Author: Antonio Cuni <[email protected]>
Branch: extradoc
Changeset: r5796:75d41d3731d6
Date: 2017-04-06 11:54 +0200
http://bitbucket.org/pypy/extradoc/changeset/75d41d3731d6/
Log: write most of my talk
diff --git a/talk/ep2011/talk/beamerdefs.txt b/talk/ep2011/talk/beamerdefs.txt
--- a/talk/ep2011/talk/beamerdefs.txt
+++ b/talk/ep2011/talk/beamerdefs.txt
@@ -93,7 +93,6 @@
:scale: 15%
-
.. nested blocks
.. ===========================
diff --git a/talk/pycon-italy-2017/asyncio.png
b/talk/pycon-italy-2017/asyncio.png
new file mode 100644
index
e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..2f8496364fe82b6a6f662f5e0d4b6d415f436ccd
GIT binary patch
[cut]
diff --git a/talk/pycon-italy-2017/beamerdefs.txt
b/talk/pycon-italy-2017/beamerdefs.txt
--- a/talk/pycon-italy-2017/beamerdefs.txt
+++ b/talk/pycon-italy-2017/beamerdefs.txt
@@ -93,6 +93,9 @@
:scale: 15%
+.. |ok| image:: ok.png
+ :scale: 25%
+
.. nested blocks
.. ===========================
diff --git a/talk/pycon-italy-2017/crossbar-io.png
b/talk/pycon-italy-2017/crossbar-io.png
new file mode 100644
index
e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..a6c7b970ea478c956920b1f4c34a8c198a7434f3
GIT binary patch
[cut]
diff --git a/talk/pycon-italy-2017/gambit.png b/talk/pycon-italy-2017/gambit.png
new file mode 100644
index
e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..f361b3e3c647f4ac1b511be0b1bb313968788ee9
GIT binary patch
[cut]
diff --git a/talk/pycon-italy-2017/magnetic.png
b/talk/pycon-italy-2017/magnetic.png
new file mode 100644
index
e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..d35a0e3a0bce5944727ed7529b486a57ed753699
GIT binary patch
[cut]
diff --git a/talk/pycon-italy-2017/speed.png b/talk/pycon-italy-2017/speed.png
new file mode 100644
index
e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..b9fda24cf9599bfc1f69725766c10bab0c421e9f
GIT binary patch
[cut]
diff --git a/talk/pycon-italy-2017/talk.rst b/talk/pycon-italy-2017/talk.rst
--- a/talk/pycon-italy-2017/talk.rst
+++ b/talk/pycon-italy-2017/talk.rst
@@ -4,13 +4,239 @@
PyPy Status Update
================================
-Hello
+About me
+---------
+
+- PyPy core dev
+
+- ``pdb++``, ``fancycompleter``, ``capnpy``, ...
+
+- Consultant, trainer
+
+- http://antocuni.eu
+
+
+What is PyPy
+---------------
+
+- Alternative, fast Python implementation
+
+- Performance: JIT compiler, advanced GC
+
+- PyPy 5.7.1
+
+ - x86, ARM, s390x, PPC
+
+- http://pypy.org
+
+
+Speed
-----
-bla bla
+.. image:: speed.png
+ :scale: 23%
+ :align: center
-World
+- 7.4 times *faster* than CPython
+
+- http://speed.pypy.org
+
+
+
+Usual questions
+----------------
+
+- **Is it stable/mature? Can I use it in production?**
+
+- Does package X work?
+
+- What about Python 3?
+
+
+PyPy in the real world (1)
+---------------------------
+
+.. image:: gambit.png
+ :scale: 20%
+ :align: left
+
+
+- High frequency trading of sport bets
+
+ * low latency is a must
+
+- PyPy used in production since 2012
+
+- ~100 PyPy processes running 24/7
+
+- up to 10x speedups
+
+ * after careful tuning and optimizing for PyPy
+
+
+PyPy in the real world (2)
+---------------------------
+
+.. image:: magnetic.png
+ :scale: 23%
+ :align: left
+
+- Real-time online advertising auctions
+
+ * tight latency requirement (<100ms)
+
+ * high throughput (hundreds of thousands of requests per second)
+
+- 30% speedup
+
+|small|
+|end_small|
+
+ We run PyPy basically everywhere
+
+ **Julian Berman**
+
+
+
+PyPy in the real world (3)
+--------------------------
+
+.. image:: crossbar-io.png
+ :scale: 23%
+ :align: left
+
+- IoT on the cloud
+
+- 5-10x faster
+
+|small|
+|end_small|
+
+ We do not even run benchmarks on CPython because we just know that PyPy
+ is way faster
+
+ **Tobias Oberstein**
+
+
+Usual questions
+----------------
+
+- Is it stable/mature? Can I use it in production? |ok|
+
+- **Does package X work?**
+
+- What about Python 3?
+
+
+cpyext
+------
+
+- Pure Python pacakges just work (TM)
+
+- C extensions goes through **cpyext**
+
+- Lots of progresses in the compatibility area
+
+- Still slow
+
+ * Crossing the Python/C border takes ages
+
+ * lots of low hanging fruits to be taken :)
+
+- Use CFFI if possible
+
+numpy
-----
-bla bla
+- It just works
+ * ~5 failing tests
+
+- Better GC integration needed
+
+- The GC does not know how much raw memory numpy objects are keeping alive
+
+ * It **looks like** a memory leak
+
+
+Other libs
+-----------
+
+- cython
+
+ * it works
+
+- pandas
+
+ * ~95% of passing tests
+
+ * most failures are because it cannot find Tkinter libs
+
+ * We saw some segfaults
+
+- scipy
+
+ * ~99% of passing tests
+
+ * Numpy memory leaks?
+
+- matplotlib
+
+ * Saving to disk works
+
+ * Jupyter inline works
+
+ * no GUI backends
+
+
+Usual questions
+----------------
+
+- Is it stable/mature? Can I use it in production? |ok|
+
+- Does package X work? |ok|
+
+- **What about Python 3?**
+
+
+PyPy 3.5
+---------
+
+- PyPy 5.7.1 supports Python 3.5 (beta)
+
+- Funded by a big grant by Mozilla (thanks!)
+
+ * |scriptsize|
https://morepypy.blogspot.com/2016/08/pypy-gets-funding-from-mozilla-for.html
|end_scriptsize|
+
+- Still some failing tests
+
+- Very good results on asyncio
+
+
+PyPy 3.5 asyncio
+----------------
+
+.. image:: asyncio.png
+ :scale: 50%
+ :align: center
+
+- Between 2.87 and 8.08 times faster
+
+
+..
+ Topics
+ -----
+
+ - STM / nogil branch :(
+
+ - vmprof
+
+ STM
+ ---
+
+ <arigato> we never really tried to fix all performance issues, and more
importantly:
+ <arigato> if you start with some multithreaded program, it is really slow
+ <arigato> it's a bit depressing because you need to dig and fix conflicts,
and not getting any speed-up at all, until with enough efforts suddenly it's
fast
+ <arigato> in other words, to make it usable we'd need something more
flexible, probably based on much shorter transactions, but that needs rethoughts
+
+
_______________________________________________
pypy-commit mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pypy-commit
