News
- IVCM petition asked Imperial County judge to overturn IID's refusal to sell it roughly 287 million gallons to cool a proposed 330 MW data center
- The company pledged in writing as recently as February that the project would not touch Colorado River water
- The case turns on whether fallowed farmland water can be reassigned to industry, a precedent worth far more than the volume in dispute
A recent petition filed in Imperial County Superior Court has asked a judge to order a public utility to start selling water to the largest data center project in the state of California.
The volume at issue is modest by the standards of the Colorado River: 880 acre-feet a year, which the petition itself calculates at ~0.03 percent of the Imperial Irrigation District's 3.1 million acre-foot entitlement.
The legal precedent that it would set if such a motion was to be granted might however have far-reaching implications beyond the relatively minuscule requirement that Imperial Valley Computer Manufacturing (IVCM) is currently gunning for.
A prolonged dispute over water that is exacerbated by who the customer isThe Imperial Irrigation District, a local agency that supplies water from the Colorado River in Imperial Valley, has declined IVCM's request to provide approximately 287 million gallons of water for its upcoming 330MW data center, the largest in the state.
The developer, Sebastian Rucci, spoke to Business Insider, stating that the project would not add to demands on the Colorado River as it would effectively stop irrigating nearby farmland to balance its consumption, calling it a "zero impact" situation.
IID rejected the application on May 1, 2026, nine days after it was submitted. The stated ground was Regulation 21, which governs small-parcel service and bars new connections within 300 feet of an accessible potable water supply; the district redirected IVCM to the City of Imperial.
With the City of Imperial already locked in a legal battle with IVCM over the very existence of the $10 billion project, citing inadequate public notice and lack of compliance with the California Environmental Quality Act (CEQA), it is unlikely to be a place where the data center builder is going to find any relief, and it has turned to courts of law to get what it feels is its fair share.
(Image credit: Microsoft)The subsequent legal action by IVCM might be a litmus test for how such interactions could unfold in the future, at a time when there is considerable blowback from communities living near such data centers, who see them as resource-greedy and driving up water and power prices, especially in drought-stricken regions.
The developer's plan to "buy and dry", as per Michael Cohen, a senior fellow at the Pacific Institute focusing on Colorado River Basin water use, might actually have made matters worse, with it being seen as detrimental to jobs in the area even as individual landowners profit from the exercise.
The water suit is one of at least three fronts, and arguably the least immediately consequential.
The City of Imperial's CEQA challenge to the project's exemption is pending. On June 16 2026, the county imposed a 45-day moratorium on data center approvals; on July 14 it extended that to a full year, blocking permits until June 2027 while an advisory committee rewrites zoning rules.
Rucci called the first moratorium defective and sought a restraining order against it, and has said he will challenge the second. Even a clean win on the water petition would result in supply permission for a facility the county currently cannot permit.
What is being litigated is whether an irrigation district chartered to serve farms can lawfully decline to serve an industry, and whether fallowing counts as conservation when the county is the buyer but not when a data center is.
One thing is for certain: the ruling will be read closely by every developer eyeing the West's agricultural water, which is roughly what the valley is afraid of. The gallons are a rounding error on the Colorado, but the precedent it sets here may determine everything for the region.
- AMD and Cerebras will split inference across two machines, with Helios racks handling prompt processing and the Wafer-Scale Engine generating tokens, available through Cerebras Cloud in H2 2026
- Nvidia is also doing something similar by licensing AI chip startup Groq's SRAM decode technology for $20 billion
- The move sees AMD and Cerebras claim 5x higher tokens per watt versus a standalone Cerebras WSE configuration
AMD and Cerebras Systems have announced a technical partnership which pairs the former's Helios rackscale system with the latter's Wafer-Scale Engine in what both companies call a disaggregated inference solution.
The move has enabled a combined AMD Helios and Cerebras WSE configuration to deliver up to five times the tokens per second per watt (TPS/W) in internal testing by both chip designers.
The move aims to address a Cerebras WSE efficiency challenge by offloading prompt processing to AMD's rackscale offering.
An efficiency gains-centric play?Both AMD and Cerebras Systems are painting the news as a win, and it very well might be, given the latter's efficiency gains in play and the former's ability to get access to SRAM decode technology without spending the $20 billion Nvidia shelled out at the end of last year for a non-exclusive deal.
It must, however, be noted that the efficiency claims of 5 tokens per second per watt are compared against an existing Cerebras WSE (Wafer-Scale Engine) as the baseline, while running the open-source Kimi 2.6 1T model, making them impressive, but without a direct comparison to figures for an Nvidia rackscale offering, one that lacks context, especially when efficiency is the metric.
The idea itself is sound and well established in the industry, with WSE known to struggle with the 'prefill' part of the equation while handling the 'decode' segment relatively well, essentially substituting AMD's hardware where Cerebras' equipment falls short.
The choice of Kimi 2.6, however, deserves a second look. Moonshot AI's model, released on 20 April 2026, is a mixture-of-experts design with one trillion total parameters but only 32 billion active per token, and it ships natively in INT4. At INT4, the full weight set runs to roughly 500 GB. A single Cerebras wafer holds 44 GB. Even before KV cache, a Cerebras-only deployment needs somewhere north of a dozen wafers just to hold the model, while one Helios rack could hold it around sixty times over.
That asymmetry means the five-times figure is measured on a model that is close to the least favorable for a WSE-only configuration. A dense model small enough to sit resident on a handful of wafers could flatter Cerebras considerably more. None of this makes the number wrong, but it does make the case for additional testing to demonstrate both its strengths and weaknesses for different models.
A partnership without numbers, for nowMore importantly, the absence of any financial information might very well be a future story, especially at a time when there are increasing concerns about 'circular financing' in an industry where Nvidia's recent move to backstop OpenAI's data center purchases was seen as a net negative by Wall St, which is already concerned about AI spend and the sustainability of such transactions.
AMD has also, in the past (and more recently with Anthropic), linked purchases of its own hardware to investments or stakes it would take in AI companies, moves that the market welcomed earlier but might view with a bit more hostility lately.
The announcement comes at a time when Cerebras might need it more than AMD: Cerebras listed on Nasdaq in May, priced at $185, opened at $350, and closed its first day at $311.07 before falling back to around $227 by late June 2026.
AMD stock, on the other hand, is up 121.48% year-to-date (YTD) as investors continue to bet heavily on its new Instinct AI processors, and the Cerebras partnership allows it to further consolidate its gains, as this might be seen as another vote of confidence in its current direction by one of its prospective customers.