Nvidia is recommending a mitigation for purchasers of certainly one of its GPU product strains that may degrade efficiency by as much as 10 % in a bid to guard customers from exploits that might let hackers sabotage work initiatives and probably trigger different compromises.
The transfer is available in response to an assault a crew of educational researchers demonstrated in opposition to Nvidia’s RTX A6000, a extensively used GPU for high-performance computing that’s accessible from many cloud companies. A vulnerability the researchers found opens the GPU to Rowhammer, a category of assault that exploits bodily weak point in DRAM chip modules that retailer information.
Rowhammer permits hackers to alter or corrupt information saved in reminiscence by quickly and repeatedly accessing—or hammering—a bodily row of reminiscence cells. By repeatedly hammering rigorously chosen rows, the assault induces bit flips in close by rows, that means a digital zero is transformed to a one or vice versa. Till now, Rowhammer assaults have been demonstrated solely in opposition to reminiscence chips for CPUs, used for common computing duties.
Like catastrophic mind injury
That modified final week as researchers unveiled GPUhammer, the primary recognized profitable Rowhammer assault on a discrete GPU. Historically, GPUs have been used for rendering graphics and cracking passwords. Lately, GPUs have grow to be the workhorses for duties resembling high-performance computing, machine studying, neural networking, and different AI makes use of. No firm has benefited extra from the AI and HPC increase than Nvidia, which final week turned the primary firm to succeed in a $4 trillion valuation. Whereas the researchers demonstrated their assault in opposition to solely the A6000, it seemingly works in opposition to different GPUs from Nvidia, the researchers stated.
The researchers’ proof-of-concept exploit was in a position to tamper with deep neural community fashions utilized in machine studying for issues like autonomous driving, healthcare purposes, and medical imaging for analyzing MRI scans. GPUHammer flips a single bit within the exponent of a mannequin weight—for instance in y, the place a floating level is represented as x instances 2y. The one bit flip can improve the exponent worth by 16. The result’s an altering of the mannequin weight by a whopping 216, degrading mannequin accuracy from 80 % to 0.1 %, stated Gururaj Saileshwar, an assistant professor on the College of Toronto and co-author of an tutorial paper demonstrating the assault.
