In June 2026, Hyperscale Data announced that it had entered into a 10-year master services agreement (the "MSA") with a California-based neocloud provider valued at approximately $1.2 billion. Expansion options contemplated within the MSA could increase the total potential contract value to more than $3.0 billion. $GPUS
Apple is evaluating PrismML’s technology, which the startup says can shrink powerful AI models enough to run directly on an iPhone while using up to 15x less memory. $MU $SNDK
I spoke to IBM CEO Arvind Krishna about the warning today and here’s what he told me:
“A few large capex deals paused near the end of the quarter. It’s not widespread. It’s mainframe and the software associated with it….Mythos is making people pause to say, wait how much do I need to spend on cyber? They’re pausing on new deals until they know…We don’t see our software being disrupted by ai at all.”
HOUSTON & DENVER--(BUSINESS WIRE)-- Global energy technology company SLB (NYSE: SLB) today announced an agreement with Liberty Energy Inc. (NYSE: LBRT) to form a strategic alliance that will deliver modular infrastructure and integrated power generation solutions for new data center projects globally.
HOUSTON & DENVER--(BUSINESS WIRE)-- Global energy technology company SLB (NYSE: SLB) today announced an agreement with Liberty Energy Inc. (NYSE: LBRT) to form a strategic alliance that will deliver modular infrastructure and integrated power generation solutions for new data center projects globally.
Q2 Results Above the High End of Range and Reiterate 2026 Guidance “We anticipate second quarter sales to come in at approximately $418 million, above our guidance of $388 to $403 million with strong sell through driven by consumer demand across our channels and products. Adjusted EBITDA is expected to be approximately $112 million. $TREX $JHX
It's nuanced, IMO:
- $TSLA has the clearest path to close the gap between US/China in mass production. But to get this done, majority of components seem sourced from China?
- $CCXI right now has a RoboFab for 10K+ units/year, and 75%+ of components sourced from the US.
So they're probably the leader in terms of Made in America supply chains + commercialization. But can't get to China scale yet through just US supply chains.
Figure is probably the other one with 12,000 target capacity, but can't comment much since their supply chain visibility isn't very public. I don't consider Hyundai/Boston Dynamics to be very American anymore so didn't include them.
BofA: Internet CapEx
> Alphabet $GOOGL: Updated to $195bn in 2026 (vs. $187bn previously) and $290bn in 2027 (vs. $257bn previously).
> Meta $META: Updated to $145bn in 2026 (vs. $130bn previously) and $185bn in 2027 (vs. $157bn previously).
> Amazon $AMZN AWS (Cloud only): Remaining at $159bn in 2026, but raised to $230bn in 2027 (vs. $196bn previously).
> Total Capacity: Combined capacity for the Big-3 mega-caps is estimated to grow from ~27GW exiting 2025 to 39GW in 2026 and 57GW in 2027.
> Capacity Additions (2026–2027): Amazon is projected to add the most capacity at 15GW, followed by Google at 9GW and Meta at ~6GW.
> Cost Advantage: Amazon AWS possesses the lowest estimated cost per incremental GW ($25bn/GW in 2026) compared to Google ($37bn/GW) and Meta ($45bn/GW). This reflects AWS's scale advantage and its reliance on custom internal silicon (Trainium/Graviton) rather than external GPUs.
> Total Sector Revenue: Hyperscaler AI capacity is projected to support roughly $900bn in total combined revenue by 2030 ($409bn for AWS, $387bn for Alphabet, and $110bn for Meta).
> Deal Premium Optimism: Current AWS and Google Cloud annual revenues per GW range from $10bn to $17bn. This sits well below recent specialized AI capacity deals signed by Anthropic and Google with SpaceX, which fetch up to $40bn–$50bn per GW, showing massive monetization upside.
> Meta Enterprise Opportunity: While Meta doesn't currently run a public enterprise cloud, selling its excess internal capacity to third parties at a conservative $12bn/GW could represent a $110bn potential enterprise opportunity by 2030.
Valuation Imbalance: The "Street" is Discounting Meta
Alphabet: Valued at an implied premium of $110bn per 2028 Cloud GW.
Amazon: Valued at $59bn per AWS GW.
Meta: Valued at just $4bn per AI GW, indicating that the Street is giving minimal to negative market value to Meta's massive infrastructure buildout.
@neruda_de No, I’m personally trying to own as much % of Agility Robotics as possible right now.
Since I see Humanoids is as next major theme following space/AI.
Don’t think historical PTSD with bad SPAC names will sway me from owning one of the US humanoid leaders.
We just published a report on Opendoor Technologies ($OPEN), whose "Opendoor 2.0" plan is basically window dressing on a model that has never worked. https://t.co/6zSXV9VJox
Year-to-date FY27, the Company has seen continued positive traction in paid consumption revenue and closed its first deals under Flex, its new consumption-based buying program, which became available to customers after the earnings call in June. $GTLB
The implied deal value of ~$48M represents a meaningful premium to recent trading. The CVRs provide additional upside. However, the 68% dilution to Scipher shareholders means existing CMMB holders are participating in a significantly reduced share of the combined upside. The stock should trade up significantly today — possibly 50-100%+ — as the market prices in the deal premium and the new strategic direction. At $2.31 pre-announcement with an implied deal value of ~$4-5/share at $150M combined valuation, there's meaningful upside to reprice. $CMMB
Just Bloomberg and $META doing damage control after crashing the market with Meta Compute framing:
Spokesperson: "Meta is still hungry for even more computing power.
It is still moving forward with plans for expensive new data centers and recently inked major computing deals with $CRVW, Google, $ORCL, and others."
Just dropped that in with the Meta Muse announcement, and evenn threw in the "expensive" framing with DCs to signal capex.
But little late given we're likely seeing a lot of margin liquidation cascades and heavy losses from media framing earlier.