ReportGem ReportGem EN

实时全球研报

North America Connectors & Other Components, Electronic Components & Equipment and Hardware & Storage: 2Q26 Earnings Preview – Agentic AI Inflection Supports Storage & Optical Momentum; Remain Constructive on AI-Levered OEMs

发布日期: 2026-07-13研究机构: Citi公司 / 股票: APH,CDW,GLW,LOGI,SNDK,STX,SMCI,TEL,WDC报告页数: 55原文语言: English证据页码: 1

研报英文原文证据摘录

North America Connectors & Other Components, Electronic Components & Equipment and Hardware & Storage: 2Q26 Earnings Preview – Agentic AI Inflection Supports Storage & Optical Momentum; Remain Constructive on AI-Levered OEMs

Action |

13 Jul 2026 03:42:28 ET │ 55 pages

North America Connectors & Other

Components, Electronic Components

& Equipment and Hardware & Storage

Asiya Merchant, CFA AC

2Q26 Earnings Preview – Agentic AI Inflection Supports Storage & +1-415-951-1752

Optical Momentum; Remain Constructive on AI-Levered OEMs asiya.merchant@citi.com

Michael Cadiz

CITI'S TAKE +1-415-951-1671

michael.cadiz@citi.com

Our coverage is up +110% YTD on average versus S&P500 +10%, led by

outperformance, in descending order, from storage, HW OEMs exposed to

AI deployments/enterprise server refresh, and connectivity components.

We hold the most conviction on storage names for very favorable S/D

fundamentals underpinned by durable AI-led datacenter demand for both

NAND and HDD storage, followed by HW OEMs where agentic-AI is driving

incremental demand for dense systems, while demand outside of AI

remains supported by ongoing refresh/modernization cycles and

incremental AI-driven demand into traditional infrastructure. Lastly

networking infrastructure is benefitting from AI-driven capex/content

tailwinds including optical. We maintain upside Short-Term Views on

SNDK, STX and WDC.

– We remain constructive on storage components where the evolution from

traditional AI to agentic AI is driving a structural inflection in infrastructure

demand, where exponentially more complex, autonomous 24/7 agent systems

continuously perform multi-step inference, interact with external tools, and

generate persistent data exhaust—fundamentally shifting the bottleneck from

本摘录由系统从所标注的 PDF 证据页直接提取并保留英文原文,不做批量翻译;登录后在阅读器切换中文时才按需翻译。

打开研报阅读器