Microsoft defines AI PC in a new way with four tier roadmap. The roadmap clearly links AI PC with memory and local computation. This is important since it gives clarity to the user what their PC can do with AI.
Microsoft defines AI PC in a new way with four tier roadmap
No more vague AI designations. Instead, there are hard numbers and precise boundaries that provide AI function capabilities. For the basic AI help, the basic PC's are available. For the professional who needs to do complex model local processing on their own, higher tier PC's are required.
- Tier 0 Memory: 8 GB
- Tier 0 Use Case: Local context and actions, basic AI tasks
- Tier 0 Hardware: Intel Core Series 3/4, AMD Ryzen 100/200, Snapdragon C/X/X Plus
- Tier 1 Memory: 16 GB to 64 GB
- Tier 1 Compliance: Copilot+ compliant
Tier 0 or entry level PC is targeted at local context and basic AI use cases. These PC's must have 8 GB memory or more with mainstream processors. The mainstream processors can be Intel Core Series 3 or 4, AMD Ryzen 100 or 200 series and Snapdragon C, X or X Plus.
This makes the very basic PC available for even the lowest price point. Those who are willing to spend just enough can have access to basic AI features without having to pay for enterprise class PC's.
Tier 1 or mainstream level PC requires 16 GB memory or more. It can go up to 64 GB. These PC's are Copilot+ compliant and have extended AI features.
These are the latest silicon devices. These include Intel Core Ultra Series 3 or 4, AMD Ryzen AI 400 or 500 and Snapdragon X2. Microsoft shows future possible PC hardware that may be used as well. These include Intel Core Ultra Series 4 "Nova Lake" and AMD Ryzen AI 500 "Medusa Point".
For those that need to build and run their own large language models locally, Tier 2 devices have 128 GB or more memory. These PC's use high end chips. Examples include AMD Ryzen AI MAX+ 400, NVIDIA RTX Spark and future devices like Intel Core Ultra Next and Snapdragon Next. The highest tier or Tier 3 is the personal supercomputer which uses server class hardware like Intel Crescent Island GPUs, NVIDIA GB300 and AMD Threadripper Halo Station.
Microsoft provides clarity for what is needed to use Windows with AI for different levels of AI performance. The user now can choose the hardware tier that meets their particular need for AI. This may range from basic local use to high volume model building.
The roadmap shows a clear path of current and future PC designs. This allows AI function level of performance to expand along with memory and processor upgrade paths.
Comments (0)