/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

← → days · ↑ ↓ browse · Enter similar · o open

Developers release source code on Github of new Linux rootkit that runs in the dedicated memory of GPUs, making it harder to detect

Dan Goodin / Ars Technica :

Ars Technica Dan Goodin

Context & Ripple Effects

This lands in a stretch where rootkits kept moving one layer beneath the operating system: within weeks of this release, Ars reported that Mac firmware from mid-2014 and earlier could be infected with a rootkit without physical access, extending the same below-the-OS logic to Apple's boot ROMs. The GitHub publication matters because it converts a research proof-of-concept into reusable tooling any attacker can build on.

The years since have validated the direction rather than retired it: Kaspersky later surfaced a malicious UEFI rootkit that survived OS reinstalls and drive replacements, and researchers showed modern GPUs themselves are attack surface, with a cross-origin attack reading pixels across sites via GPU image compression in Chrome and Edge.

First-order effects

  • Antivirus and endpoint-security vendors lose their primary vantage point: a Linux rootkit executing in GPU dedicated memory sits outside main-system RAM and disk images their scanners inspect, so detections built for kernel-land malware miss it outright.
  • Operators of GPU-equipped Linux servers and workstations inherit a new persistence threat whose source code is public, meaning the barrier to deployment drops from research skill to copy-paste.

Second-order effects

  • Security firms are pushed toward telemetry beyond the host OS — hypervisor-level inspection, DMA/IOMMU controls, and GPU driver instrumentation — because anything confined to CPU-side visibility cannot see the threat.
  • GPU and motherboard vendors face pressure to treat their firmware and driver stacks as security boundaries, the same way the UEFI rootkit discovery later forced BIOS/UEFI suppliers into the malware-fight conversation.

Third-order effects

  • If hiding spots keep migrating outward — firmware, then dedicated GPU memory — the industry's trust anchor shifts from software scanners toward hardware-rooted attestation, where the silicon itself vouches for what runs beside it.
  • Public code releases of offensive techniques accelerate a dual-use cycle in which each disclosure hardens defenders on known vectors while arming attackers with the next unmonitored layer, raising the baseline cost of endpoint assurance permanently.

The trend: Malware persistence is steadily relocating below the operating system — from firmware to dedicated accelerator memory — outpacing traditional file-and-kernel-scanning defenses.