But it wouldn’t be long before a very real threat to computers surfaced: that which has since been dubbed the capacitor plague.
Such effusive language is sure to frighten even the sternest of PC gamers, though this isn’t just some tall tale told to scare budding builders into cleaning their chassis every once in a while. There really was an event by such a name in the early 2000s, though unlike Y2K’s unwarranted panic, the capacitor plague would genuinely threaten catastrophic consequences for millions of PCs.
The events leading up to the so-called capacitor plague may have begun as early as 2001, where capacitors that were used in a range of electrical appliances and equipment—most famously of all, Dell PCs—were said to be fundamentally flawed.
I don’t imagine many PC gamers today will have any memory of the capacitor plague. I certainly didn’t until 2018. In a strange turn of events, I first heard of the event during a visit to Dell’s testing facilities in Texas. I wasn’t there to report on capacitor failure rates, yet during a test facility tour those fated capacitors were mentioned, serving as prime example of why Dell now doubles down on ensuring its suppliers’ products match its own standards.
Dell certainly hasn’t forgotten about the capacitor plague—I suppose one brush-in with mass component failure will do that to a company. Yet for me it became a bit more of a techie morbid fascination: a tech plague sounds like something out of a Warhammer 40,000 novel.
The electrolyte within these doomed caps was prone to deteriorating at a rapid rate.
It should come as no surprise that this all stems from one of the smallest discrete electrical components inside your PC. Storing a small amount of electrical charge, a capacitor, or cap, is a key component in any piece of electrical equipment, and there’s no escaping the hundreds of them required for your gaming PC to function. Most among us would probably recognise the coiled electrolytic cap—made up of a dielectric material separating metal plates wound tightly together within a single insulated cell. Though there are also capacitors made from ceramic, polystyrene, paper, film, and a whole lot else.
You’ll absolutely recognise the ceramic cap as those littering the space around or overleaf from your GPU or CPU. Though the caps we’re talking about today are the coiled electrolytic kind—and pretty dodgy ones at that—many of which would be fitted in millions of PCs and electronic devices for the first half decade of the 2000s.
With capacitors today being produced in their trillions annually, and respectively in very high demand of the 2000s, there have to be certain accepted rates of failure. However, this issue was much, much worse than expected, and perhaps even endemic to an industry.
The electrolyte within these doomed caps was prone to deteriorating at a rapid rate, meaning the cap could swell or even burst as the electrolyte changed from a liquid into a gas. Once a cap fails like this, it is no longer building charge and discharging, which will lead to subsequent failures of the attached component, such as a PC motherboard.
But it’s not just that dodgy capacitors were being manufactured and installed into heaps of machines. The wildest part is how that is alleged to have come to pass. According to one report by The Independent published in 2003 (no longer available on its website but accessible via the Wayback Machine archive), the reason for these particular capacitors’ tendency to fail was down to an act of industrial espionage.
“A scientist steals a secret formula for an electrical product from his Japanese employer and takes it to China. Then it is stolen again and turns up in Taiwan. But something goes wrong—and thousands, perhaps millions, of computers and electrical goods in the West begin to burn out or explode.”
The Japanese employer said to have had its secret blend of electrolytes stolen is a company called Rubycon, which was at the bleeding edge for electrolytic capacitors at the time of the alleged theft. The company that ultimately made off with the supposedly functional formula is said to have got something wrong, however, which led to the capacitors tendency to, uh, burst.
“After a few hours of operation, the electrolyte would leak hydrogen gas, before bursting the metal body of the capacitor. The electrolyte would then leak its brownish filling and could cause a fire.”
That’s the popular theory as to how these caps came to be manufactured and why they were flawed, though admittedly the exact details are muddier than I’d prefer. The Independent report is most often the only source for the matter of corporate espionage, and even that article no longer exists on The Independent’s website—scrubbed from the internet over time. A recount of the story was published in The Guardian by the same journalist, Charles Arthur, in 2010, however.
I also can’t find any reports of fires being started this way, but at the very least I’d guess a few affected capacitors popped in a moderately dramatic style.
Though there are other reports of the issue as it came to be. A marketing manager from at least one company affected by the failing caps told Passive Component Industry Magazine in its Nov/Dec 2002 issue (PDF warning) that it had sourced caps from Taiwan that were affected by the issue. Though they failed to mention any specific company was responsible. The same report goes on to cite the owner of a PC building business who states just how widespread the issue may have been at one point.
“ABIT PC motherboards were the most problematic, although he has seen several other Taiwanese-made motherboards with seeping capacitors, including ASUSTek Computer Inc., Gigabyte Technology Co. Ltd., PC Chips Manufacturing Ltd., and Microstar International Co. Ltd.”
You’ll likely recognise many of those same brands making PC gaming gear today. Though this was a long time ago—you’d have to be dusting off some very old gaming gear to get anywhere close to these capacitors today, and at this point they’re most likely dead anyways.
Bad caps were causing a very real and widespread issue. Most of all, or so it appeared, for US PC goliath Dell.
Court documents from a lawsuit between Dell and Advanced Internet Technologies, reported on by the New York Times (paywall) in 2010, would reveal the full extent of the issue for the company. Dell was sued for high rates of failure in the computers AIT leased from the company in 2003 to 2004, but the two companies would later settle out of court.
“A judge in the Federal District Court in North Carolina unsealed hundreds of documents linked to a lawsuit filed by Advanced Internet Technologies that had accused Dell of trying to hide defects in its desktop computers from customers,” the NYT report said.
Defect rates for Dell machines at the time were noted in the court documents to be as high as 20.2% of machines in a batch of 5,000 purchased by the city of New York. Microsoft had an issue with 11% of its machines during the same period, and other companies are noted with similar failure rates.
A document noted by the NYT expected a minimum of 12% of Dell’s SX270 Optiplex computers to result in an “incident report” over the course of three years, though later revised that number to 45% and noted it could get as high as 97%.
Crikey.
A company spokesperson later said that Dell’s actual failure rates were much lower than those it forecast, confirming it repaired 22% of the 21 million Optiplex computers it shipped during 2003 to 2005. That’s still 4.62 million PCs with potentially faulty capacitors, and a lot of repair work for perhaps a handful of dodgy caps in any given machine.
This was an industry-wide problem.
Dell
Dell pointed to capacitors manufactured by Japanese company Nichicon as the primary cause for concern in its products. The company mentioned Nichicon by name quite a lot in a blog post published in 2010, in fact.
“As noted in a New York Times article about the lawsuit, faulty capacitors were manufactured by Nichicon, a respected, long-term supplier to many industries. These capacitors were used by Dell suppliers at certain times from 2003 to 2005. The faulty Nichicon capacitors affected many manufacturers, including Hewlett-Packard, Apple and others, as discussed in the initial story and several blog posts afterward. Again, this was an industry-wide problem,” Dell said.
“Dell suspended use of Nichicon capacitors after we discovered a problem in its manufacturing process. As we routinely do with product issues, we actively investigated the failures, audited the Nichicon plants and worked with customers to fix OptiPlex computers on a case-by-case basis. Beyond that, Dell voluntarily extended the warranties on all potentially affected OptiPlex motherboards up to January 2008 to address the Nichicon capacitor problem. The capacitor failure rates varied depending on customers’ environments and the number of Nichicon capacitors in the customer’s motherboards.”
This is where things get a little confusing. Some attribute the Nichicon issue as a separate capacitor issue, unrelated to the aforementioned Taiwanese caps, yet occuring at near enough the same time. There was certainly a large level of overlap between these incidents, whether related or not, and that is probably justification enough to constitute that entire period as the so-called capacitor plague, anyways.
What’s clear is that the entire period would be disastrous for the lifespan of electronics built during it. Though not everyone was running scared. If you knew what you were doing, there was also an opportunity to buy up mountains of faulty hardware, replace the caps, and get some good use out of them or sell them on for profit. There were even bespoke repair businesses set up during this time to help with this process, such as badcaps.net, which is still going today.
Thankfully, if we think back over the previous decades since the capacitor plague, no one has suffered from such an issue on anywhere near a scale like some did at that time. At least none we know about yet, anyways. Though there have been a few run-ins with other sorts of capacitor-related issues. Capacitor supply did take a big knock during recent shortages, which only worsened the chip deficit for PC gamers and further afield, and you might also remember the MLCC versus POS-CAP (SP-CAP) capacitor debacle from 2020 involving Nvidia’s RTX 3080 graphics card. That was later mitigated to some extent by a new driver version, however, so you couldn’t pinpoint capacitors exclusively at fault there.
Perhaps PC gamers are never too far from a dodgy capacitor or two, but let’s hope a plague the scale of the one seen in the early 2000s won’t come our way ever again.
