sure what else we can do to make the board listen.
If Jensen isn't reinstated by end of year, I will have no choice but to short $NVDA.
Please don't make me do this.
4
402
Nvidia hasnβt abandoned the neocloud buildout. Itβs still backing the ecosystem, but making support more conditional and project-specific. Thatβs all
π¨$NVDA cuts neocloud support plans down $1B q/q and adds tighter language around conditions; removes revenue share language from last quarter; adds new language that certain commitments "could be reduced or eliminated" if contingencies are not satisfactorily resolved. Cuts previously disclosed 2028 capex $1B to zero (sounds like a project pushout or cancellation).
Now says commitments take effect "only if the neoclouds secure financing to purchase the infrastructure products and the related compute capacity is deployed, made operational, and accepted by us."
Previous language: "We continue to make commitments....Under these agreements we will earn revenue on upfront sale of our infra and if certain criteria are met, we will participate in revenue share."
Re: secondary equity investments, future commitments fall from $25B to $20B but 2028 plans rise from $3B --> $11B. Unclear if this is pushouts from 2027 investments not made (such as Figure AI).
Bottom line: Total commitments grew 41% QoQ, driven almost entirely by a surge in 2028 supply/capacity commitments and long-dated data center leases, but the new contingency clause means a large portion is conditional on neoclouds securing external financing.
Also 10-Q to arrive "in coming weeks" seems like a delay.
5
283
While the market debates an AI spending slowdown, Kress says hyperscalers are feverishly looking for more of the compute NVIDIA has available and says customer demand hasnβt changed across customer types.
The quarter backs up that broad demand: hyperscale revenue grew 23% QoQ, while AI cloud, industrial and enterprise grew 18%.
The question is how quickly NVIDIA can deliver more compute. $NVDA
102



























