Bumpgate

Bumpgate, also known as Nvidiagate, was a scandal where Nvidia and ATI Graphics Processing Units (GPUs) were prone to high failure rates due to a design flaw that led to cracked solder bumps under the die. Despite the "Nvidiagate" name, this defect not only affected many Nvidia GPUs made from approximately 2006 to 2010, but it also affected ATI GPUs from 2006 to 2008. Among retro console enthusiasts, the defect is best known to have likely been the culprit behind the high failure rate of Nvidia GPUs in Sony's early PlayStation 3 models and ATI GPUs in Microsoft's early Xbox 360 models.

Background

Graphics Processing Unit (GPU) Faults

From approximately 2005-2010, GPU manufacturers Nvidia and ATI developed some GPUs that had a serious design flaw. This flaw led to failures in many of their GPUs during that time period, and Nvidia even saw a lawsuit from it. In order to understand what truly happened during this controversy, though, it's important to understand what exactly led to the faults.

These manufacturers had engineered GPUs that electrically connected the silicon chip (die) to the substrate (the "green square part") using high-lead solder bumps. High-lead solder bumps were chosen in order to fit the power delivery specifications that these GPUs needed. To give these solder bumps more strength (especially when operating at high temperatures), it's standard to use an epoxy with silica filler known as underfill. Underfill needs to fit certain specifications, depending on how hot the processor it's used on is expected to get. If it's too hard, the underfill will crack the die. If it's too soft, it'll crack the bumps. It also needs to still fit within the right specifications at both high and low temperatures.

In 2006, IBM and Amkor published a study that explained that use of a low T_g underfill was not acceptable with high-lead solder bumps, and high T_g would be necessary to avoid defects. Therefore, because Nvidia and ATI chose to use high-lead solder bumps, they needed a high T_g underfill. However, this study was not out at the time that GPUs from 2005 were made, and the companies ended up using low T_g underfill in these processors.

This low T_g underfill would become too soft at high, but normal operating temperatures for these GPUs. When the processor went through normal thermal changes, the solder bumps would soften under heat and harden as they cooled. This would happen over and over again until they cracked under the thermal stress. When enough solder bumps cracked, it would cause a failure in the unit, hence the term "Bumpgate".

Companies involved and responses

Bumpgate was a worldwide issue for any consumer that purchased these defective graphics processors. However, it's not fully clear who was the most responsible for the incident. Building a GPU is a specialized process that requires it to go through multiple partner companies before ultimately ending up with the company that sells the unit (i.e., Microsoft, Sony, etc.). With this in mind, it is still important to note the response of each of the companies that were the public faces involved in this incident.

Microsoft's Response - The "Red Ring of Death"

A defective 90nm "Xenos" GPU from an early Xbox 360 (Xenon revision). Note the large die in the center of the chip, and the dark greenish underfill. The color of the underfill is one way to tell if the chip has the low T_g underfill.

"The breakthrough came when we understood that the connections that were being broken were not located on the motherboard, but they were actually located inside the components." - Leo Del Castillo, member of Xbox's hardware engineering group; Power On: The Story of Xbox (2021)

In November 2005, Microsoft released the Xbox 360 with the ATI Xenos GPU. According to leaked internal Microsoft documents from the time, 1.2 million 360s shipped out to consumers by November 30th, 2005. 3% of customers had some type of issue with their system. Out of that 3%, 19% had three flashing red rings, and 24% had freezing problems.

Therefore, 43% of consumers shortly after launch had these familiar symptoms to retro console enthusiasts: the "Red Ring of Death"- which is what consumers around the world called the Xbox 360's Bumpgate-related fault. However, "three flashing red rings" on the Xbox 360 simply means "core digital error"- in other words, a general hardware failure. As it takes multiple power-on cycles to cause a failure in the solder bumps, it is unlikely that these systems were failing specifically because of the Bumpgate defect. 56% of systems worked on the first try, and after component reworks, that number improved to 71%. Only 200,000 consoles went to what Microsoft termed "the bonepile", a collection of broken consoles that did not work after a component rework. However, it is also worth noting that some afflicted Xbox 360s will work again after a "reflow"- which is effectively redoing all of the components on the motherboard at once. In addition, Microsoft wasn't aware that the "Red Ring of Death" was a specific widespread problem yet, so they would not have been taking major note of consoles with a GPU problem. At the height of the Bumpgate-related defect, approximately 600,000 to one million Xbox 360s were suffering from the "Red Ring of Death".

A non-defective 65nm "Xenos" GPU from a later revision of Xbox 360 (Jasper V2 on "Kronos 1" package). Note the smaller die and the light, whitish underfill.

For several months, Microsoft had consumers pay to repair their consoles. However, in 2007, they chose to extend the warranty for Xbox 360 consoles displaying an E74 error (an on-screen error also associated with "Red Ring of Death") from one year from date of purchase, to three years from date of purchase. This allowed most consumers who had consoles with the issue to get their consoles refurbished. In addition, they refunded any consumers who had paid to have their afflicted systems repaired.

Microsoft primarily seemed to have done this to protect the Xbox brand. The Xbox 360 was only their second console, and the original Xbox hadn't done as well as they'd hoped it would. Microsoft confirmed this in Chapter 5 of their 2021 documentary, Power On: The Story of Xbox.

By 2009, the cause of the issue was confirmed to be "within the components"- the low T_g underfill. ATI and Microsoft completely fixed the issue in Xbox 360s made after this point.

Sony's Response - The "Yellow Light of Death"

"We think it is highly unfair to suggest that from an installed base of 2.5 million that the numbers you mention somehow are evidence of a 'manufacturing defect'..." - Ray Maguire, managing director and senior vice president of Sony Computer Entertainment Europe in 2009; in a letter responding to a BBC Watchdog segment covering the "Yellow Light of Death".

In November 2006, Sony released the PlayStation 3. Like the Xbox 360, the early models of PlayStation 3 had what consumers believe to be Bumpgate-related issues with their Nvidia RSX GPU. Many consumers who had bought launch models- what are now known as "PlayStation 2 backwards compatible" models- found that their systems were malfunctioning. PlayStation 3 systems that were exhibiting the issue may freeze suddenly during gameplay and shut off, but all users who dealt with the defect had a common experience: when turning their PlayStation 3 on, the LED power indicator would turn green for a moment, then the system would beep three times, very briefly flash to a yellow light, then it would blink red. This was coined by consumers as the "Yellow Light of Death".

BBC Watchdog aired a segment in 2009 covering the "Yellow Light of Death". The hosts stated that when viewers called Sony because their console was malfunctioning, Sony told them that it could have been for a variety of reasons, and they couldn't determine exactly what the problem was without disassembling the console having the issue. While all of this is true- the LED indicators do simply indicate a general hardware failure that requires detailed troubleshooting- it is worth noting that by August 2008, Sony was building consoles with the 65nm RSX, which has been confirmed by enthusiasts to not suffer from the likely Bumpgate-related fault.

Sony never issued a recall, or extended warranty for these systems; consumers had to pay out of their own pocket to get them repaired. According to one of the individuals interviewed in BBC's Watchdog segment, if the system was outside of its one year warranty period, they would have to pay £128 (~$173) to get the system fixed, and they would only provide customers with a three-month post-repair warranty, and if it failed again after that point, they would have to pay out of pocket again. Keeping in mind that the fault was very likely caused by a GPU defect, this possibility was very likely unless Sony's repair technicians replaced the defective 90nm GPU with a non-defective one, which they were able to do for some consoles.

Instead of admitting that there was a defect and issuing an extended warranty to consumers (like Microsoft did) Sony simply quietly fixed the defect around the same time that Microsoft did for the Xbox 360, so PS3 "slim" revisions and newer were not affected. Specifically, it was only 90nm GPUs that appeared to have low T_g underfill- and hence, the defect.

As of 2025, Sony has still never made a statement confirming that "Yellow Light of Death", their likely Bumpgate-related issue, was a widespread problem in early PlayStation 3 models. The only reason that consumers know that the RSX was likely afflicted by Bumpgate is thanks to console repair and modding enthusiasts- who had to do extensive research to find out the true cause of the problem. It was clear from the timeline- what systems have been observed to be reliable or unreliable revisions- as well as tests performed by console repair and modding enthusiasts (such as "poking" the underfill on suspected bad GPUs with a soldering iron) that it was a Bumpgate-related defect.

Nvidia's Response

[Discuss initial response. Briefly touch on the lawsuit; there's a dedicated section below]

Lawsuit(s)

If applicable, add any information regarding litigation around the incident here.

Claims

Main claims of the suit.

Rebuttal

The response of the company or counterclaims.

Outcome

The outcome of the suit, if any.

---

Add your text below this box. Once this section is complete, delete this box by clicking on it and pressing backspace.

Consumer response

Summary and key issues of prevailing sentiment from the consumers and commentators that can be documented via articles, emails to support, reviews and forum posts.

---

Add your text below this box. Once this section is complete, delete this box by clicking on it and pressing backspace.

[General Consumer Response (frustration at/satisfaction with Microsoft, Sony, etc.)]

Xbox 360/Microsoft Consumer response

At first, consumers were angry at Microsoft. During the months that Microsoft would not acknowledge the widespread "Red Ring of Death", consumers felt as if they'd been scammed and made a bad investment. However, when Microsoft extended the warranty on Xbox 360s experiencing an E74 error to three years after purchase, consumer sentiment improved. After Microsoft's explanation and confirmation of the exact issue that caused the "Red Ring of Death" in Power On: The Story of Xbox, many consumers who remembered dealing with this issue felt further validated. The release of this information also helped enthusiasts learn more about how to fix the issue not only in early Xbox 360s, but even in Sony's early PlayStation 3 systems.

Misdiagnosis/Poor Repair of Faults

There was a lot of speculation among affected consumers as to why so many GPUs were failing, and theories tended to vary between communities for devices. For example, consumers who purchased an Xbox 360 from this era and even some independent repair technicians assumed that the issue was caused by Microsoft using lead-free solder balls- used to connect the Xenos GPU to the motherboard- in order to meet new-at-the-time RoHS standards in the European Union. Many blamed it on the more brittle nature of non-leaded solder balls compared to traditional leaded ones. Others thought it was related to the solder balls' melting point, and that the console getting too hot and literally "desoldering" the GPU. These misconceptions largely faded into obscurity when Microsoft released the Jasper and Falcon revisions of the Xbox 360, and announced their extended warranty on their defective consoles. They were dispelled entirely with the release of Chapter 5 of Power On: The Story of Xbox, where hardware engineers for Xbox during this era explained the problem.

With the PlayStation 3, much of the same speculation happened as with the Xbox 360. However, because the defect was never acknowledged as such by Sony, it was far less clear to consumers what exactly was causing the fault in consumers' machines. Long after the PS3 was no longer supported by Sony, this led many retro console and repair enthusiasts to make poorly informed choices in how to attempt to repair their devices. By 2018, many owners were also convinced for some time that the fault wasn't even related to their RSX GPU, but rather, the NEC-Tokin capacitors surrounding it. The reason for the misconception was because Toshiba laptops from the same era had their NEC-Tokins fail prematurely due to a design flaw specific to those laptops. Some general hardware failures can be caused by the PS3's NEC-Tokin capacitors failing (primarily due to age), but Sony had designed the motherboard correctly when placing the NEC-Tokins- so they would not fail prematurely. It was not until 2022 that this misconception was largely corrected amongst the community, and the true defect was identified as likely to be Bumpgate-related.

A common bad practice between both the Xbox 360 and the PS3 was to reflow the entire motherboard. Even alleged professional repair technicians did this, as shown in BBC's Watchdog segment for the "Yellow Light of Death". However, because this fault was caused by a defect in the GPU, this would only fix the problem temporarily, if at all. In addition, a motherboard is not designed to go through this process more than once (specifically, at the time of manufacture), so this would damage the motherboard over time until it would be impossible to fix the system again.

The only way to permanently fix systems with a Bumpgate-affected GPU is to replace the defective GPU with one that doesn't have the defect. How possible that is, however, depends much on a person's skills, tools, and how easy or difficult it is for regular consumers to repair and modify the system. For example, the PS3 has a modification called "Frankenstein", which makes it possible to replace a defective 90nm RSX GPU with a non-defective 65nm or 40nm GPU from one of the newer PS3s. This modification was inspired by an "official" fix that Sony performed on some faulty PS3s. However, for a person to perform this repair, they would need a BGA rework station (and the skills to use it properly) and a softmodded PS3. A similar modification is also possible for early Xbox 360s, but it requires the same tools and modding experience as the PS3 does. Due to the specialized nature of such modifications, most people will still recommend for consumers to find someone skilled enough to do the modification for them, buy a system that is already modified, or to simply avoid purchasing consoles impacted by Bumpgate.

References

Add a category with the same name as the product, service, website, software, product line or company that this article is about.

The "Incidents" category is not needed.

---

Add your text below this box. Once this section is complete, delete this box by clicking on it and pressing backspace.