company

Advanced Micro Devices (AMD)

Semiconductor / AI Hardware & Software
72Good
Aio Composite Score
Confidence
ApplicationHeavy
high
ImpactMixed
medium
OpennessTransparent
medium
AMD is an aggressive adopter and developer of AI technology, positioning itself as both a manufacturer of AI hardware (Instinct MI300X/MI250 GPUs, Ryzen AI NPUs) and a contributor to the open-source AI software ecosystem (ROCm, open-weight models). The company's AI activity is overwhelmingly product-facing — building and selling AI compute infrastructure — rather than deploying AI internally to automate its own workforce. In November 2024, AMD cut approximately 1,000 employees (4% of workforce), framing the move as a resource reallocation toward 'largest growth opportunities' including AI chips; however, the causal link between internal AI adoption and these specific cuts is circumstantial at best, as the stated rationale was strategic competitive repositioning against Nvidia. AMD has published a Responsible AI Use Policy and maintains a public 'Responsible AI' page emphasizing transparency, human agency, and fairness, consistent with its open-source posture.
Application
How is AI being used?
Heavy
high
AMD is deeply embedded in AI at the product level — developing and selling AI accelerator hardware (Instinct GPUs, Ryzen AI NPUs), releasing…
Impact
How have people been affected?
Mixed
medium
AMD's AI hardware and open-source contributions have enabled significant positive outcomes for external developers, enterprises, and researchers.…
Openness
Company honesty about AI use.
Transparent
medium
AMD has published a formal Responsible AI Use Policy, a Responsible AI webpage with explicit human-centric design commitments, and open-sourced AI…
AI Tools & Platforms Detected
AMD ROCmAMD Instinct MI300X / MI250 GPUsAMD Ryzen AI / XDNA NPUAMD-Llama-135M (AMD-135M SLM)Instella-Long (3B-parameter open LLM)Vitis AI QuantizerONNX Runtime with AMD GPU EPMIGraphX / rocMLIR / MIOpen / HIPBlasLTAmuse (AI image generation app)AMD Silo AIStable Diffusion (via Ryzen AI Software)Whisper / Zipformer (speech-to-text via Ryzen AI)DeepSeek-R1 Distill Series (supported on Ryzen AI)
AI Hiringmoderate
30AI roles/ 2180 total
Infrastructure Architect – GPU Test Automation Farm (AMD)ML Compiler Engineer (AMD-XILINX, via Elevated Resources)Strategic Business Development Specialist – AI Inference / Edge (AMD-XILINX, via Elevated Resources)Software Engineer, Inference – AMD GPU Enablement (OpenAI, AMD-facing role)Software Engineering Manager – AI Hardware Systems Co-Design (Meta, AMD ecosystem)
Key Evidence (13)
AMD released the Instella-Long 3B-parameter open LLM trained on 64 Instinct MI300X GPUs with fully open weights, data, and code
web searchSource ↗
AMD released its first small language model (AMD-135M) trained from scratch on AMD Instinct MI250 accelerators on 690B tokens, open-sourcing weights, code, and data
2024-09-17
web searchSource ↗
AMD Ryzen AI Software platform supports LLM deployment (including DeepSeek-R1, Phi-4, Qwen-3) on NPU hardware with up to 16K context length, plus Stable Diffusion image generation and Whisper speech-to-text
web searchSource ↗
AMD ROCm open-source software stack is adopted by OpenAI, Microsoft, Meta, and Oracle for AI workloads
web searchSource ↗
AMD confirmed laying off approximately 1,000 employees (~4% of its ~26,000 workforce) in November 2024, citing reallocation toward 'largest growth opportunities' including AI chips
2024-11-13
web searchSource ↗
AMD layoffs of 4% were described by company as strategic pivot to focus on high-growth areas like artificial intelligence, not attributed to internal AI automation replacing workers
2024-11-01
web searchSource ↗
What insider reports could clarify
  • Are AMD employees using internal AI tools (e.g., Copilot, internal LLMs) in daily engineering or business workflows?
  • Were any of the 1,000 November 2024 layoffs in roles where AI tooling directly replaced the function?
  • How are AMD's own open-source AI models reviewed or red-teamed before public release?
  • Do AMD's product teams use AI-generated code or designs in hardware/software development pipelines?
  • Has AMD backfilled any of the cut roles with AI tooling rather than re-hiring?
Submit a ReportClaim This Profile