In short: Google researchers recently disclosed a secretly safety defect, called “Entrycene”, which allowed malicious code execution through a microCode patch through the AMD processor through Zen 1. Team Red has just confirmed that its upcoming Zen 5 chips are also affected.
The main problem is a defect in the signature verification of AMD for microcode updates-the low-level patch chipmakers posted after the CPUS ship to fix the safety issues. Typically, the operating system or firmware only loads the microcode that AMD has signed and approved. Entricing bypassed the security guard on the affected chips with the ring 0 (kernel-level) to the attackers.
Last month, AMD stated that Entrycene had influenced the first four generations of Zen CPU in its entire product line. Everything from mainstream ryzen chips to beef epic server processor was unsafe.
Team Red updated its safety bulletin this week, confirming that the new Zen 5 chips are also unsafe for the bug. The affected systems include:
- Ryzen 9000 “Granite Ridge” Cpus
- EPYC 9005 “Turin” Server Chips
- AI-focused Ryzen AI 300 processor Strikes Hello, Strikes Point and Crackon Point
- Ryzen 9000hx “Fire Range” laptop CPUS.
The good news is that AMD has already given a fix to motherboard vendors through Comboam 5PI 1.2.0.3C Agesa update. Therefore, if you have not already done this, see your motherboard vendor’s website for BIOS updates.
The server position is slightly more complicated. While AMD has issued mitigation for entricing -affected desktops and old EPYC chips, patch for the unprotected new EPIC Turin models for the bug is not expected until the end of this month.
On the bright side, high-level system privileges are required to execute this hack. Continuing to frequent malware, a system restarts any malicious microcode loaded in this way. While practical risk for typical consumers is relatively low, the possibility of misuse in data centers and cloud settings. It creates a significant safety concern that AMDs and its partners are working quickly.
