AMD Announces Versal Tamper-Resistant Adaptive SoCs With a Backside Protection Die and ML-DSA and SLH-DSA Accelerators on 12 November 2026, Sampling in 2027: Readiness Not Applicable
On 12 November 2026 AMD announced the 2VE3358TR, 2VE3558TR and 2VE3858TR, Versal AI Edge Series Gen 2 adaptive SoCs with a sensor-based backside protection die bonded to the compute die by 3D active-on-active technology, so the device refuses to boot if the die is tampered with or removed. The same parts add ML-DSA and SLH-DSA hardware accelerators, a hybrid boot flow, an improved key vault, a memory element aging sensor and a second glitch detector, with samples in the second half of 2027 and production in the first half of 2028. The post sets the announcement beside AMD security bulletin AMD-SB-8034 of 2 September 2026 and the Bochum Leakback CRC paper, which showed optical backside probing reading encrypted bitstreams on 7-Series FPGAs, reads the development as a defensive post-quantum root of trust in hardware whose readiness cannot yet be placed on the ladder, and lists the silicon, threat model, independent evaluation and CNSA 2.0 calendar a program office should wait for.
Georgia Tech's Gupta Puts a Ceiling on What One Twiddle Fault Leaks From ML-KEM and ML-DSA Key Generation, arXiv 2610.30383, 21 October 2026: 254 of 256 Coefficients Exposed on Paper, TRL 2
Shandong University's Chen and Li Recover 16 Private-Key-Dependent Bits From BIKE's BGF Decoder by Deep-Learning Power Analysis at 95.68 Percent, Cybersecurity 9:215, 16 September 2026
On September 16, 2026, the journal Cybersecurity published a paper by Geng Chen, Yanbin Li and colleagues at Shandong University that reads a private-key-dependent 16-bit sequence out of the BIKE post-quantum decoder from a single power trace on an ARM Cortex-M4 board, at 95.68 percent perfect recovery. We read the deep-learning side-channel attack as an offensive extraction method, weigh what it does and does not transfer to the ML-KEM and HQC code a force will field, and set out the missing key-recovery stage, the untested countermeasures and the single-board evidence that stand between this post-quantum implementation security result and anything a program office would plan against.
The Post-Quantum Migration Reaches the Silicon Layer
In one August week, a Wuhan university team released a 28-nanometer post-quantum cryptography chip with a naval research institute in the room, and a Canada-Taiwan consortium validated a compute-in-memory rival for the same NIST algorithms. A close look at what a hardware root of trust means for forces whose platforms outlive their ciphers, and at the evaluation gates still standing between both chips and a program office.