Micron Technology
-
I think it's somewhat moot what is decided at the Govt level as the Chinese vendors don't have any product to sell, and based on the evidence are asking a premium. Plus, the over supply argument has been put to bed, regardless of what certain vested interest will continue to promote.
We will have severe shortages for at least 5 years and possibly much more. A very very long time in this game-trying to predict markets in 5 years is pointless so ignore that too.
If anything the infection of physical AI is a matter of when not if. Cars will get to level 4 soon, robots will proliferate. Both will need material memory to allow real time (zero shot) inference. It's obvious, well at least to some.
We will discuss this in due course but the next phase in the memory road map is custom silicon where companies like SK/Micron are working with customers to produce bespoke silicon. We know they are working with Anthropic and Nvidia. Why it's a big deal. It makes their products not just commoditised, it means these co-designs are deeply embedded in their AI architecture which effectively locks out competitors.
-
Jensen Huang (NVIDIA founder and CEO) and Tim Cook (Apple CEO) both provided supportive statements in Micron’s August 20, 2026 announcement of Micron Research Labs (a U.S.-based long-horizon research institution headquartered in Boise, Idaho, backed by a planned $10 billion investment
Jensen Huang:
“Micron is taking on one of the great challenges of the AI era — reinventing memory technologies and architectures to fuel the next generation of increasingly powerful AI systems. By bringing together advanced manufacturing and the broader technology ecosystem, Micron is helping drive the breakthroughs that will define the next era of AI and computing.”and surprising, given the rift (restoring relations?)
Tim Cook:
"For more than two decades, Micron has been an important partner, providing memory technologies for Apple’s groundbreaking products that people love and use every day around the world. With the launch of Micron Research Labs, they are building on a legacy of leadership in semiconductor research to drive breakthroughs in memory and computing for decades to come. Apple believes deeply in American innovation, and we’re proud to support Micron as it expands leading-edge manufacturing and R&D in the United States.” -
Yesterday Aug 20),Micron Technology (MU) CEO Sanjay Mehrotra said AI has fundamentally changed the memory business, creating a more durable source of demand for an industry historically prone to boom-and-bust cycles.
“Today there is no AI without memory. AI systems need more memory. They need higher performance memory. They need lower power memory,” Mehrotra told Jim Cramer on “Mad Money.” “So, the value of memory, that equation has totally changed.”He added that data-centre customers currently want roughly 50% more supply than Micron can commit, underscoring the tightness in the memory market.Mehrotra also called memory “the strategic infrastructure of the AI era,” pointing to growing demand beyond data centres from autonomous vehicles, robots, and AI-enabled consumer devices.
“Without memory, you cannot make AI smarter or faster, without memory you can’t scale up AI. AI is advancing and a lot of content is getting generated. All of that content has to be stored, has to be processed through memory,” he added.How bullish this is / re-rating the company.
This is strongly bullish and supports a meaningful re-rating of Micron. Historically, memory (DRAM/NAND) has been a classic cyclical industry: sharp price collapses when supply outstrips demand, followed by recoveries. Investors therefore applied low valuation multiples reflecting that volatility and limited visibility. Mehrotra’s comments mark a clear structural shift. He frames AI as creating durable, multi-year demand that is no longer purely cyclical: AI systems require more, higher-performance and lower-power memory; data-centre customers already want ~50% more supply than Micron can currently commit (tight market conditions); and memory is positioned as “the strategic infrastructure of the AI era,” with demand also expanding into autonomous vehicles, robots and AI-enabled devices. In valuation terms this implies: Higher and more sustainable utilisation/pricing power.
Better earnings visibility and potentially elevated through-cycle margins.
A transition from “cyclical commodity” to “structural AI enabler.”Markets typically reward that shift with higher multiples (P/E, EV/EBITDA, etc.). If investors accept the durability thesis, Micron can re-rate from a traditional memory multiple toward something closer to other AI infrastructure names, supporting a higher share price even before further earnings upgrades.
-
Nvidia is shifting its primary HBM4 configuration from 12-high to 8-high stacks for the second half of the year, driven by thermal-management needs and supply-chain resilience.
Samsung and SK Hynix, previously focused on denser 12-high products for the Vera Rubin platform, are now increasing their 8-high share to meet Nvidia’s request.
Micron has been the only major supplier committed to 8-high HBM4 from the start.
This change validates Micron’s strategy, eliminates a potential competitive disadvantage, and positions it for stronger volume and share as Nvidia seeks a broader, more balanced mix of 8-high product across all three suppliers. The same preference is also emerging for next-gen HBM4E. In short, Nvidia just endorsed the path Micron was already on — clear product validation and a meaningful win for MU.
-
from nvidia earnings prepared notes-wow
rerating?
-
Micron share price is like a day trading one at the moment …
-
Micron share price is like a day trading one at the moment …
@Ducati996R said in Micron Technology:
Micron share price is like a day trading one at the moment …
Indeed - it's all over the place. I await some nice steady growth in the coming months but do wish it would get on with it ......
-
Micron 512GB DDR5 RDIMM
Micron has just demonstrated the world’s first 512GB DDR5 RDIMM, aimed squarely at next-generation AI/data-centre servers. It is a significant step up in memory density: one module holds 512GB, with speeds of up to 9,200 MT/s. Micron expects volume production in H2 2027. AMD and Intel are already validating it on future server platforms.The power saving is particularly striking. One 512GB module consumes approximately 16W, versus 44.2W for four 128GB modules providing the same capacity — a reduction of more than 60%.
The key fact to add is that Micron's memory opportunity isn't solely an AI/HBM story. The new 512GB DDR5 module targets the much broader non-AI data-centre/server market, while HBM4/HBM4E captures the AI accelerator market.
That gives MU two separate demand drivers. Even at $15k per module, you're looking at $360k of DRAM per server.
Validation of durability!
-
Whats your thoughts on the upcoming MU earnings @Adam-Kay ? MU is flying at the moment.
-
Hi,
Guidance was $50b and $31 EPS(say $32 billion net income). I expect a significant beat. The market expects $53B. I think it will be quite a bit more. $58, maybe as high at $60B (eps $35?). Will they guide $70b+ for Q1(fiscal 27). I don't know, nobody does. Just remember one simple fact, there is no peak in sight, they are constrained. Demand is increasing every quarter, every year for as far as they can see. Micron are now producing quarterly earnings only a few elites can match. More than Apple, as much as GOOG. 15X more than AMD (same valuation
) 1000X more than Crowdstrike
The question is, are their earnings durable-I think they. You dont have to maintain 90% margins to justify a higher valuation because the market has never valued them accordingly. My view is long term the company will be able to produce EPS as a base line in the range $200 with higher peaks before.I have no idea what the reaction will be-it's not priced accordingly(is it) so im only interested in the numbers and what management have to say
My thoughts are-everyone is talking about HBM, however HBM is actually very expensive to produce as it hoovers up a large wafer count vs the end bit yield and is actually a drag on margins. Inference re agentic Ai is making other Dram more important when it comes to margins.
The AI memory race is shifting beyond pure HBM market share towards CPU-attached memory, where margins look structurally stronger and Micron currently holds the advantage(over SK).
Agentic AI workloads require agents to maintain long-running state, plan and manage expanding KV caches. While model inference still runs on GPUs with HBM, much of the supporting work and context lives on the CPU side in system memory and storage.
This is pushing demand towards high-capacity, power-efficient LPDRAM in SOCAMM form factors, high-capacity server DDR5 RDIMMs and data-centre SSDs for offload. Nvidia’s own software already moves KV cache out of scarce HBM into these cheaper, higher-capacity tiers .In 2026 HBM pricing was largely locked in at lower levels from earlier negotiations, while conventional DRAM contract prices surged several-fold.
Because HBM also consumes far more wafer capacity per bit, its revenue and profit per wafer fell behind ordinary server DRAM. Suppliers with heavier HBM exposure therefore faced a margin drag; those able to tilt mix towards CPU-side products enjoyed superior pricing power and profitability.
DDR5 server memory and SSDs are emerging as the huge revenue and margin drivers. Surging prices, strong volume growth from agentic demand and far better wafer economics than locked-in HBM make them highly profitable. Micron, with the smallest HBM share among the big three, grew DRAM revenue fastest and posted the highest operating margins. It pioneered data-centre SOCAMM, led capacity ramps, and saw its Core Data Center Business Unit (covering high-capacity DDR5, LPDRAM and storage) become the company’s fastest-growing segment. The winners in this phase are therefore those strongest in these CPU-centric memory types, not simply the HBM volume leaders.
I will be listening to the call for comments re the above