Intel’s new tech has stopped physical attacks

Invasive cyber attacks on data centers cost billions of dollars, but can be mitigated by sophisticated software and hardware-based security defenses. However, a physical attack on machines in a data center using various fault injection methods can cause damage without compromising security. Therefore, Intel developed the Tunable Replica Circuit (TRC) telemetry control technology supported by the 12th generation core Alder Lake platform to detect similar attacks.

“Software protections have been strengthened through virtualization, layered canaries and code verification,” said Daniel Nemiroff, Intel’s principal engineer. “This has led malicious actors to turn their attention to physical attack computer platforms. A favorite tool of these attackers is fault injection attacks with flash voltages, clock pins and electromagnetic radiation that cause circuit timing errors and allow them to execute malicious instructions and potentially flash secrets.”

Intel’s Alder Lake 600-series chipsets (which Intel calls the Platform Control Hub, or PCH) have a TRC module that constantly monitors the operation of other components in the system, including the CPU, which supports the company’s Converged Security and Management Engine (CSME). himself.

TRC monitors timing failures due to voltage, clock, temperature, or electromagnetic malfunctions, which can be caused by a variety of factors. For example, when certain CPU times are out of the ordinary, this could indicate a cyber attack (using a malicious command or exploiting holes in security). A set of TRC sensors are tuned to detect errors caused by fault injection rather than normal workload voltage drops, so Intel is confident that the technology won’t use mitigation techniques to ensure data integrity under normal conditions due to false positives.