Micron
Micron in the enterprise
As of November 5, 2026. Every figure links to its source; company-reported numbers are labeled as such.
Micron makes the memory and storage chips that AI systems run on. Its most important AI product is high-bandwidth memory (HBM), stacked memory chips packaged right next to AI processors from Nvidia and AMD. It also sells server memory and data-center flash drives to the large cloud companies. Micron does not sell AI software. Its business customers are chipmakers, cloud operators and AI labs that buy memory in bulk, increasingly under multi-year contracts.
The AI boom has transformed Micron's results. Fiscal 2026 revenue reached a record $133.19 billion, up from $37.38 billion the year before. In the fourth quarter alone, revenue was $54.23 billion, about 4.8 times the year-earlier figure. Its two data-center units, Core Data Center and the HBM-heavy Cloud Memory unit, brought in $34.29 billion of that quarter's revenue. Micron no longer reports HBM sales separately; The Next Platform estimates them at about $14.3 billion for the quarter. Demand is locked in for years ahead. Micron says it has agreed terms for most of its 2027 HBM supply, has taken $32 billion in customer prepayments and has $150 billion in contracted backlog. It also has 26 multi-year Strategic Customer Agreements that cover about 35% of revenue through fiscal 2030.
Micron is in high-volume production of its newest HBM generation (HBM4) and other memory for Nvidia's Vera Rubin platform. It is also co-developing a custom version of the following generation (HBM4E) with Nvidia. Its AMD wins include the Instinct MI350 accelerators, and it signed a supply agreement with Anthropic in June 2026. Micron still ranks third in HBM, behind Samsung and SK hynix. What to watch: whether Micron gains HBM market share, whether the long-term contracts keep prices firm when supply catches up, and its revenue guidance of about $61.5 billion for fiscal Q1 2027.
Key numbers
| Measure | Figure | As of | Source |
|---|---|---|---|
| Fiscal 2026 revenue | $133.19 billion (vs. $37.38 billion in FY2025) | FY2026 (ended Sep 2026) | Micron Reports Record Fiscal Q4 Revenue Of $54.23 Billion And FY2026 Revenue Of $133.19 Billion (Pulse 2.0) |
| Data-center revenue (Core Data Center + Cloud Memory units) | $34.29 billion in the quarter, up 5.6x year over year | Fiscal Q4 2026 | Micron's Numbers Show Why HBM4 Memory Is Worth 15X Its Weight In Gold (The Next Platform) |
| HBM revenue (analyst estimate; Micron no longer discloses it) | About $14.3 billion in the quarter, up about 7.6x | Fiscal Q4 2026 | Micron's Numbers Show Why HBM4 Memory Is Worth 15X Its Weight In Gold (The Next Platform) |
| Contracted backlog / customer prepayments | $150 billion backlog; $32 billion prepaid by customers | Sep 2026 | Micron's Numbers Show Why HBM4 Memory Is Worth 15X Its Weight In Gold (The Next Platform) |
| Multi-year Strategic Customer Agreements | 26 agreements covering about 35% of revenue through FY2030 | Sep 2026 | Micron's Numbers Show Why HBM4 Memory Is Worth 15X Its Weight In Gold (The Next Platform) |
| Next-quarter revenue guidance | About $61.5 billion (plus or minus $1.5 billion) | Fiscal Q1 2027 outlook, Sep 2026 | Micron posts record quarterly sales of $54.23b, eyes even stronger fiscal 2027 (The Korea Herald) |
What it sells to businesses
| HBM4 / HBM3E high-bandwidth memory | Stacked memory chips packaged alongside AI processors so they can move data fast enough to train and run large models. |
| HBM4E (custom base die) | Next-generation AI memory, targeted for 2027, that customers can have customized for their own chips. |
| SOCAMM2 server memory modules | Compact, low-power memory modules for AI servers such as Nvidia's Vera Rubin systems. |
| Micron 9650 PCIe 6.0 data-center SSD | High-speed flash storage drive for AI data centers, billed as the industry's first PCIe 6.0 data-center SSD. |
| Server DRAM (DDR5, LPDDR) | Standard and high-capacity main memory sold to cloud providers and server makers. |
Named customers and deployments
| Customer | What they did | When | Source |
|---|---|---|---|
| Nvidia | Micron entered high-volume production of HBM4 (36GB 12-high stacks) for Nvidia's Vera Rubin GPU platform, alongside 192GB SOCAMM2 modules for Vera Rubin NVL72 systems and its 9650 SSD. Micron says it was the first supplier shipping all three in volume for Vera Rubin. | Mar 2026 | Micron enters high-volume production of HBM4 for Nvidia Vera Rubin (Tom's Hardware) |
| Anthropic | AI infrastructure supply agreement for Micron memory and storage, plus joint study of memory performance on AI workloads; Micron also invested in Anthropic's Series H round. Financial terms not disclosed. | Jun 2026 | Micron, Anthropic sign AI infrastructure supply agreement (The Star / Reuters) |
| AMD | Micron HBM3E 36GB 12-high memory designed into AMD's Instinct MI350 Series accelerators (288GB of HBM3E per GPU). | Jun 2025 | Micron HBM Designed into AMD AI Platform (StorageNewsletter) |
Partnerships
| Partner | Deal | When | Source |
|---|---|---|---|
| Nvidia | Co-developing what Micron calls the industry's first custom HBM4E implementation (NVHBM) for Nvidia's next-generation GPUs and NVLink Fusion platforms. | Sep 2026 | Micron's Numbers Show Why HBM4 Memory Is Worth 15X Its Weight In Gold (The Next Platform) |
| TSMC | TSMC will manufacture the base logic die for Micron's HBM4E, including custom designs for customers willing to pay a premium; production targeted for 2027. | Sep 2025 | Micron teams up with TSMC to deliver HBM4E, targeted for 2027 (Tom's Hardware) |