Anthropic Uses AI to Slash GPU Kernel Development Time
Anthropic has utilized its Claude Mythos model to automate the complex process of porting AI workloads to AMD hardware, potentially challenging Nvidia's long-standing software dominance.
🔌Chips & HardwarePhoto: Hacker News Front Page
Key points
- Anthropic used Claude Mythos to port AI kernels to AMD GPUs in three days.
- The task previously required 6-12 months of manual engineering effort.
- The shift threatens Nvidia's CUDA ecosystem by automating code translation.
- AMD is actively using AI to optimize its ROCm software stack and kernel development.
Anthropic recently revealed that its Claude Mythos model successfully automated the development of low-level software kernels required to run AI models on AMD’s MI355X GPUs. According to company co-founder Tom Brown, the task was completed in approximately three days with virtually no human intervention. This process, which involves writing the foundational code that bridges AI models to specific hardware architectures, previously required dozens of engineers and between six to 12 months of development time.
The ability to rapidly port software between different chip architectures addresses a significant barrier in the semiconductor market. Historically, Nvidia has maintained a strong competitive advantage through its CUDA platform, which creates a deep software ecosystem that ties developers to its hardware. Anthropic’s breakthrough suggests that AI-driven automation could effectively neutralize this "moat," allowing companies to treat different GPU providers as a fungible pool of compute resources.
AMD executives have confirmed that AI-generated kernels can sometimes outperform manually tuned versions, signaling a shift in how data center infrastructure is optimized. As AI agents increasingly handle the translation of code between platforms, the reliance on proprietary, legacy software environments like CUDA may decline, potentially accelerating the adoption of alternative hardware in large-scale AI deployments.
Source Hacker News Front Page — tomshardware.com/tech-industry/artificial-intelligence/anthropic-sa… ↗
Read next
🧠AI AnalysisPedestal AI Secures Seed Funding to Automate Logistics With Intelligent Agents
Pedestal AI has launched publicly with $4.7 million in seed backing to deploy agentic artificial intelligence aimed at streamlining logistics and supply chain operations.
🏢Big TechUnited Launch Alliance Investigates Vulcan Rocket Failure
United Launch Alliance is investigating a mission failure of its Vulcan rocket, which suffered from low booster performance shortly after liftoff. The incident raises que…
🌐Work & SocietyUS Southwest Hydropower Output Hits Record Lows Amid Persistent Drought
Hydropower generation in the US Southwest dropped 28% below the 10-year average this summer as record-low water levels in the Colorado River Basin threaten regional energ…
🏢Big TechYouTube Updates View Metrics to Include TV Co-Viewing Estimates
YouTube is incorporating estimated co-viewing data into its public view counts to better reflect group consumption on television sets.
⚖️Regulation & PolicyOpenAI Partners with Chilean Ministry to Advance Public Sector AI
OpenAI has signed a non-binding agreement with Chile's science ministry to integrate artificial intelligence into public administration and support local startups.
🌐Work & SocietyAmbience Healthcare Expands Clinical AI Suite for Hospital Operations
Ambience Healthcare has launched new AI tools designed to assist hospital staff with patient rounding and discharge planning, aiming to improve both clinical efficiency a…