New DDRop Attack Breaks Intel TDX and AMD SEV-SNP Confidential Computing
New DDRop attack bypasses Intel TDX and AMD SEV-SNP confidential computing memory protections.
Researchers have disclosed a new hardware attack named DDRop that targets Intel TDX and AMD SEV-SNP confidential computing technologies. By using a custom-built interposer circuit, the attack silently drops writes to a server's memory, causing the processor to read outdated encrypted data, thereby breaking memory integrity and potentially allowing attackers to gain control of protected virtual machines.
AI summary of the linked source. Not reviewed by an editor. How this works
Run Google?
Get an email when a new story names Google, usually within the hour.
Free. Your list stays private and never appears in a subject line. One click stops it. How Vendor Watch worksPrivacy
The lesson from this incident
Full lessonThe DDRop attack exploits a physical hardware vulnerability by inserting a custom interposer circuit between the CPU and memory modules, silently dropping memory write operations so the processor reads stale, outdated encrypted data. This undermines the core integrity guarantees that confidential computing technologies like Intel TDX and AMD SEV-SNP are designed to provide, potentially allowing an attacker with physical or supply-chain access to hijack protected virtual machines. The attack is particularly dangerous in cloud and data center environments where tenants rely on hardware-enforced isolation to protect sensitive workloads from even privileged insiders. This highlights that cryptographic and software-layer protections are insufficient when physical hardware integrity cannot be guaranteed, making supply chain security and physical access controls critical.
What helps against this
How we list vendors- OT and industrial control system security
Factory, building and utility control systems get visibility and protection.
- Encryption and key management
Stolen data stays unreadable, and the keys stay under your control.
Indicators of Compromise
Extracted by AI from the linked source. Not reviewed by an editor. How this works
- mitre_attack — T1560.001