Your Chip Works. Prove It Is Secure. - ARTERIS INC

Financial Performance:

• No specific financial metrics (revenue, earnings, profit margins, cash flow, debt, or deferred revenue) were disclosed as this presentation is a technical product webinar rather than a quarterly earnings call.

Guidance and Future Outlook:

• Anticipates a rapid acceleration in hardware vulnerability discovery, with the potential for new attacks to be identified and automated at scale using AI tools.
• Outlook emphasizes increasing regulatory pressure from the EU Cyber Resilience Act (CRA), ISO 21434 in automotive, and U.S. government requirements for third-party IP security evidence, driving demand for pre-silicon security assurance.
• Foresees growing adoption of security verification in AI, virtualization, and multi-tenant environments where data isolation is critical.

Business Segments and Product Lines:

• Focuses on Arteris Harbor Security and its proprietary PsyQ (Information Flow Tracking) technology, which tracks assets (e.g., cryptographic keys) through chip designs during simulation and emulation.
• Product distinguishes between functional security verification (does the crypto block work?) and security protection verification (did the key stay within protection boundaries?).
• Capabilities include tracking assets to identify blind spots, ensuring coverage of ~80% of hardware weaknesses before tape-out, and generating sign-off reports for regulatory compliance.
• Explicitly states the company does not provide hardware acceleration for post-quantum cryptographic algorithms; its focus remains on pre-silicon verification.

Market and Competitive Landscape:

• Market trends indicate a shift from expert-led attack discovery to automated, scalable attacks, increasing the necessity for rigorous pre-silicon checks.
• Key customers and use cases include ARM (expanding usage of security solutions in new designs) and U.S. federal government programs (collaborating with BAE Systems on third-party IP risk reduction).
• The product is gaining traction in the automotive industry, federal defense sector, and physical AI environments.
• Competitors or alternative methods are noted as requiring extensive mathematical calculations or failing to visualize data flow, whereas PsyQ uses symbolic analysis independent of data values.

Risks and Challenges:

• Hardware Vulnerabilities: Rapid growth in reported hardware weaknesses (e.g., Spectre, Meltdown, reset-based lock clearing) has eclipsed previous years' totals, with attacks becoming easier to replicate (e.g., a Spectre-like attack in under four minutes using low-cost hardware).
• Blind Spots: Functional testing often fails to detect if sensitive data leaks into debug interfaces, unused buffers, or cross-tenant areas despite passing functional tests.
• Regulatory Non-Compliance: Failure to provide evidence of hardware security diligence could lead to non-compliance with EU CRA and automotive standards.
• Design Complexity: The increasing complexity of modern chips (secure boot, shared caches, speculative execution) makes manual verification of data flows nearly impossible without automated tracking.

Management Commentary and Tone:

• Management (Leon Brown, Pete Marr) maintains a confident and educational tone, emphasizing that "passing functional tests does not prove assets stayed protected."
• Emphasizes a paradigm shift from testing end-states to tracking asset lifecycles to generate "sign-off evidence."
• Expresses urgency regarding the speed of modern attack discovery and the liability associated with post-silicon vulnerabilities.

Analyst Questions and Answers:

• Q: What makes pre-silicon security assurance more important?
  • A: AI tools are accelerating the conceptualization and automation of hardware attacks, raising the bar for defensive design verification.
• Q: How can a chip pass functional tests yet remain exposed?
  • A: Functional tests verify expected behavior, whereas security verification tracks where protected information travels and if it leaks to unintended locations (e.g., debug ports, shared memory).
• Q: How is sensitive asset tracking achieved?
  • A: Using Information Flow Tracking (PsyQ) with symbolic analysis to visualize asset movement regardless of native data values, independent of specific test vectors.
• Q: Does Arteris accelerate post-quantum algorithms?
  • A: No; Arteris focuses on verification, not hardware acceleration.

Other Key Points:

• The presentation cites the MITRE Common Weakness Enumeration (CWE) database, noting 112 known hardware weaknesses in version 5.0, with PsyQ targeting roughly 80% of these pre-tape-out.
• Specific security failure examples highlighted include keys lingering in third-party IP after boot, security engine locks being cleared by chip resets, and race conditions in firewall policies where data moves before checks are enforced.
• The webinar was sponsored by EE Times and broadcast by AspenCore, featuring collaboration on the U.S. Department of Defense's Third-Party IP (3PIP) program.