On January 9, 2026, Oklo and Meta announced an agreement supporting development of a nuclear-energy campus in Pike County, Ohio capable of scaling to 1.2 GW.
The project is intended to support Meta's growing Ohio data-center operations, including AI infrastructure.
The current planned schedule is:
| Milestone | Target |
|---|---|
| Pre-construction/site work | 2026–2027 |
| General construction | Early 2028 |
| First-phase operation | 2030 |
| Full 1.2 GW buildout | 2034 |
The most important clarification is:
Meta is not currently receiving 1.2 GW from operating Oklo reactors. The agreement supports development of a future phased nuclear campus.
AI workloads require large computing clusters.
Those clusters require:
For hyperscalers, the challenge is not merely buying renewable-energy certificates. They increasingly need new physical power generation that can support long-duration, high-load operations.
Nuclear power can potentially provide:
Advanced-reactor developers argue that smaller units can also be deployed closer to large industrial or computing loads.
Oklo says the agreement provides a mechanism for Meta to:
Oklo plans to use this funding to secure nuclear fuel and advance Phase 1.
This is more economically meaningful than a simple statement that the companies are “exploring nuclear.”
Nuclear projects require substantial capital before revenue begins.
A customer that provides early money can potentially help with:
That may reduce part of the funding burden that would otherwise fall on equity investors.
Oklo's disclosed model remains centered on Oklo developing and operating its own powerhouses.
The agreement is structured around the supply of future power rather than Meta simply purchasing a reactor from Oklo.
1.2 GW equals:
1,200 MW
With current Aurora units designed around 75 MW each, a full-scale campus represents a large multi-unit deployment. Oklo's project materials refer to a 16-unit site when discussing staffing.
Oklo controls more than 200 acres in Pike County at the former Portsmouth Gaseous Diffusion Plant area.
The site also sits near Centrus' Piketon uranium-enrichment operations, creating potential strategic links between:
Oklo and Centrus announced plans to explore a HALEU deconversion joint venture in Pike County.
Then in June, they signed an LOI under which Centrus intends to supply enough HALEU for multiple years of operation for up to five Aurora units, with deliveries scheduled from 2029.
This links:
Fuel production
↓
Aurora
↓
Meta-supported demand
within the same broader Ohio region.
Not yet.
The June announcement explicitly states that the LOI anticipates negotiation of a future definitive agreement.
Therefore, articles should not say:
“Centrus has guaranteed all fuel needed for the entire 1.2 GW campus.”
That would overstate the current commitment.
Oklo currently targets first-phase operations for the Ohio campus as early as 2030.
Between now and then, important steps include:
The project therefore has substantial execution risk despite the size of the commercial opportunity.
The full 1.2 GW campus is targeted for 2034.
That means investors valuing all 1.2 GW as if it were operating today are ignoring nearly a decade of:
The agreement can potentially improve investors' confidence in:
A major hyperscaler wants future nuclear capacity.
Customer prepayments can support development.
The contemplated campus is far larger than a single Aurora unit.
Success could create a blueprint for other data-center customers.
2030 and 2034 are targets, not guarantees.
Large fleets require large quantities of qualified fuel.
Many repeated units must be manufactured and installed.
Commercial regulatory approvals remain necessary.
Oklo must ultimately deliver electricity at economics acceptable to both itself and customers.
No.
It materially strengthens the commercial story, but future revenue still depends on project execution and the detailed commercial arrangements.
Development funding should not automatically be treated as equivalent to decades of realized electricity revenue.
OKLOON is linked to OKLO.
If the market increases its valuation of Oklo because the Meta campus progresses, OKLO may respond and OKLOON may follow.
But OKLOON does not itself represent:
1.2 GW of Meta electricity capacity.
For the product distinction, read What Is OKLOON?.
Up to 1.2 GW.
Pike County, southern Ohio.
As early as 2030 under the current target.
No.
The agreement includes mechanisms for prepayment and early-stage funding.
The Ohio project remains under development. Its capacity, economics, timing, licensing, financing and commercial structure can change.

美国6月就业报告发出了明确的经济放缓信号,数据不及预期,降低了对美联储加息的担忧。 尽管美国国债收益率下降,但AI和科技股依然下跌,这证明宏观利率环境的改善已不再是高估值科技股的绝对催化剂。 比特币和黄金有望从美元走弱中获益,但需要更多确认信号来维持涨势。 交易员应将注意力从美联储政策转向第二季度的企业财报质量,以此来验证当前市场的溢价估值。

辉达的毛利率刚刚连续第三季度维持在接近 75% 的水准。 而在同一份新闻稿里,公司下修了这个数字的指引。 营收仍在加速——截至 2026 年 7 月 26 日的当季达 962 亿美元,比一年前的两倍还多,而下一季的指引则是 1,080 亿美元。 也就是说,公司一边加速增长,一边在每一美元营收上赚得更少;而这一组张力解释了大部分的原因,这组张力也是为什么覆盖同一家公司的分析师,连一年后的目标价都无法

AI 推理需要的不只是算力,大规模部署同样需要不断增长、能够快速访问且具备成本效率的存储容量。 NAND 正逐步成为与 HBM、DRAM 并存的 AI 容量层,而不再只是传统商品型存储产品。 Sandisk 预计,到 2030 年企业数据中心闪存需求将达到 1.2 ZB,并正在开发专门面向 AI 推理的 High Bandwidth Flash。 Sandisk、Samsung 等存储厂商开始采用

Key Takeaways Microsoft、Amazon、Alphabet 和 Meta 在 2026 年仍在大规模投入 AI 基础设施和数据中心。 Nvidia 仍是最直接的 AI 资本支出受益者之一,但 AI 云、服务器、网络和光通信也开始呈现明确增长。 CoreWeave、Nebius、Dell 和 Broadcom 提供了 AI 支出转化为营收、订单和待履约收入最清晰的证据之一。 随着