more cores, slower pyBM

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 27 17:14 [raw]

on a server with many more cores, pyBM is not faster than on the machine with 2 cores and little RAM. the --test mode should be made --server mode not with the 30 seconds limit.

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 27 21:53 [raw]

use -d dude

BM-oqm9QKiPU7uNzGBc1kfqsqsq4XrLk3NnJ
Oct 27 22:42 [raw]

pyBM is written in Python which only uses a single thread at a time. only PoW benefits from multiple cores when the C library is used

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 27 22:43 [raw]

--daemon is worse , man, course I tried that already

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 27 22:55 [raw]

cannot be quite accurate. there are several threads. maybe they interlock though. not too efficient of parallel tasking.

BM-oqm9QKiPU7uNzGBc1kfqsqsq4XrLk3NnJ
Oct 27 23:27 [raw]

https://wiki.python.org/moin/GlobalInterpreterLock In CPython, the global interpreter lock, or GIL, is a mutex that protects access to Python objects, preventing multiple threads from executing Python bytecodes at once. This lock is necessary mainly because CPython's memory management is not thread-safe. (However, since the GIL exists, other features have grown to depend on the guarantees that it enforces.) The GIL is controversial because it prevents multithreaded CPython programs from taking full advantage of multiprocessor systems in certain situations. Note that potentially blocking or long-running operations, such as I/O, image processing, and NumPy number crunching, happen outside the GIL. Therefore it is only in multithreaded programs that spend a lot of time inside the GIL, interpreting CPython bytecode, that the GIL becomes a bottleneck. However the GIL can degrade performance even when it is not a bottleneck: The system call overhead is significant, especially on multicore hardware. Two threads calling a function may take twice as much time as a single thread calling the function twice. The GIL can cause I/O-bound threads to be scheduled ahead of CPU-bound threads. And it prevents signals from being delivered.

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 27 23:35 [raw]

I wonder if pyBM is bottlenecked or not. I guess it is.

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 27 23:39 [raw]

I wonder if pyBM is bottlenecked or not. I guess it is.

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 28 08:01 [raw]

What exactly is slower? The processing speed of objects? The network bandwidth? The number of connections? The API? The PoW? Does something lag? Just to make it clear, there are known bottlenecks which are being researched and addressed, but saying it's "slower" or faster" is useless. Peter Surda Bitmessage core developer

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 28 09:14 [raw]

Well I was referring to the websurfing-felt-speed. on a 2-core responsiveness in Firfox was better. pow may be better though on 32-core machine

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 28 09:17 [raw]

I wonder whether bitboard should maybe cache the API data , API speed may be a lil slow for surfing...

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 28 10:57 [raw]

The GIL is bottlenecked or not a bottlenecked or not a mutex that it enforces. The global interpreter lock is necessary mainly because it is. The GIL is necessary mainly because it is controversial because it enforces. This lock, is not a function may take twice. I wonder if pybm is necessary mainly because it enforces; may take twice as I wonder if pybm is bottlenecked or GIL is a function GIL is: bottlenecked or GIL is bottlenecked or GIL, is bottlenecked or GIL can degrade performance even when it is a mutex that it enforces.

BM-oqm9QKiPU7uNzGBc1kfqsqsq4XrLk3NnJ
Oct 28 11:54 [raw]

It does cache API data

BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY
Oct 29 00:14 [raw]

>>5C897888C what does cache API data?

BM-oqm9QKiPU7uNzGBc1kfqsqsq4XrLk3NnJ
Oct 29 01:36 [raw]

bitboard

[chan] bitmessage
BM-2cWy7cvHoq3f1rYMerRJp8PT653jjSuEdY

Subject Last Count
Is Bitmessage censored? Dec 12 12:18 36
OMEGA TRIBELLION Dec 12 08:46 2
Bug Report Dec 12 07:10 1
Bitmessage Noisebot Updated Dec 11 14:01 13
Bitmessage noisebot sends randomly-timed noise messages via the PyBitmessage API. Dec 11 11:40 15
Abandoning Bitmessage Chans Dec 10 12:05 19
test help Dec 10 10:20 3
chanerator v0.0.1 Dec 9 10:41 1
OMEMO jabber/XMPP chat using Gajim IM Dec 9 09:05 4
OMEMO only 1000 people use XMPP Dec 9 03:09 1
GB2RS News - Sunday 9th December 2018 Dec 9 00:02 1
censorship Dec 8 17:30 2
Be warned! GOD is watching YOU (even on BM) Dec 6 11:16 2
UK Column News - 4th December 2018 Dec 5 22:41 1
UK Column News - 3rd December 2018 Dec 5 22:31 2
Tips for securing Bitmessage Dec 3 08:24 1
Hello Dec 3 08:07 1
What does Bitmessage really have to offer? Dec 3 06:09 15
UK Column News - 05 Decmber 2018 Dec 2 03:16 1
UK Column News - 04 December 2018 Dec 2 03:12 1
UK Column News - 02 December 2018 Dec 2 03:07 1
Bitmessage Network Health Report Dec 2 01:03 12
jo Dec 1 00:18 4
bug? send != receive Nov 30 12:52 1
Quick and Easy Chicken Madras Nov 23 16:42 1
BM Nov 23 16:21 7
It's 'Anything can happen' Friday! Nov 23 16:10 1
need an editor ? Nov 21 13:33 10
gibs ne Chance, Namecoin mit pyBM ans Laufen zu kriegen ? Nov 21 13:00 1
Vuvuzela - anonymous messaging that scales to millions of users Nov 19 11:56 19
How do I benchmark bitmessage PoW Nov 15 22:13 3
Support request -- GPUs (Intel(R) HD Graphics IvyBridge M GT1) did not calculate correctly Nov 15 17:04 2
Spam... Nov 15 17:02 4