Apple may launch an AI‑focused server product in 2029 that will run on two to four next‑generation M8 Ultra ARM chips and could incorporate Nvidia’s NVLink Fusion interconnect, according to a report by The Verge citing The Information.
Company background and recent financial health
Apple Inc. is headquartered in Cupertino, California, and trades on the Nasdaq under ticker AAPL. The SEC filing for the quarter ended 27 June 2026 shows net income of USD 101.464 billion and total assets of USD 383.266 billion (Form 10‑Q filed 31 July 2026). Shareholders’ equity stood at USD 107.520 billion and the company had 14.608963 billion shares outstanding. Nvidia Corp., based in Santa Clara, trades under ticker NVDA. Its most recent Form 10‑Q (filed 26 August 2026) reports revenue of USD 177.837 billion, net income of USD 118.010 billion, total assets of USD 320.272 billion, and shareholders’ equity of USD 228.984 billion with 24.304 billion shares outstanding. Both firms remain financially robust, giving them the capacity to invest in new hardware platforms.
Rumor details: a server line that never existed
Apple discontinued its Xserve line in 2011 and has not offered a dedicated enterprise server since, as noted in the Verge article. Recent shortages of the Mac Mini and Mac Studio among AI developers have revived speculation that Apple sees a market gap for power‑efficient, high‑performance compute. The Verge report states that the planned server would be built around two to four M8 Ultra ARM processors and that “Apple could borrow tech from Nvidia to make it happen,” specifically Nvidia’s NVLink Fusion interconnect that links multiple chips as a single logical unit. The same source adds that Apple has already used Nvidia chips to power its revamped Siri servers, indicating a willingness to work with Nvidia’s silicon.
Sector impact: what a 2029 Apple server could mean
If Apple follows through, the move would re‑introduce a major U.S. hardware vendor into a space now dominated by cloud providers and specialist chip makers. The rumored specifications place Apple’s offering between Amazon’s Graviton3‑based instances (ARM, 1‑8 CPUs, custom interconnect) and Google Cloud’s TPU‑v5 ASICs (1‑4 CPUs, proprietary interconnect). The table below summarises the projected Apple server against these competitors.
| Company | Server Name (rumored) | Launch Year | CPU Architecture | Number of CPUs | Interconnect Technology |
|---|---|---|---|---|---|
| Apple | Apple Server (rumored) | 2029 | ARM (M8 Ultra) | 2‑4 | Nvidia NVLink Fusion (rumored) |
| Amazon (AWS) | Graviton3‑based instances | 2023‑present | ARM (Graviton3) | 1‑8 | Custom interconnect |
| Google Cloud | TPU‑v5 | 2025‑present | ASIC | 1‑4 | Google’s proprietary interconnect |
Source: The Verge (16 Sept 2026) and public product announcements from AWS and Google Cloud.
For AI developers, the appeal would be Apple’s reputation for energy‑efficient silicon combined with Nvidia’s proven multi‑chip scaling. If the NVLink Fusion interconnect is realised, Apple could compete on throughput per watt, a metric increasingly important as data‑center operators seek to curb power costs.
Outlook and unanswered questions
The plan remains speculative. The Verge article frames the information as a rumor sourced from The Information, and no official comment from Apple or Nvidia has been published. Key unknowns include the final pricing, the form factor (rack‑mount vs. blade), and whether Apple will market the product under the historic Xserve brand or a new name.
From an investor perspective, the potential re‑entry could diversify Apple’s revenue beyond consumer devices, but the timeline—launch not before 2029—means any impact on near‑term earnings will be limited. Nvidia could benefit from a new high‑volume partner, yet the partnership would be a departure from the historically “frosty” relationship between the two firms, as noted in the source excerpt.
Analysts will watch for any filing or patent activity that signals hardware development, as well as supply‑chain signals such as orders for M8 Ultra silicon or NVLink components. Until Apple confirms the plan, the market should treat the story as a high‑confidence rumor rather than a firm commitment.
In summary, Apple’s alleged 2029 AI‑server project—two to four M8 Ultra chips, possible Nvidia NVLink Fusion, and a return to enterprise hardware—could reshape the competitive dynamics of high‑performance compute if it materialises. The sector impact hinges on Apple’s ability to marry its energy‑efficient ARM designs with Nvidia’s interconnect expertise, a combination not yet seen in a commercial server offering.