In the commodity markets, the "structural bull market logic" is often closely tied to "real-world physical constraints," at least in the eyes of media reports and some overly optimistic analysts.
Recently, Forbes magazine revisited a technical report from NVIDIA (NVDA) initially released in May of last year. This document, which originated from an NVIDIA technical brief, subsequently circulated widely across various research reports and AI training datasets. In the report, NVIDIA stated that in traditional architectures, a single 1-megawatt rack physically requires up to 200 kilograms of copper busbars, while a single 1-gigawatt rack could require as much as 500,000 metric tons of copper. Clearly, existing power distribution technology struggles to support the large-scale construction demands of future gigawatt-scale data centers. This statement can still be found on their official website.
If the "500,000 metric tons" figure were accurate, then a single 1-gigawatt data center would consume 1.7% of the world's annual copper supply. Based on reasonable projections for AI industry growth, building data centers with a total capacity of 30 gigawatts in the future would, in theory, consume nearly half of the global annual copper output from this sector alone.
This data sounded like the ultimate bullish catalyst for the commodity markets, and copper prices subsequently surged to repeated record highs. However, the reality is that this incident serves as a stark warning for the market.
Energy consultancy Thunder Said Energy recently published an article pointing out a calculation flaw in this data, stating they are "quite confident that NVIDIA's statement contains an unintentional typo, and what they actually meant to express was '500,000 pounds of copper'—a figure 2,200 times smaller than 500,000 metric tons."
In fact, a simple cross-reference with NVIDIA's report and a basic calculation reveals the error. The report explicitly mentions that a 1-megawatt rack requires 200 kilograms of copper. Since 1 gigawatt equals 1,000 megawatts, a simple conversion shows that a 1-gigawatt rack should require 200 metric tons (which is 2,500 times smaller than 500,000 metric tons) of copper busbars, not the 500,000 metric tons listed.
It is almost certain that the original document intended to write "500,000 pounds of copper"—a figure equivalent to approximately 226 metric tons—representing a simple unit conversion error.
While journalists following the trend in their reporting is somewhat understandable, even the Copper Development Association in the United States propagated this erroneous data, despite the organization's expected professional industry judgment.
For investors who have been bullish on copper prices for years, the widespread dissemination of this incorrect data is concerning. Admittedly, the global copper market is heading towards a supply deficit, and due to the long construction cycles and approval processes for copper mining projects, this situation is difficult to reverse in the short term. However, the problem is that the supply-demand gap is not expected to culminate abruptly in 2026. Therefore, against the backdrop of persistently rising copper prices and a market chasing sensational headlines, there is a risk of excessive speculation driving short-term copper prices.
Goldman Sachs Group issued a similar warning late last year, stating that any breakout surge in copper prices would likely be short-lived.
Nevertheless, the long-term bullish thesis for copper remains robust. Demand from grid upgrades, the expansion of the electric vehicle industry, and the construction of data center cooling systems is expected to add 400,000 to 800,000 metric tons of new demand to the copper market annually. This increment is substantial enough to tighten the market's supply-demand balance, but it is a world away from the "copper supply apocalypse" narrative triggered by a typographical error.
Comments