From c196fbf33d35d7e4110f78b400140711328b5fec Mon Sep 17 00:00:00 2001
From: ml server <ml@spacecruft.org>
Date: Thu, 16 Jan 2020 16:00:25 -0700
Subject: [PATCH] Tensorflow Proxmox

---
 README-distributed.md |  7 +++++++
 README.md             | 14 +++++++++++++-
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/README-distributed.md b/README-distributed.md
index 1afbdc2..687c3bd 100644
--- a/README-distributed.md
+++ b/README-distributed.md
@@ -22,6 +22,13 @@ Major steps.
 1. Install Tensorflow.
 1. Set up Ceph.
 
+## Proxmox
+Setting up Proxmox is outside the scope of this document.
+All you really need is some virtual machines, however
+they are created.
+
+* https://www.proxmox.com/en/proxmox-ve
+
 # Usage
 `top`
 
diff --git a/README.md b/README.md
index 72d0b41..d594d2d 100644
--- a/README.md
+++ b/README.md
@@ -91,7 +91,7 @@ installed separately.
 
 The verions of Tensorflow installed with `pip3` on Debian
 Buster crashes. It is perhaps best to do a custom install,
-based preferred build options, of the most preferred version.
+best preferred build options, of the most preferred version.
 At this point, the `remotes/origin/r2.1` branch is preferred.
 
 
@@ -125,6 +125,18 @@ bazel build --config=opt //tensorflow/tools/pip_package:build_pip_package
 pip3 install --user /tmp/tensorflow_pkg/tensorflow-2.1.0-cp37-cp37m-linux_x86_64.whl
 ```
 
+### Tensorflow KVM Notes
+Recent versions of Tensorflow can handle many more CPU build options
+to optimize for speed, such as
+[AVX](https://en.wikipedia.org/wiki/Advanced_Vector_Extensions).
+By default, Proxmox and likely other virtual machine systems pass
+kvm/qemu "type=kvm" for CPU type. To use all possible CPU options
+available on the bare metal server, use "type=host".
+For more info about this in Proxmox, see
+[CPU Type](https://pve.proxmox.com/pve-docs/chapter-qm.html#qm_cpu)
+If you don't have this enabled, CPU instructions will fail or
+Tensorflow will run slower than it could.
+
 ### Tensor Configuration
 ```
 $ ./configure