added technotes and MVM compute note

_data/sidebars/technotes_sidebar.yml

@@ -0,0 +1,34 @@
+entries:
+- title: Tech Notes
+  product: cacao - Tech Notes
+  folders:
+
+
+  - title: Overview
+    output: web, pdf
+    folderitems:
+
+    - title: About
+      url: /cacao_technotes_about.html
+      output: web
+
+
+
+
+  - title: Compute Hardware
+    output: web, pdf
+    folderitems:
+
+    - title: Requirements
+      url: /cacao_comphardw_requirements.html
+      output: web
+
+
+  - title: Real-time OS
+    output: web, pdf
+    folderitems:
+
+    - title: Linux Kernel
+      url: /cacao_RTlinux.html
+      output: web
+

_data/topnav.yml

@@ -20,6 +20,10 @@ topnav_dropdowns:
       folderitems:
         - title: cacao by example
           url: /cacao_examples.html
+    - title: Tech Notes
+      folderitems:
+        - title: Tech Notes
+          url: /cacao_technotes_about.html
     - title: Contributing
       folderitems:
         - title: Writing Documentation

pages/cacao/cacao_comphardw_requirements.md

@@ -0,0 +1,42 @@
+---
+title: Compute Hardware Requirements
+keywords:
+last_updated: May 17, 2024
+tags: [CPU, GPU]
+summary: "Hardware Requirements: compute bandwidth needed to close the AO loop."
+sidebar: technotes_sidebar
+permalink: cacao_comphardw_requirements.html
+folder: cacao
+---
+
+
+## 1. Matrix-Vector Multiply (MVM) and GPU/CPU Specs
+
+The most compute-heavy operation in closing the AO loop is often the matrix-vector-multiply (MVM) converting the input WFS pixel values to output wavefront modes. This MVM must be completed in a fraction of the AO loop period, typically well under 1 ms.
+
+
+The MVM is most often memory-bandwidth limited, so when choosing the compute hardware (for example GPU), the device's memory bandwidth is the most relevant parameter. This is described in [this MVM technical note](https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html), where N=1 (Matrix-vector multiply, special case of matrix-matrix multiply), K is the number of WFS elements, and M is the number of modes reconstructed, or in zonal control, the number of DM actuators.
+
+Taking, for example, a large system with 87k input pixels, 33k output modes :
+
+```
+M=33k
+K=87k
+N=1
+
+Assuming FP16 input, FP32 accumulation
+
+For each MVM:
+Compute load : 2.9 GFLOP
+memory load : 2.9 GB
+
+Arithmetic intensity ~ 1 (need one FLOP per byte)
+```
+
+
+
+As of today (year 2024), current GPU have memory bandwidth of approximately 2 TB/s (note this is terabytes, not terabits), and have compute bandwidth of about ~200 TFLOPS. Comparing these specs with the requirements derived above reveals that the MVM will be memory bandwidth limited, not compute bandwidth limited.
+
+In this example, the MVM would take 1.45 ms (700 Hz maxmum AO frame rate).
+
+{% include links.html %}

pages/cacao/cacao_technotes_about.md

@@ -0,0 +1,19 @@
+---
+title: cacao
+keywords: 
+last_updated: May 17, 2024
+tags: [getting_started]
+summary: "cacao Tech Notes"
+sidebar: technotes_sidebar
+permalink: cacao_technotes_about.html
+folder: cacao
+---
+
+
+Tech notes relevant to Adaptive Optics and cacao.
+
+
+
+
+
+{% include links.html %}