Merge branch 'main' into main

Author: odincodeshen

.wordlist.txt

@@ -4590,4 +4590,81 @@ xdp
 xhci
 JFR
 conv
-servlet
+servlet
+tv
+gpiozero
+lgpio
+TinyLlama
+Superscalar
+automations
+gemma
+tinyllama
+pinout
+Makatia
+Omusilibwa
+EdgeAI
+Raspi
+abyz
+fidel
+javascript
+makatia
+uk
+STDDEV
+Stdev
+BytesToBytesMap
+HashMap
+LongToUnsafeRowMap
+Stdev
+UnsafeRow
+UnsafeRowhash
+agg
+arrayEqual
+codegen
+hashmap
+hugeMethodLimit
+ints
+kurtosis
+stddev
+wholestage
+RDD
+TEEs
+paravirtualization
+WholeStageCodegen
+Wholestage
+zshrc
+hadoop
+CBL
+DataFrame
+exitCode
+Gerganov's
+Radoslav
+rgerganov
+NSS
+spatio
+upsampling
+UE
+VGF
+NNE
+RDG
+Configurator
+RHI
+RHIs
+NNERuntimeRDGMLExtensionsForVulkan
+Unreal's
+ORT
+MLEmulationLayerForVulkan
+RenderDoc's
+vgf
+dataflow
+Sandboxed
+sandboxed
+Termina
+LXC
+Crostini
+ChromeOS's
+crosh
+Sommelier
+chromeos
+linuxcontainers
+
+

assets/contributors.csv

@@ -96,4 +96,5 @@ William Liang,,,wyliang,,
 Waheed Brown,Arm,https://github.com/armwaheed,https://www.linkedin.com/in/waheedbrown/,,
 Aryan Bhusari,Arm,,https://www.linkedin.com/in/aryanbhusari,,
 Ken Zhang,Insyde,,,,
-Ann Cheng,Arm,anncheng-arm,hello-ann,,
+Ann Cheng,Arm,anncheng-arm,hello-ann,,
+Fidel Makatia Omusilibwa,,,,,

content/install-guides/fm_fvp/fvp.md

@@ -11,15 +11,20 @@ multi_install: false             # Set to true if first page of multi-page artic
 multitool_install_part: true    # Set to true if a sub-page of a multi-page article, else false
 layout: installtoolsall         # DO NOT MODIFY. Always true for tool install articles
 ---
-Arm Fixed Virtual Platforms (FVPs) are provided as a library of ready to use platforms.
 
-{{% notice  Arm Development Tools%}}
-An appropriate subset of the FVP library is installed with [Arm Development Studio](/install-guides/armds) and [Keil MDK](/install-guides/mdk) Professional Edition.
+{{% notice Note %}}
+Arm Fixed Virtual Platforms (FVPs) were available as a library of ready to use platforms (and as a component of Arm Development Studio) up until version 11.28.
+
+From 11.29 onwards the FVPs are provided solely as part of Arm Development Studio.
+
+This install guide is only applicable to the legacy FVP library.
+
+See the [Arm Development Studio Install Guide](/install-guides/armds) and the [Introduction to FVPs](https://developer.arm.com/documentation/110379/1129/Introduction-to-FVPs) documentation.
 {{% /notice %}}
 
 ## Download installer packages
 
-You can download the FVP library installer from the [Product Download Hub](https://developer.arm.com/downloads/view/FM000A).
+You can download the FVP library installer from the [Product Download Hub](https://developer.arm.com/downloads/view/FMFVP).
 
 Linux (AArch64 and x86) and Windows (x86 only) hosts are supported.
 
@@ -47,11 +52,6 @@ For full list of available options, use:
 
 FVPs are license managed. License setup instructions are available in the [Arm Licensing install guide](/install-guides/license).
 
-{{% notice  Arm Development Tools%}}
-The FVPs provided with Arm Development Studio and/or Keil MDK Professional Edition use the license of that product, not that of the FVP library.
-{{% /notice %}}
-
-
 ## Verify installation
 
 To verify everything is working OK, navigate to the install directory, and launch any of the supplied FVP executables. No additional command options are needed.

content/install-guides/llm-jetson-xavier.md

@@ -50,7 +50,7 @@ Run the commands below to set up the ExecuTorch internal dependencies:
 
 ```bash
 git submodule sync
-git submodule update --init
+git submodule update --init --recursive
 ./install_executorch.sh
 ```
 

content/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/2-env-setup.md

@@ -1,5 +1,6 @@
 ---
-title: Overview
+title: Run LLMs locally on Raspberry Pi 5 for Edge AI
+
 weight: 2
 
 ### FIXED, DO NOT MODIFY
@@ -8,66 +9,65 @@ layout: learningpathall
 
 ## Overview
 
-This Learning Path walks you through deploying an efficient large language model (LLM) locally on the Raspberry Pi 5, powered by an Arm Cortex-A76 CPU. This will allow you to control your smart home using natural language, without relying on cloud services. With rapid advances in Generative AI and the power of Arm Cortex-A processors, you can now run advanced language models directly in your home on the Raspberry Pi 5.
+This Learning Path walks you through deploying an efficient large language model (LLM) locally on the Raspberry Pi 5, powered by an Arm Cortex-A76 CPU. This setup enables you to control your smart home using natural language without relying on cloud services. With rapid advances in generative AI and the power of Arm Cortex-A processors, you can now run advanced language models directly in your home on the Raspberry Pi 5.
 
-You will create a fully local, privacy-first smart home system that leverages the strengths of Arm Cortex-A architecture. The system can achieve 15+ tokens per second inference speeds using optimized models like TinyLlama and Qwen, while maintaining the energy efficiency that makes Arm processors a good fit for always-on applications.
+You will create a fully local, privacy-first smart home system that leverages the strengths of Arm Cortex-A architecture. The system can achieve 15+ tokens per second inference speeds using optimized models like TinyLlama and Qwen, while maintaining the energy efficiency that makes Arm processors well suited for always-on applications.
 
-## Why Arm Cortex-A for Edge AI?
+## Why Arm Cortex-A76 makes Raspberry Pi 5 ideal for Edge AI
 
 The Raspberry Pi 5's Arm Cortex-A76 processor can manage high-performance computing tasks like AI inference. Key architectural features include:
 
-- The **superscalar architecture** allows the processor to execute multiple instructions in parallel, improving throughput for compute-heavy tasks.
-- **128-bit NEON SIMD support** accelerates matrix and vector operations, which are common in the inner loops of language model inference.
-- The **multi-level cache hierarchy** helps reduce memory latency and improves data access efficiency during runtime.
-- The **thermal efficiency** enables sustained performance without active cooling, making it ideal for compact or always-on smart home setups.
+- **Superscalar architecture**: Executes multiple instructions in parallel, improving throughput for compute-heavy tasks
+- **128-bit NEON SIMD support**: Accelerates matrix and vector operations, common in the inner loops of language model inference
+- **Multi-level cache hierarchy**: Reduces memory latency and improves data access efficiency during runtime
+- **Thermal efficiency**: Enables sustained performance without active cooling, making it ideal for compact or always-on smart home setups
 
-These characteristics make the Raspberry Pi 5 well-suited for workloads like smart home assistants, where responsiveness, efficiency, and local processing are important. Running LLMs locally on Arm-based devices brings several practical benefits. Privacy is preserved, since conversations and routines never leave the device. With optimized inference, the system can offer responsiveness under 100 ms, even on resource-constrained hardware. It remains fully functional in offline scenarios, continuing to operate when internet access is unavailable. Developers also gain flexibility to customize models and automations. Additionally, software updates and an active ecosystem continue to improve performance over time.
+These characteristics make the Raspberry Pi 5 well suited for workloads like smart home assistants, where responsiveness, efficiency, and local processing are important. Running LLMs locally on Arm-based devices brings several practical benefits. Privacy is preserved, since conversations and routines never leave the device. With optimized inference, the system can offer responsiveness under 100 ms, even on resource-constrained hardware. It remains fully functional in offline scenarios, continuing to operate when internet access is unavailable. Developers also gain flexibility to customize models and automations. Additionally, software updates and an active ecosystem continue to improve performance over time.
 
-## Arm Ecosystem Advantages
+## Leverage the Arm ecosystem for Raspberry Pi Edge AI
 
 For the stack in this setup, Raspberry Pi 5 benefits from the extensive developer ecosystem:
 
 - Optimized compilers including GCC and Clang with Arm-specific enhancements
 - Native libraries such as gpiozero and lgpio are optimized for Raspberry Pi
-- Community support from open-source projects where developers are contributing Arm-optimized code
-- Arm maintains a strong focus on backward compatibility, which reduces friction when updating kernels or deploying across multiple Arm platforms
+- Community support from open-source projects where developers contribute Arm-optimized code
+- Backward compatibility in Arm architecture reduces friction when updating kernels or deploying across platforms
 - The same architecture powers smartphones, embedded controllers, edge devices, and cloud infrastructure—enabling consistent development practices across domains
 
-## Performance Benchmarks on Raspberry Pi 5
+## Performance benchmarks on Raspberry Pi 5
 
 The table below shows inference performance for several quantized models running on a Raspberry Pi 5. Measurements reflect single-threaded CPU inference with typical prompt lengths and temperature settings suitable for command-based interaction.
 
-| Model               | Tokens/Sec | Avg Latency (ms) |
+| Model               | Tokens/sec | Avg latency (ms) |
 | ------------------- | ---------- | ---------------- |
 | qwen:0.5b           | 17.0       | 8,217            |
 | tinyllama:1.1b      | 12.3       | 9,429            |
 | deepseek-coder:1.3b | 7.3        | 22,503           |
 | gemma2:2b           | 4.1        | 23,758           |
 | deepseek-r1:7b      | 1.6        | 64,797           |
 
+## LLM benchmark insights on Raspberry Pi 5
 
-What does this table tell us? Here are some performance insights:
-
-- Qwen 0.5B and TinyLlama 1.1B deliver fast token generation and low average latency, making them suitable for real-time interactions like voice-controlled smart home commands.
-- DeepSeek-Coder 1.3B and Gemma 2B trade off some speed for improved language understanding, which can be useful for more complex task execution or context-aware prompts.
-- DeepSeek-R1 7B offers advanced reasoning capabilities with acceptable latency, which may be viable for offline summarization, planning, or low-frequency tasks.
+- Qwen 0.5B and TinyLlama 1.1B deliver fast token generation and low average latency, making them suitable for real-time interactions such as voice-controlled smart home commands
+- DeepSeek-Coder 1.3B and Gemma 2B trade some speed for improved language understanding, which can be useful for complex tasks or context-aware prompts
+- DeepSeek-R1 7B offers advanced reasoning capabilities with acceptable latency, which may be viable for offline summarization, planning, or low-frequency tasks
 
-## Supported Arm-Powered Devices
+## Supported Arm-powered devices
 
-This Learning Path focuses on the Raspberry Pi 5, but you can adapt the concepts and code to other Arm-powered devices:
+This Learning Path focuses on the Raspberry Pi 5, but you can adapt the concepts and code to other Arm-powered devices.
 
-### Recommended Platforms
+## Recommended platforms
 
-| Platform         | CPU                              | RAM            | GPIO Support                  | Model Size Suitability      |
-|------------------|----------------------------------|----------------|-------------------------------|-----------------------------|
-| **Raspberry Pi 5** | Arm Cortex-A76 quad-core @ 2.4GHz | Up to 16GB     | Native `lgpio` (high-performance) | Large models (8–16GB)        |
-| **Raspberry Pi 4** | Arm Cortex-A72 quad-core @ 1.8GHz | Up to 8GB      | Compatible with `gpiozero`        | Small to mid-size models     |
-| **Other Arm Devices** | Arm Cortex-A             | 4GB min (8GB+ recommended) | Requires physical GPIO pins       | Varies by RAM                |
+| Platform            | CPU                              | RAM            | GPIO support                   | Model size suitability      |
+| ------------------- | -------------------------------- | -------------- | ------------------------------ | --------------------------- |
+| **Raspberry Pi 5**  | Arm Cortex-A76 quad-core @ 2.4GHz | Up to 16GB     | Native `lgpio` (high-performance) | Large models (8–16GB)       |
+| **Raspberry Pi 4**  | Arm Cortex-A72 quad-core @ 1.8GHz | Up to 8GB      | Compatible with `gpiozero`        | Small to mid-size models    |
+| **Other Arm devices** | Arm Cortex-A                    | 4GB min (8GB+ recommended) | Requires physical GPIO pins       | Varies by RAM               |
 
-Additionally, the platform must:
+Additionally, the platform must meet the following requirements:
 
 - GPIO pins available for hardware control
-- Use Python 3.8 or newer
+- Python 3.8 or newer
 - Ability to run [Ollama](https://ollama.com/)
 
-Continue to the next section to start building a smart home system that highlights how Arm-based processors can enable efficient, responsive, and private AI applications at the edge.
+In the next section, you’ll set up the software dependencies needed to start building your privacy-first smart home system on Raspberry Pi 5.