[erlang-questions] Erlang's Performance

[Ivan Carmenates García]

Hi all, I was taking a view of the new stuffs since I had left Erlang, and I
found the observer application, I have to say that’s a very nice tool. But
unfortunately I had a problem with that, the Load Charts tab has a very high
load, it suppose to show the load of other processes and the whole system,
but instead of that it take all the load for it self painting the graphics,
in my case I have an old pc, 512 GB of RAM and just one CPU Intel Celeron
2.7 GHz, Intel D915GL Board, Socket 775.

 

In some messages from you that I just saw, you speak about running Erlang in
very good machines, like 8 GB of RAM and stuffs like that, four CPU cores
and so on. I was old programming in Erlang, I mean with a very set of old
machines, and I was always happy with Erlang because of the performance it
had and has. I had left Erlang for a while, now I’m back and I keep having
my old pc, and now I tried some of the new apps that comes with the OTP, for
example the ‘observer’ a very good tool actually, but I noticed that in the
Load Charts tab it take the 100 % of my CPU and left none to any other
processes. The Erlang console has its problems too with performance,
printing things and that kind of stuffs that consoles does, that the oldest
consoles had not, I observed that in this new release the console is more
slower than for example the v5.7.5.

So my point is, I know that the hardware revolution is now more than ever
growing up, and every day we have double faster computers and in a few years
we will don’t care about doing things so right because computers will
support our laziness. So I think if we want to make the best CO programming
language ever known for a concurrency world, we need to improve every
chain’s link and always look back to the base and optimize there to. Maybe
in the Load Charts of the observer application was used the accelerations
advantages that the Direct X or OpenGL brings, maybe that’s why my old pc
can not enjoy that, or maybe not, and was made just using lineal direct draw
as I think because with new computers it was enough. Well my final point is
that I can not use that part of the observer application in my pc Intel
D915GL, Socket 775 with no graphic card. And so maybe the same can happen
with others applications because we program using the new technologies and
we forget about the real world that is having an Intel Pentium III 800 MHz,
I already test Erlang, with the framework EVO having hundreds of message
passing, two years ago and work very fine in real-time with only 256 GB of
ram on a Pentium III. That’s why I love Erlang. 

 

Best Regards, Ivan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20130302/eee2bc40/attachment.htm>