Nvidia manufactures roughly 20 unique products from the same GA100 silicon die by relying on internal eFuses to restrict performance tiers and memory. Through a targeted voltage fault injection attack, a researcher blew specific fuses to transform a CMP 170HX mining processor into an enterprise-grade A100-80G equivalent. The exploit unlocked previously inaccessible hardware capabilities, including 80 GB of High Bandwidth Memory, PCIe Gen4 speeds, and NVLink interconnects. The underlying vulnerability was shared directly with Nvidia's product security team prior to public disclosure.
Key points
Nvidia's unusual silicon architecture allows attackers to obtain broader hardware access…: Nvidia's unusual silicon architecture allows attackers to obtain broader hardware access by blowing additional fuses rather than enforcing tighter restrictions.
Physical voltage glitching allows operators to bypass factory product segmentation and un…: Physical voltage glitching allows operators to bypass factory product segmentation and unlock enterprise-tier features like NVLink and Error-Correcting Code memory.
The security flaw was formally disclosed to Nvidia's product security team before the tec…: The security flaw was formally disclosed to Nvidia's product security team before the technical research was made public.
Sources checked by our analysis pipeline: 1
On this page
Summary
I’ve discovered a voltage fault injection vulnerability on the GA100 die, allowing an operator to blow arbitrary fuses. Utilizing this vulnerability, I’ve converted my CMP 170HX 8GB card into what is effectively an A100-80G! Further, my unlocked cards actually perform ~1.2x faster than a retail A100 by having all functional SMs enabled (~120-126, die-lottery dependent), rather than the yield-safe 108 SMs NVIDIA chose for all production A100 cards. Finally, CMP-disabled features like NVLink, PCIe Gen4, NVJPG, NVDEC, MIG are all unlockable with this approach, though I haven’t specifically tested NVJPG/NVDEC/MIG.
Here’s how my “A100-80G” appears and functions.
NVIDIA uses fuses for product segmentation, security features, and other fun access control features - and crucially, there are several functions where blowing a fuse on an NVIDIA die can grant you more access, not less. In a properly secured design, a blown fuse should always induce a more restrictive state, because a blown fuse cannot be unblown in an economically feasible way. NVIDIA’s design allows one to relax restrictions by blowing more fuses!
Public institutions and international cultural bodies are deploying new digital platforms to broaden access to civic records, historical artifacts, and cinema. Concurrently, government agencies are modernizing legacy operations, even as new disclosures highlight persistent cybersecurity and privacy vulnerabilities across commercial and tracking technologies. These parallel developments underscore an accelerating shift toward open data alongside the urgent need to protect sensitive digital infrastructure.
An inquiry into 11 smart television and streaming hardware makers highlights how the industry harvests household audio and viewing behaviors. While several major tech brands defended their audio safeguards, academic findings reveal that opting out of integrated trackers does not completely stop data transmission to corporate servers. This continuous background tracking raises persistent questions regarding consumer consent and digital privacy in living rooms.
Early adopters of Apple's new $2,000 foldable are returning the device as initial two-week trial periods expire. Customers report dissatisfaction with the hardware's bulk, underwhelming cameras, and the practical challenges of combining a smartphone and tablet into a single unit. These hardware grievances, however, have not pushed users away from Apple's broader ecosystem, with most returnees simply reverting to conventional iPhone models.