Siloing security: A response to Meltdown and Spectre
Earlier this week, Jann Horn of Google’s Project Zero published a detailed blog post titled “Reading privileged memory with a side-channel.” The post confirmed that CPU data cache timing can be exploited to efficiently leak information out of mis-speculated execution. This could lead to – at worst – arbitrary virtual memory read vulnerabilities across local security boundaries in various contexts.
Put simply, the above-mentioned security issue could allow cyber criminals to steal the entire memory contents of computers, including mobile devices, personal computers and servers running in cloud computer networks. Categorized as two distinct security flaws, Meltdown and Spectre were independently disclosed by a number of security experts, including senior Rambus technology advisor Paul Kocher and senior Rambus security engineer Mike Hamburg.
To read the full article, click here
Related Semiconductor IP
- nQrux® Root of Trust IP
- AXI to UCIe Bridge IP
- UCIe 2.x Controller IP
- SWI3S (SoundWire I3S Interface) Peripheral Controller Core IP
- OpenTitan-based RISC-V Secure Element
Related Blogs
- What Does a GPU Have to Do With Automotive Security?
- The Silent Guardian of AI Compute - PUFrt Unifies Hardware Security and Memory Repair to Build the Trust Foundation for AI Factories
- From Classical CAN and CAN FD to CAN XL: Functional Safety and Security for Next-Generation In-Vehicle Communication
- Want to Mix and Match Dies in a Single Package? UCIe Can Get You There
Latest Blogs
- From Bug Hunting to Engineering Methodology: Lessons from AI-Driven Development
- Manufacturing Intelligence: Turning EDA Data into Trusted Action
- Execute-in-Place: Getting More from Embedded NVM
- Rethinking RTL flows with AI-driven hybrid formal verification
- Arm and NVIDIA: Building the trusted compute foundation for the agentic AI era