Micron's $22B Commitments, SpaceX Slides 25%, Meta Chip Update

Micron locked in $22 billion worth of customer commitments for AI memory supply, while SpaceX’s post-IPO slide deepened and Meta’s in-house chip push moved closer to production.

Key points

• Micron disclosed 16 strategic customer agreements totaling $22 billion in supply commitments, announced on June 24.
• SpaceX (NASDAQ: SPCX) has fallen roughly 25% since its June IPO, with debt markets widening on concerns about losses in its launch and AI units.
• Meta’s newest in-house AI chip is reported to enter production in September, a meaningful step in the company’s push to reduce reliance on third-party silicon.
• Nvidia’s forward P/E has actually declined as its stock price rose, with earnings growth outpacing a roughly 12% year-to-date gain.
• Jim Cramer argued that Nvidia’s valuation is misunderstood by the market, calling it the most proprietary chip company ever built.

What does Micron’s $22 billion in commitments actually mean?

The headline number is large, but context matters. These are customer commitments to secure supply, not prepaid purchase orders. Still, for Micron, they represent a structural shift in how the memory market operates. AI infrastructure spending has created a class of hyperscale and enterprise buyers willing to lock in access to high-bandwidth memory and other advanced chips years in advance, rather than buy on the spot market.

That dynamic is exactly what Micron’s management has been telegraphing. Sixteen agreements across what are presumably major cloud and AI customers gives the company better visibility into future revenue and justifies continued capital investment in leading-edge memory capacity. It also reduces the boom-and-bust volatility that has historically punished memory stocks during downturns.

The timing is notable. Jim Cramer recently pointed out, with some irritation, that investors have been assigning Micron a higher forward multiple than Nvidia. Whether or not that view holds, the $22 billion commitment figure does give Micron a credible demand story to point to.

Why is SpaceX stock still falling?

SpaceX joined the Nasdaq-100 after its June IPO, but the fanfare has not translated into price support. Shares are down about 25% from the IPO, and the reasons cited are straightforward: a trillion-dollar-plus valuation sitting on top of reported deep losses in both its launch operations and AI units, combined with debt markets that appear to be repricing the company’s risk upward.

Adding pressure, lock-up expiry dates are approaching. When insiders become free to sell, markets often anticipate that supply increase and adjust prices before the expiry hits. That dynamic, layered on top of existing valuation concerns, creates a difficult near-term setup for buyers.

Jim Cramer went further, suggesting SpaceX’s IPO weight in index funds contributed to a broader Nasdaq selloff as passive flows adjusted. That is a speculative causal claim, but the correlation between SPCX’s slide and broader tech weakness has drawn attention.

Longer-term bulls point to SpaceX’s potential across satellite internet, launch services, and AI-adjacent infrastructure. Some projections suggest a $500,000 investment today could grow to over $1.6 million by 2030, though analysts like Jeremy Grantham have publicly pushed back on that kind of optimism. The range of outcomes here is unusually wide, even by high-growth tech standards.

Where does Meta’s chip production timeline fit in?

Meta’s in-house AI chip moving to production in September is the clearest new data point in an ongoing story. The company has been investing heavily in custom silicon to reduce dependence on Nvidia’s GPUs and manage the cost of running its AI workloads at scale. A September production date suggests the chip is past the tape-out and validation stages and moving toward deployment.

This is part of a broader hyperscaler trend. Google, Amazon, and Microsoft have all built or are building custom AI accelerators. Meta joining that group in a more serious way matters for the chip supply chain, and potentially for Nvidia’s long-term addressable market among the largest AI spenders. For now, Nvidia still dominates training workloads, but inference at Meta’s scale is enormous, and that is where in-house chips tend to attack first.

No financial figures around the chip program were disclosed in the sourced reporting. Investors should watch for any commentary during Meta’s next earnings call for volume and deployment specifics.