AMD’s new Versal tamper-resistant chips combine AI-capable processing with hardware designed to detect physical attacks on the silicon. The devices are expected to sample in the second half of 2027, with production planned for the first half of 2028, so this is a product announcement rather than hardware customers can order today.
AMD Watch analysis
What happened
The Versal TR family is based on AMD’s Versal AI Edge Series Gen 2 adaptive SoCs. AMD says a sensor-based protection die is bonded to the main compute die, shielding the silicon’s backside; if that protection die is tampered with or removed, the device can prevent booting. The company also describes expanded post-quantum cryptography support, hybrid boot flows and additional glitch detection.
The product combines Arm CPUs, AI Engines and programmable logic. AMD lists three models and says customers can begin development on selected standard Versal AI Edge Series Gen 2 devices before moving to footprint-compatible tamper-resistant versions. Engineering.com reports the announcement and its technical details.
Our top picks
- Protection built into the device
A bonded sensor die is intended to detect attacks on the silicon’s backside and block boot if it is tampered with or removed. - Post-quantum cryptography support
AMD says the chips add acceleration for ML-DSA and SLH-DSA signature algorithms. - More layers against glitches
The design adds an orthogonal detector intended to catch glitches across two voltage rails at once. - A route from development to deployment
Footprint-compatible standard devices let customers start designs now and later transition to the TR versions. - AI-capable processing in the same package
The family combines AI Engines and programmable logic with Arm CPUs, rather than treating security as a separate add-on.
Why it matters
Security-critical systems often rely on custom enclosures and separate components to protect chips from physical interference. AMD says integrating protection and processing can reduce that extra hardware and simplify board design. For teams building edge systems, the combination of compute and tamper detection is the practical point: security need not arrive as a second, custom-built box bolted onto the first.
The useful test is still some way off. AMD expects sampling in late 2027 and production in early 2028. The announcement describes intended protections, not a guarantee against every kind of tampering; the company itself says tamper-resistant does not mean tamper-proof.
Our read
This is a more concrete security pitch than a new badge on an existing chip: the proposed sensor die, boot response and cryptographic features give engineers specific design choices to assess. The catch is timing. Customers can start with standard devices, but they will need technical documentation and hands-on evaluation before judging how well the protections fit their threat models. The silicon has made a promising introduction; it has not yet passed the entrance exam.
What to watch
- Whether AMD publishes the detailed threat model and technical documentation.
- When the three Versal TR models begin sampling in the second half of 2027.
- Whether the planned production start in the first half of 2028 holds.
- How the protection and performance compare in customer evaluations.
Discussion spark: For security-critical edge systems, would you prioritise integrated tamper protection in the chip, or keep security components separate so each can be independently tested and replaced?
Sources and evidence
- AMD introduces Versal devices with built-in physical attack protection – Engineering.com (16 November 2026, 07:37 UTC)
Independent WittyWires tracker for public updates about AMD. Not affiliated with or endorsed by AMD; this is not an official account.