Battering RAM Exploit Bypasses Modern Protections in Intel, AMD Cloud Processors

Cloud providers utilize hardware-based memory encryption to safeguard user data, including sensitive information such as passwords, financial records, and personal files. Technologies like Intel SGX and AMD SEV-SNP are implemented to ensure that data remains protected, even if a cloud host or administrator is compromised.

However, a new attack method, termed Battering RAM , has emerged, demonstrating that these safeguards can be bypassed using a hardware add-on that costs less than $50.

Researchers have developed a low-cost interposer, a small board that fits between the processor and memory module. During the system startup, the interposer operates normally, passing all built-in security checks. Once the operating system and cloud platform have validated the memory, the attacker can activate a hidden switch on the interposer.

Subsequently, the device redirects encrypted memory requests to locations controlled by the attacker, enabling data corruption or replay without detection.

Stealthy Startup: The interposer remains transparent during boot, allowing the system to validate memory encryption keys and setup routines. Triggering Malicious Mode: After boot, the attacker flips a hardware switch. The interposer modifies address metadata, redirecting protected memory pages to attacker buffers. Ciphertext Capture: The attacker captures encrypted data by reading the aliased buffer, especially targeting secure enclaves or virtual machines. Replay Attack: The interposer is reconfigured, allowing the attacker’s enclave to occupy the victim’s physical address. The captured ciphertext is replayed, forcing decryption and execution in the attacker’s context. Plaintext Exposure: Decrypted data appears in the attacker’s enclave, granting full read or write access to memory regions protected by Intel SGX or AMD SEV-SNP.

Intel SGX: The attack bypasses enclave isolation by replaying ciphertext into an attacker’s enclave. AMD SEV-SNP: The attack undermines integrity checks and remote attestation used to verify VM memory authenticity.

Both Intel and AMD have acknowledged the findings. Mitigating the Battering RAM attack requires a redesign of memory encryption protocols, with potential defenses including:

Adding cryptographic integrity checks on memory metadata. Verifying memory mapping throughout system runtime, not just at boot. Introducing tamper-evident hardware that alerts on unexpected metadata changes.

Until these changes are implemented in future processors, cloud users should include physical memory attacks in their threat models and consider additional safeguards such as multi-party memory encryption or real-time hardware attestation to protect critical workloads.

Based on reporting by GBHackers.