Memory chips have emerged as the driving force behind the semiconductor industry’s latest growth cycle as the artificial intelligence boom fuels demand for data-center hardware and pushes prices higher.
Memory has become the “king” of the semiconductor industry, according to Susquehanna analyst Mehdi Hosseini, who believes “its reign is here to stay.”
Investments in AI data centers have put semiconductor industry revenue on track to double to about $1.5 trillion this year, with memory chips accounting for much of the increase, Hosseini said in a Tuesday note to clients, citing Semiconductor Industry Association data on global shipments and average selling prices in July, MarketWatch reported.
Memory now represents between 50% and 55% of total semiconductor revenue, compared with its historical share of roughly 20% to 30%, according to Hosseini.
Memory prices to rise between 50-60% this quarter
The rapid construction of AI data centers has created an enormous appetite for memory and storage as companies deploy increasingly powerful systems capable of processing large volumes of data.
That demand has contributed to severe supply shortages and a sharp increase in memory prices.
While average semiconductor prices are already moving higher, Hosseini said the increase would be “more moderate” if memory were excluded from the calculation.
The analyst expects average prices for dynamic random-access memory, or DRAM, to rise 50% sequentially during the current quarter, followed by another 20% increase in the fourth quarter of 2026.
NAND flash memory prices are expected to climb even more sharply in the near term, with Hosseini forecasting a 60% sequential increase this quarter and a further 25% rise in the following quarter.
The price increases are providing a significant earnings tailwind for memory manufacturers, which have spent much of the previous semiconductor cycle dealing with volatile demand and periods of oversupply.
Goldman Sachs says Micron, SanDisk stocks breaking out of downtrends
Micron Technology has been one of the strongest-performing stocks in the S&P 500 this year as investors have increasingly priced in the benefits of the memory shortage.
Micron shares have risen about 250% so far this year. SanDisk, which specializes in NAND flash memory, has performed even more strongly, with its stock up 531%.
Hosseini has positive ratings on both companies.
Meanwhile, Goldman Sachs has identified early signs that investor interest in the memory trade is returning after a quieter summer.
The firm said Micron and SanDisk are beginning to break out of downtrends that had constrained their shares during the summer.
The setup comes as positioning in the broader AI trade remains relatively subdued. Goldman’s prime brokerage data showed gross leverage among US fundamental long/short hedge funds at the 27th percentile of the past year, while net leverage stood at just the fourth percentile.
Options activity has also eased, with the CBOE Semiconductor ETF Volatility Index falling to around 36 from roughly 65 in July, pointing to a significant decline in expected swings across semiconductor stocks.
That combination could leave room for investors who reduced their exposure during the summer to rebuild positions if the strength in AI-related memory demand persists.
High-bandwidth memory adds to supply pressure
One of the most important areas of the memory market is high-bandwidth memory, or HBM, which is increasingly critical for AI accelerators.
Earlier this month, South Korean news outlet Electronic Times reported that Micron was targeting a doubling of its HBM production capacity by the end of the year.
HBM consists of multiple layers of DRAM stacked together and is designed to provide the high-speed data access required by AI processors.
Nvidia is among the major chip companies relying on HBM to handle increasingly intensive AI workloads.
The production process is more silicon-intensive than conventional memory manufacturing, meaning the rapid expansion of HBM capacity can also constrain supplies of traditional DRAM.
That dynamic has contributed to the broader tightening of the memory market.
Logic chips face a period of stabilization
Memory is not the only part of the semiconductor industry benefiting from higher prices.
Hosseini said logic chips, including central processing units, have also seen average selling prices rise.
However, he noted that “recent trends suggest a period of stabilization.”
He linked some of the moderation to delays in the ramp-up of Nvidia’s next-generation Rubin AI platform, which includes Vera central processing units and Rubin graphics processing units.
The timing of new AI hardware remains important for semiconductor companies because each generation can require significant increases in memory capacity and bandwidth.
At the same time, Hosseini argued that higher upfront hardware costs do not necessarily undermine the longer-term economics of AI.
Future generations of AI hardware will carry “a higher upfront cost,” he said, but falling prices for tokens — the units of data processed by AI models — combined with rising token production should support “the economic value proposition of AI at scale.”
Earnings will test the memory rally
Despite the strength of the memory trade, investors still face a familiar problem: the industry remains highly cyclical.
Memory companies can benefit rapidly when demand exceeds supply, but production increases can eventually reverse the balance and put pressure on prices and margins.
Consumer spending and broader electronics demand can also shift quickly.
That makes the recent technical strength in Micron and SanDisk only part of the story.
For Micron, the next major test will come with its fiscal fourth-quarter results on Sept. 30.
Investors will be looking for evidence that the surge in memory prices is translating into sustained revenue and profit growth rather than simply a temporary improvement in the industry’s cycle.
For now, however, the fundamental backdrop remains favorable.
The post Memory prices could jump 60%: why Micron and SanDisk are breaking out again appeared first on Invezz
