docs: trim site nav and focus messaging on AMD NPU

docs/_config.yml

@@ -40,26 +40,18 @@ defaults:
 nav_links:
   - label: Technology
     children:
-      - label: How It Works
-        url: "/how-it-works/"
       - label: Models
         url: "/models/"
       - label: Benchmarks
         url: "/benchmarks/"
       - label: Demos
         url: "/demos/"
-  - label: Testimonials
-    url: "/testimonials/"
   - label: Company
     children:
       - label: Team
         url: "/team/"
-      - label: Community
-        url: "/community/"
       - label: News
         url: "/news/"
-      - label: Roadmap
-        url: "/roadmap/"
   - label: Docs
     url: "/docs/"
 

docs/benchmarks.md

@@ -54,30 +54,5 @@ sections:
       alt: "Gemma3 4B benchmark overview of power efficiency (TPS/W) for both prefill and decoding"
       title: "Ultra-high power efficiency"
 
-
-  # - type: media
-  #   variant: alt
-  #   kicker: "Qualcomm reference telemetry"
-  #   # title: "Documenting partner power + thermal sweeps"
-  #   media:
-  #     # title: "Power, temp, and workload overlays"
-  #     body: |
-  #   gallery:
-  #     - src: "/assets/bench/qualcomm15.png"
-  #       alt: "Qualcomm benchmark instrumentation on a reference laptop"
-  #     - src: "/assets/bench/qualcomm14-1.png"
-  #       alt: "Qualcomm bench capture 1"
-  #     - src: "/assets/bench/qualcomm14-2.png"
-  #       alt: "Qualcomm bench capture 2"
-  #     - src: "/assets/bench/qualcomm14-3.png"
-  #       alt: "Qualcomm bench capture 3"
-  #     - src: "/assets/bench/qualcomm14-4.png"
-  #       alt: "Qualcomm bench capture 4"
-  #   items:
-  #     - heading: "On-device telemetry"
-  #       body: "Photos include wattage, temperature, and throughput overlays straight from the Qualcomm reference UI."
-  #     - heading: "Replayable sessions"
-  #       body: "Visual artifacts are stored next to CSV logs so OEM teams can validate every published datapoint."
-
 ---
 

docs/index.md

@@ -9,7 +9,7 @@ sections:
     kicker: "NPU-first runtime"
     title: "The fastest, most efficient LLM inference on NPUs"
     body: |
-      FastFlowLM (FLM) delivers an Ollama-style developer experience optimized for tile-structured NPU accelerators. Install in seconds, stream tokens instantly, and run context windows up to 256k — all with dramatically better efficiency than GPU-first stacks. Our GA release for AMD Ryzen™ AI NPUs is available today, with betas for Qualcomm Snapdragon and Intel Core Ultra coming soon.
+      FastFlowLM (FLM) delivers an Ollama-style developer experience built exclusively for AMD Ryzen™ AI NPUs. Install in seconds, stream tokens instantly, and run context windows up to 256k — all with dramatically better efficiency than GPU-first stacks.
     ctas:
       - label: "Download FastFlowLM (Windows)"
         href: "https://github.com/FastFlowLM/FastFlowLM/releases/latest/download/flm-setup.exe"

docs/news.md

@@ -93,7 +93,7 @@ sections:
       This isn't a "someday" promise. It is working **right now.**
 
       * **Powering AMD:** FastFlowLM is the chosen AI runtime software engine inside the official **[AMD Lemonade Server](https://lemonade-server.ai/)** 🍋, bringing production-ready NPU AI capability to developers and partners.
-      * **Expanding Platform Support:** FastFlowLM is also prepared for next-generation Qualcomm devices and additional NPU platforms.
+      * **Expanding Platform Support:** FastFlowLM continues to deepen support across the AMD Ryzen™ AI NPU lineup (Strix, Strix Halo, Kraken).
 
       ---