DanielRomero

$NBIS announces proposed private offering of $3.75 billion of convertible senior notes
The offering consists of:
•$2.0 billion of convertible notes due 2031
•$1.75 billion of convertible notes due 2033
Nebius also expects to give the initial purchaser an over-allotment option for up to:
•$300 million of additional 2031 notes
•$262.5 million of additional 2033 notes
The company said the proceeds will be used to support the continued growth of the business, including:
•construction and build-out of data centers
•development of its full-stack AI cloud
•expansion of its data center footprint
•proc

I dont think most people are aware of how fast things are advancing
NVIDIA’s last GTC event was eye-opening in several aspects
A quick summary
Bullish
•Energy access to cheap clean energy first, then behind the meter, then utilities, solar, nuclear developments
•Photonics (watch for those working with NVIDIA partners)
•Memory (the shortage isnt ending anytime soon, focus on the leaders)
•Physical AI suppliers (sensors, cameras, rare earths, all crucial and likely to hit a bottleneck at some point)
•Advanced packaging specially next-gen 3D packaging. For this, and for silicon photonics, hybrid

$MU has completed the acquisition of PSMC’s Tongluo P5 site in Taiwan for $1.8B in cash
The site adds about 300,000 square feet of existing 300mm cleanroom space and will support Micron’s expansion of leading-edge DRAM and HBM production to meet rising AI demand
Micron expects the facility to support meaningful product shipments in fiscal 2028
A second expansion phase is also planned, with construction set to begin by the end of fiscal 2026, adding roughly 270,000 square feet of additional cleanroom space

$AMD and Celestica Announce Collaboration to Advance the Next Era of AI with Helios Rack-Scale AI Platform
Celestica will undertake the R&D, design, and manufacturing of scale-up networking switches in the $AMD Helios rack-scale AI architecture, based on the Open Compute Project OCP, Open-Rack-Wide ORW form factor
The scale-up switches will utilize advanced networking silicon to enable the high-speed interconnect of the next-generation AMD Instinct MI450 Series GPUs, enabling leading-edge computing, optimized for large-scale AI clusters.
Consistent with the open standards-based design of the

$NBIS deal with $META is 40% larger than their deal with $MSFT
Just to put things into context

🚨 JUST IN
$NBIS has signed a new $27 billion deal with $META
Nebius has signed a 5-year agreement with Meta worth up to $27 billion
The deal includes $12 billion of dedicated AI compute capacity across multiple locations, based on one of the first large-scale deployments of the NVIDIA Vera Rubin platform, with delivery starting in early 2027
In addition, Meta has committed to purchase up to $15 billion of extra compute capacity across certain upcoming Nebius clusters over the same period. Nebius plans to sell this capacity to third-party AI cloud customers first, with any remaining capacity g

$TEM CEO says interest from OpenAI and Anthropic in its genomics database has recently risen from 1 to 5 out of 10, citing particular interest in longitudinal patient histories at scale

TrendForce projects that co-packaged optics will grow from 0.5% penetration of optical communication modules in AI data centers in 2026 to 35% by 2030
$LITE $COHR $AAOI $CIEN

$LITE is completely sold out through the end of 2027, with current capacity falling 25–30% short of customer demand, according to the CEO

$INTC 高管警告CPU短缺正在冲击"各行各业"，价格上涨即将来临
英特尔全球渠道主管Dave Guzzi通过CRN表示，英特尔的CPU供应限制正在对整个供应链产生更广泛的影响，影响包括云服务提供商、OEM和系统集成商
CPU需求大幅增加，部分原因是超大规模数据中心运营商在部署更多与代理AI相关的纯CPU工作负载。就英特尔而言，Sapphire Rapids等较老的Xeon芯片仍然需求强劲，因为许多客户认为它们比新一代芯片更可靠
这为英特尔造成了困难的权衡。公司必须决定是将更多生产产能分配给仍然支持高需求产品的成熟工艺节点如Intel 7，还是专注于Intel 3和18A等较新工艺节点，希望企业客户随时间推移实现过渡
与此同时，Raptor Lake的需求也有所上升，因为消费者在PC市场持续不确定的情况下转向购买上一代CPU以提前锁定购买。由于Raptor Lake也依赖于Intel 7，英特尔正在努力平衡消费类和数据中心市场之间的需求。目前，公司似乎优先考虑数据中心领域，因为其利润率更高、订单量更大

You could buy $MU at a $70B market cap less than a year ago
Absolutely insane

$AAOI expands its ATM equity offering program, doubling the capacity to $500 million

$PLTR and $NVDA Team to Deliver Sovereign AI Operating System Reference Architecture
Palantir and NVIDIA announced a joint sovereign AI OS reference architecture that provides a turnkey AI data center, covering everything from hardware infrastructure to application deployment
The system combines NVIDIA Blackwell Ultra AI systems and Spectrum-X networking with Palantir’s full software stack, including AIP, Foundry, Apollo, Rubix, and AIP Hub
The goal is to allow governments and enterprises to deploy advanced AI in sovereign, on-premise, edge, or controlled cloud environments, maintaining full c

Photonics stocks in layman's terms
$TSEM = $TSMC
$LITE = $NVDA
$COHR = $AMD
$BESI = $AMAT
$XFAB = $INTC
$AAOI = $MU
$POET = Groq
$ALMU = Cerebras
$AXTI = $SNDK
$CIEN = $CRDO
$SMTC = $ALAB
$SIVE = $LITE Lite