Morgan Stanley said it did not see worsening U.S. data-center power bottlenecks putting Nvidia’s or Broadcom’s 2027 forecasts at risk. The bank’s assessment, reported by Reuters on October 5, 2026, is not a guarantee: it rests on the companies’ visibility into chip placement, ability to serve projects in different geographies, and coordination with data-center and power-sector participants.
What Morgan Stanley says about Nvidia and Broadcom
The distinction is relative resilience, not immunity from power constraints. Reuters reported that Morgan Stanley believed Nvidia and Broadcom had enough visibility and flexibility that the bottlenecks were not putting their 2027 forecasts at risk. The report describes an analyst judgment, not a confirmed outcome or assurance about either company’s results. Reuters, reproduced by Investing.com, published the assessment on October 5, 2026.
Why their position may be different
According to the reported rationale, Nvidia and Broadcom have better visibility into where chips can be placed and can adapt to expansion across geographies. Coordination among data centers, semiconductor suppliers, and the power supply chain may also help align chip deliveries with projects that can proceed. Quartz relayed a further possibility—that chips could be redirected to projects with secured power—attributing that idea to Benzinga; it is a secondary explanation, not a guarantee that redirection will occur. Quartz’s October 6, 2026 report discusses the assessment.
How large is the U.S. data-center power gap?
Morgan Stanley estimated a net U.S. data-center power shortfall of 34%, or 32 gigawatts, through 2028, after accounting for measures such as behind-the-meter generation and fuel cells. Reuters reported that estimate on October 5, 2026. It is a forecast of a potential gap, not a measurement that a 32-GW shortfall has already occurred. The estimate applies to U.S. data-center power needs; it should not be read as a global figure or as a count of projects already canceled. Reuters’ report is the source for this newer figure.
#1 Best Overall
- AI Performance: 767 AI TOPS
- OC mode: 2632 MHz (OC mode)/ 2602 MHz (Default mode)
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Axial-tech fan design features a smaller fan hub that facilitates longer blades and a barrier ring that increases downward air pressure
- A 2.5-slot design maximizes compatibility and cooling efficiency for superior performance in small chassis
Which suppliers could be more exposed if projects slip?
The reported risk falls more directly on component suppliers if customers cannot bring data-center capacity online and delay or cancel related orders. Reuters identified memory, optical, power-management, and analog components as categories most exposed to possible inventory disruption. That is a conditional scenario: the report does not establish that customers have already canceled orders or that suppliers have already accumulated excess inventory.
| Supply-chain position | Exposure if deployment is delayed | What the reporting establishes |
|---|---|---|
| Nvidia and Broadcom, described as large chip suppliers | Potentially more flexibility to place chips across projects and geographies, according to Morgan Stanley’s reported assessment | The bank did not see the power bottlenecks putting their 2027 forecasts at risk; this is a forecast judgment, not proof of immunity. |
| Memory, optical, power-management, and analog component suppliers | Potential inventory disruption if customers push out or cancel orders when deployments cannot proceed | Reuters reported these categories as most exposed in that scenario; it did not report that the disruption had already happened. |
The difference is about supply-chain position and flexibility, not a definitive ranking of company performance. A delayed project can affect the timing of many orders, but the available reporting supports a conditional exposure distinction rather than a conclusion that any named supplier will lose revenue.
Rank #2
- Chipset: GeForce RTX 3050
- Boost Clock / Memory: 1492 MHz / 14 Gbps
- Video Memory: 6GB GDDR6
- Memory Interface: 96-bit
- Output: DisplayPort x 1 (v1.4a) / HDMI 2.1a x 2
Why power constraints can delay data-center investment
Securing power is one part of bringing a data center online. Interconnection and construction schedules can stretch across years, making it possible for hardware spending and deliveries to occur before a facility is ready to operate. Morgan Stanley Research’s report, “Bridging the Data Center Gap”, cites Lawrence Berkeley National Laboratory figures showing that projects built in 2022 took five years from interconnection request to commercial operation, compared with two years for projects built in 2015 and under two years in 2008. These are historical timelines for the specified project cohorts, not a universal estimate of how long a current data center will take to build.
The broader data-center spending forecast
Morgan Stanley Research estimated about $2.9 trillion in global data-center capital spending from 2025 through 2028. Its model allocated roughly $1.3 trillion to construction and $1.6 trillion to hardware, and estimated approximately $1.7 trillion of the total would be spent in the United States. These are model estimates, not audited spending totals or committed budgets. The report notes that construction takes multiple years and that some capital is spent before a facility’s operating year, so projected investment and operating capacity do not necessarily arrive at the same time.
Recommended Free Tools
Rank #3
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Powered by GeForce RTX 5070
- Integrated with 12GB GDDR7 192bit memory interface
- PCIe 5.0
- NVIDIA SFF ready
The same report anticipated that some workloads could be served outside the United States if domestic power is insufficient. That possibility is consistent with the geographic-flexibility rationale cited for Nvidia and Broadcom, but it does not establish how much demand will move or which suppliers will benefit. The 34% and 32-GW figures above come from Morgan Stanley’s newer U.S. shortfall estimate as reported by Reuters; older or differently scoped estimates should not be substituted for it.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What this means for investors
The reporting offers a supply-chain explanation for why the power crunch may not affect all chip-related companies equally: Nvidia and Broadcom are described as having more flexibility to manage placement, geography, and coordination, while component suppliers could face order timing and inventory risk if projects are postponed. It does not establish realized losses, guarantee either large supplier’s forecast, or support a buy-or-sell recommendation. Investors should treat the power-gap and spending figures as attributed estimates and the supplier disruption as a risk scenario, not as reported results.
Quick Recap
Rank #4
- Powered by the NVIDIA Blackwell architecture and DLSS 4. System Requirements: Minimum 850W PSU with 16-pin 12V-2x6 (12VHPWR) connector required. Verify before purchasing.
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability. Compatibility: 348mm (13.7") length, 3.6 slots, 4.3 lbs. Confirm case clearance and slot spacing. GPU bracket included.
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.6-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
