The market is about to open: storage chip prices are rising collectively, Bloom Energy is included in the S&P 500, Dell is included in the S&P 100, both are experiencing a double increase, what other news is worth paying attention to?

Source: BIT Securities

As of last Friday's U.S. stock night session close, the storage chip and AI hardware industry chain collectively surged against the trend, with SanDisk soaring 11.9% in a single day, and SK Hynix and Micron Technology rising over 8% and 6% respectively. After the close, S&P Dow Jones Indices announced quarterly component adjustments, with Bloom Energy, Illumina, and Everpure being added to the S&P 500 Index, while Dell, Palo Alto Networks, Arista Networks, and SanDisk were included in the S&P 100 Index. Notably, Bloom Energy's stock jumped over 5% after hours following a 7.35% increase during the day due to a significant surge in performance. NVIDIA's latest acquisition is reshaping storage capacity allocation, and the AI capital expenditure boom is spreading from individual stock performance to the core index weights of U.S. stocks. As the market opens, what information regarding AI, chips, and cybersecurity should we pay attention to?

1. AI Storage Power Fully Ignited: From NVIDIA's Acquisition to the Structural Shortage in Storage

The recent surge in storage chip stocks is not driven by a single favorable factor but is the result of three tightening threads.

First, the catalyst is NVIDIA's acquisition of Hugging Face, moving from computational power competition into the development process. On September 3, NVIDIA confirmed it would acquire the AI development platform Hugging Face for approximately $12.9 billion—this open-source model and dataset hosting community, often referred to as "the GitHub of AI," has gathered over 18 million developers and more than 200,000 companies using it to find, evaluate, customize, and deploy AI models. NVIDIA has already dominated the computational power market; its acquisition intention is not about the power itself but about getting involved in the entire process of AI "from selection to usage" earlier: in recent years, the industry has competed on model capabilities, but another race is accelerating—who can make these capabilities usable by more people and continuously generate revenue. Open models precisely bring the competition into the entire process of "what to do after finding the model," and Hugging Face is positioned at the intersection of validating effectiveness, integrating business, and deploying operations. What NVIDIA is truly acquiring is the community trust and developer habits that this platform has built over the long term and are difficult to replicate in the short term.

Open models lower the threshold for companies to use AI, and whether for training or deployment, these applications ultimately need to land on AI servers and data centers: they require NVIDIA's GPUs and high-bandwidth memory and large-capacity flash storage from companies like SanDisk to store model weights and data. That evening, SanDisk surged 11.9% to $1,740, with a cumulative increase of over 550% for the year; SK Hynix rose 8.14% on the U.S. stock market; Micron Technology increased by 6.1%; Intel rose 4.51%.

Second, the supply structure is being redefined by AI. A deeper driving factor is that the allocation of storage industry capacity is shifting towards AI: to meet the strong demand for high-margin products like high-bandwidth memory (HBM) from AI servers and data centers, major manufacturers like Samsung, SK Hynix, and Micron are continuously shifting capacity towards enterprise/AI server product lines, directly compressing the supply of consumer-grade storage—this is also why, even with weakening demand for PCs and smartphones, DRAM and NAND prices remain high. Dell's Chief Operating Officer previously stated that the current supply bottleneck "can be summed up in a few words: DRAM, DRAM, DRAM, followed by NAND, NAND, NAND." According to industry data, the quarter-on-quarter increase in DRAM and NAND prices in the second quarter was once close to 60%, and although there has been some cooling in the third quarter, they are still expected to rise by 13%—18% and 10%—15% respectively.

Third, the shortage is structural rather than a one-time cyclical fluctuation. New advanced storage wafer fabs typically take 3 to 5 years to form effective capacity. Institutions generally believe that the current storage shortage driven by AI is structural rather than cyclical, and it is difficult to quickly alleviate it through capacity expansion in the short term. Because of this, even with astonishing increases, SanDisk's current valuation is only about 8 times its forward P/E ratio, which, according to storage authorities, is still not considered expensive—this is also the underlying support that allows this transmission chain to ultimately reach the adjustments in the S&P index components: a real supply-demand gap, rather than just emotional capital speculation.

2. Bloom Energy's Surge: Performance Far Exceeds Expectations with Inclusion in the S&P 500 Index

On the individual stock level, the most eye-catching is the continuous rise of Bloom Energy (BE), a hydrogen fuel cell power generation company. The company recently announced that its second-quarter revenue reached $1.07 billion, a year-on-year increase of 166%; adjusted EBITDA reached $253.4 million, far exceeding Wall Street's expectation of $149.4 million. The strong demand for electricity from AI data centers: currently, all major hyperscale cloud service providers in the U.S., along with more than a dozen emerging cloud computing, AI labs, and data center hosting operators, have verified and adopted Bloom's fuel cell power generation solutions; data from the International Energy Agency shows that the electricity demand growth rate for AI data centers is as high as 50%, far exceeding the global overall electricity growth rate of about 3%.

On the day of the earnings release, Bloom Energy's stock price surged 7.35% during the day to $252.87, with a market capitalization of $74.48 billion, an increase of 506.8% compared to before; after the close, S&P Dow Jones Indices announced its inclusion in the S&P 500 Index, and the stock price jumped over 5% after hours to $266.14.

3. Full Analysis of the S&P Dow Jones Index Quarterly Adjustment: Chips, Cloud Hardware, and Cybersecurity Fully Positioned

After the close on September 4, S&P Dow Jones Indices announced the list of adjustments for the S&P 500 Index series for the third quarter. The newly added companies are almost entirely concentrated in the fields of chips, cloud hardware, and cybersecurity, with the index weights shifting from traditional consumer, real estate, and industrial stocks to the AI infrastructure industry chain, aided by passive fund reallocation.

New additions to the S&P 500: Bloom Energy, as a core incremental target, with a second-quarter revenue surge of 166% and an upward revision of the annual guidance; gene sequencing company Illumina, which was added after raising its annual revenue guidance in July, saw its stock rise about 1%—1.7% after the announcement; computer storage company Everpure also saw its stock rise about 1.5%—2.2% after exceeding expectations. The four new slots in the S&P 100 Index were all taken by technology companies in the AI capital expenditure industry chain—Dell's stock has risen about 316% this year, making it the largest gainer among the newly added stocks; Palo Alto Networks has increased by about 78% this year, outperforming the S&P 500 Index (which has risen about 12%) by 66 percentage points; Arista Networks has also seen an increase of over 40%, with its switch products being key infrastructure connecting AI data center GPU clusters; SanDisk is the stock with the most significant increase in the entire index system, with a cumulative increase of over 550% this year.

4. Worth Noting: Among the New Additions, These AI Chip and Cybersecurity Companies Are Worth Continuous Tracking

As passive funds begin to chase the AI capital expenditure chain, which companies in this new index addition list have fundamental stories that are still in the early stages and are worth continuous tracking?

Palo Alto Networks: After exceeding expectations in its fourth-quarter performance announced on September 1, Benchmark analyst Yi Fu Lee raised the target price from $340 to $400, citing that AI infrastructure construction is driving a simultaneous explosion in cybersecurity demand, with the company's Prisma Cloud cloud security platform and unified AI security platform Prisma AIRS performing strongly; however, its P/E ratio has reached about 173.6 times, and whether it can continue to deliver on the narrative of the "AI security stack" with subsequent performance is the most critical variable to verify.

Arista Networks: As a switch supplier connecting AI data center GPU clusters, the company has seen an increase of over 40% this year, with a market average target price of about $241, indicating about 24% upside potential from the current stock price, making it relatively undervalued among the newly added list.

Bloom Energy: The company's management cited industry data indicating that by 2027, the global AI data center's new installed capacity will reach 30—40 gigawatts, while Bloom currently has only about 1.5 gigawatts of deployed fuel cell capacity, meaning the growth space has just begun to open; combined with a $25 billion financing agreement with Brookfield and approximately $20 billion in hand orders (of which product orders are about $6 billion, a year-on-year increase of about 2.5 times), Wall Street's average target price is about $275, indicating some upside expectation from the latest stock price after inclusion in the S&P.

SanDisk: Despite a cumulative increase of over 550% this year, making it the stock with the most significant increase in the entire index system, its current forward P/E ratio of about 8 times is still significantly lower than the overall valuation center of the storage and AI hardware sector during the same period, making it one of the few companies in this round of market that has "huge increases but is not considered expensive."

Dell: As a representative of AI servers, the company has an AI order backlog of $95 billion, with an AI demand conversion scale of $131.7 billion over the past 12 months. The index inclusion itself has limited marginal impact; the real pricing anchor remains the pace of fulfillment of subsequent quarterly AI server orders.

Overall, these companies are distributed across different segments of the AI industry chain—cybersecurity, data center networks, power infrastructure, storage chips, and AI servers. The index inclusion is more of a "confirmation" of this fundamental market rather than a starting point. Whether they can continue to outperform will still depend on whether their quarterly performance can meet the current high expectations.

Disclaimer: This article is for reference only and does not constitute any investment advice or product offer. Data is as of the close of the U.S. stock night session on September 4, 2026 (Eastern Time), sourced from public information, and our company does not guarantee its accuracy or completeness. The article contains forward-looking statements, and actual results may differ significantly. Investment involves risks, and prices can rise or fall; past performance does not represent future performance, and investors may lose all principal. Product availability is subject to local laws and regulatory restrictions. Please evaluate independently and consult independent professional advice.