Three of the world's largest memory chip manufacturers are facing a federal antitrust lawsuit accusing them of illegally coordinating to restrict DRAM supply and drive prices to record highs, with plaintiffs claiming the companies used the industry-wide pivot toward high-bandwidth memory as cover for cutting output of widely used commodity modules.
The Lawsuit and Its Claims
Samsung, SK Hynix, and Micron were sued on June 25th in the U.S. District Court for the Northern District of California. The case, filed as Garciaguirre v. Samsung Electronics and assigned to Judge Noel Wise, was brought by 17 plaintiffs who allege the three companies illegally coordinated to restrict DRAM supply and inflate prices.
The complaint invokes Section 1 of the Sherman Act and has been structured as a class action. According to the complaint, DRAM prices have risen roughly 700% over four years.
The three defendants collectively hold around 90% of the global DRAM market, a concentration that the plaintiffs argue gave the companies both the motive and the means to manipulate prices without fear of competitive pressure from outside the existing oligopoly.
READ NEXT: VivoPower Names Unnamed Global AI Leader as Preferred Tenant for Data Center in Norway, Europe
HBM Transition Used as Cover, Plaintiffs Argue
At the center of the complaint is a specific theory about how the alleged price fixing was executed. The plaintiffs argue that Samsung, SK Hynix, and Micron used their coordinated shift toward high-bandwidth memory, the stacked DRAM architecture used to feed AI accelerators, as a pretext to curtail production of older DDR3 and DDR4 modules. That contraction in commodity DRAM supply, the suit contends, pushed prices to record highs.
The complaint further argues that the market structure made it nearly impossible for any new competitor to step in and fill the gap. Building a new DRAM fabrication facility costs tens of billions of dollars and takes years to complete, leaving the three incumbents free to cut output without facing the risk of being undercut by a new entrant. That structural barrier, the plaintiffs argue, made the alleged coordination particularly effective.
Trusted by Leading EPCs & Manufacturers
Find the Latest Data Center Projects Around the World
Gain exclusive access to our industry-leading database of data center opportunities with detailed project timelines and stakeholder information
Request Free Trial → Learn More →
No credit card Up-to-date coverage
Who Filed the Suit
The 17 named plaintiffs include 14 individuals and three small PC businesses. Among the business plaintiffs are Troy's Computers and a Florida repair operation called My Florida PC.
The plaintiffs are seeking class certification, an injunction against the alleged conduct, and treble damages, which under antitrust law can amount to three times the actual damages proven.
The plaintiffs also point to Apple's recent price increases on its iPad and Mac product lines as evidence of the downstream impact of the alleged DRAM supply squeeze, framing those consumer-facing cost increases as a consequence of the conduct attributed to the three chipmakers.
Why This Matters for the Data Center Industry
Although the lawsuit is attracting attention because of the AI boom, it is important to understand that the case is not about AI data center memory pricing.
Modern AI data centers rely on High Bandwidth Memory (HBM), a specialized memory technology integrated with AI accelerators such as NVIDIA's H100, H200, and Blackwell GPUs, as well as AMD's Instinct accelerators.
HBM delivers the massive memory bandwidth required for AI training and inference and has become one of the fastest-growing segments of the semiconductor industry.
The lawsuit, however, focuses on conventional DRAM, including DDR3 and DDR4 memory, which is commonly used in personal computers, laptops, enterprise servers, networking equipment, industrial systems, and many non-AI data center workloads.
Consumer DRAM and AI Data Center Memory Markets
One of the key distinctions highlighted by the case is that the consumer DRAM market and the AI data center memory market operate as separate segments.
Traditional DRAM products such as DDR3, DDR4, and DDR5 are designed for general-purpose computing and conventional servers. These products are sold in high volumes across consumer electronics, enterprise IT, and cloud infrastructure.
By contrast, AI data centers primarily depend on HBM, a premium memory technology manufactured using advanced packaging techniques and supplied to hyperscalers and AI infrastructure providers building GPU clusters for large language models and other AI workloads.
Prior Litigation in the Same Court
The current case is not the first time similar allegations have been leveled against DRAM manufacturers in the same courthouse. A 2018 class action brought by law firm Hagens Berman made comparable claims about DRAM pricing conduct, establishing that the Northern District of California has previously been the venue for antitrust challenges targeting the memory chip industry. The new lawsuit revisits that precedent as plaintiffs attempt to build on the legal foundation laid by earlier litigation.
Market Concentration at the Core of the Complaint
The lawsuit places significant emphasis on the degree to which the global DRAM market is dominated by the three defendants. With Samsung, SK Hynix, and Micron together accounting for roughly 90% of global DRAM supply, the complaint argues that the market structure itself creates conditions under which coordination, whether explicit or tacit, can have outsized effects on pricing across the entire industry.
The inclusion of small PC repair businesses among the plaintiffs reflects the breadth of the alleged impact, with the complaint suggesting that the pricing effects reach not just large technology companies but also small enterprises dependent on commodity memory components for their day-to-day operations.
Winning Data Center Contracts Around the World Starts With Knowing What's Next
How many opportunities in the data center sector have slipped past your team simply because the intelligence arrived too late? In a market defined by precision timing and high-stakes capital decisions, delayed data is as costly as no data at all.
The Global Project Tracking (GPT) platform by Blackridge Research was built for exactly this environment, aggregating global data center project activity into a single, continuously updated intelligence source that keeps your pipeline full and your proposals sharp.
From hyperscale facilities breaking ground to data center expansions taking shape around the world, the platform surfaces the projects that matter before your competitors have a chance to act on them.
Upcoming Projects
Tender Notices
Contract Awards
Projects Under Construction
Completed Projects
See the full depth of data center project intelligence available across the World for yourself. Contact the Blackridge Research team today to Book a Free Demo of the Global Project Tracking (GPT) platform.
Leave a Comment
We love hearing from our readers and value your feedback. If you have any questions or comments about our content, feel free to leave a comment below.
We read every comment and do our best to respond to them all.