Skip to content
Tech News
← Back to articles

New GPUBreach attack enables system takeover via GPU rowhammer

read original get GPU Rowhammer Attack Kit → more articles
Why This Matters

The GPUBreach attack highlights a significant vulnerability in GPU memory, enabling attackers to escalate privileges and potentially take over entire systems without disabling IOMMU protections. This revelation underscores the evolving threat landscape in hardware security, emphasizing the need for stronger safeguards in GPU and driver design. For consumers and the industry, it signals a critical need for updated security measures and vigilance against emerging hardware-based exploits.

Key Takeaways

A new attack, dubbed GPUBreach, can induce Rowhammer bit-flips on GPU GDDR6 memories to escalate privileges and lead to a full system compromise.

GPUBreach was developed by a team of researchers at the University of Toronto, and full details will be presented at the upcoming IEEE Symposium on Security & Privacy on April 13 in Oakland.

The researchers demonstrated that Rowhammer-induced bit flips in GDDR6 can corrupt GPU page tables (PTEs) and grant arbitrary GPU memory read/write access to an unprivileged CUDA kernel.

An attacker may then chain this into a CPU-side escalation by exploiting memory-safety bugs in the NVIDIA driver, potentially leading to complete system compromise without the need to disable Input-Output Memory Management Unit (IOMMU) protection.

GPUBreach attack steps

Source: University of Toronto

IOMMU is a hardware unit that protects against direct memory attacks. It controls and restricts how devices access memory by managing which memory regions are accessible to each device.

Despite being an effective measure against most direct memory access (DMA) attacks, IOMMU does not stop GPUBreach.

“GPUBreach shows that GPU Rowhammer attacks can move beyond data corruption to real privilege escalation,” the researchers explain.

“By corrupting GPU page tables, an unprivileged CUDA kernel can gain arbitrary GPU memory read/write, and then chain that capability into CPU-side escalation by exploiting newly discovered memory-safety bugs in the NVIDIA driver.”

... continue reading