hello
just planning on buying a generic laptop pc soon, for realtime music applications under linux (pd, ardour) and under windows xp (max/msp), but I need some information / experiences about it (on the internet the main topic I found is this eternal amd/intel debate and I don't really care about this)
regarding the cpu, there are two main technologies
or amd mobile athlon xp): these ones give you the same performance as a desktop computer, but you shouldn't unplug the computer from the external energy source for too long
on mobility (low consumption, embedded WLAN, less heat hence less fan noise, etc) but almost nothing is said about their performance compared to the above category; as these processors work differently (other instructions, other architecture), I would like to know if any of you use such machines for making music in realtime, because I didn't find anything serious about it on the internet
So the question is not "amd or intel ?" but "sempron/centrino or mobile-athlon/- pentium4 ?"
For a given price, I can't make any compromise on calculus performance and I don't plan on using the laptop on batteries too much (anyway my sound card needs external power), so maybe the first category is the best for me, but as there are more and more second-category laptops, I certainly have to take them in account too.
Most probably I will have to choose between amd sempron, amd mobile athlon xp, intel mobile pentium 4 and intel pentium M / centrino. I don't need a huge cpu power but I don't want this laptop to be less performant/reliable than my current 2 years old 2.4 GHz shuttle desktop computer.
notes: I fear 64-bit will be too expensive so I won't focus on it. Also, I don't see clearly the difference between a "mobile xxx processor" and its "xxx processor for notebooks equivalent that you can find in both amd and intel ranges :)
thanks for your help, -j
hi Julien,
my (and others') problems with mobile Pentium4 processors are very well-documented on this list! I simply cannot recommend them for *any* kind of realtime audio processing, especially under Linux! This is due to the "denormal numbers" bug in their architecture. When the mantissa of a floating point number gets X number of places below the decimal point, it kicks into "extremely accurate mode", and there goes your DSP. Calculating the tail signal of a reverb (very very small, almost infinitely quiet sample level) is a classic situation where this occurs. Later generation P4s may have some fixes around this, but the problem still exists.
I got rid of my last P4 latop and got a Athlon64 because if this. If you don't mind a hot, heavy, loud computer (I could blowdry my hair in the morning with mine!) that lasts about an hour (in my case) then you might investigate this. For me, besides the fact that I put an extra kink in my spine every time I go on tour with this machine from carrying it around, the biggest issue now isn't CPU power. I can't remember the last time I hit bottom with this laptop. It's the noise, and all those desktop CPUs in a small casings will have this problem. So doing any recording work with microphones in the same room is completely out!
good luck, d.
thanks derek and ix for your anwsers
I will certainly avoid the mobile pentium4. Also, that's right you can find some cheap laptops with a 64 bit processor. But most programs are optimized for the 32 bit operating systems currently, so they may run slower on a 64 bit ...
-j
Hi again,
julien.breval@tremplin-utc.net wrote:
thanks derek and ix for your anwsers
I will certainly avoid the mobile pentium4. Also, that's right you can find some cheap laptops with a 64 bit processor. But most programs are optimized for the 32 bit operating systems currently, so they may run slower on a 64 bit ...
Not really true. Lots of (most?) Linux apps seem to run fine on 64bit. The issue here wouldn;t be so much the speed by the compatibility. I don't know how far PD is with 64bit optimizations, for example, and you might still have some problems with binary-only drivers under Linux, so 64bit may or may not be worth your money. I end up running PD in 32bit mode because I had problems loading soundfiles to tables in 64bit. Check the list for postings from thewade about other 64bit issues.
best, d.
hi Derek,
most programs are optimized for the 32 bit operating systems currently, so they may run slower on a 64 bit ...
Not really true. Lots of (most?) Linux apps seem to run fine on 64bit.
Yes, but actually the problem rather comes with ms windows; apart from my linux pd/ardour/scilab music system, I use some windows xp programs like max/msp and have no information about their performance under windows xp64 (in 32bit mode)
The issue here wouldn't be so much the speed by the compatibility. I don't know how far PD is with 64bit optimizations, for example, and you might still have some problems with binary-only drivers under Linux, so 64bit may or may not be worth your money. I end up running PD in 32bit mode because I had problems loading soundfiles to tables in 64bit. Check the list for postings from thewade about other 64bit issues.
I saw this. This looks like one more reason to choose 32 bit. According to what I read across the Internet about processors for notebooks, the Pentium M (centrino technology) looks very good for realtime audio. The AMD equivalent (Sempron) should be good too, but I didn't find much information about it; maybe someone here has a laptop with this type of cpu ?
thanks for your information, -j
I saw this. This looks like one more reason to choose 32 bit. According to what I read across the Internet about processors for notebooks, the Pentium M (centrino technology) looks very good for realtime audio. The AMD equivalent (Sempron) should be good too, but I didn't find much information about it; maybe someone here has a laptop with this type of cpu ?
I have a centrino and in comparison to desktop p4's it's much faster at the same clockspeed. It's said that a 1.7GHz Pentium M (dothan) is faster than a 3.0GHz P4. http://www.tecchannel.de/imgserver/bdb/341800/341851/C2E2D1F2D6C75DD830F7D60...
i've never noticed any denormal problems. and denormals are not a problem when using sse2 anyway. so if you have denormal probs you should try to force the compiler to use sse2 instructions. closed source software which has denormal probs is just badly coded or improperly compiled. (sorry, i'm not sure, if what i've wrote is perfectly true - but polemic is fun) the only processors which are a little faster than the PM2.0Ghz are the P4 3.8Gz and the the Athlon 4000+. This small advantage of additional speed is outweighted by the fact that these cpu are electricity-hungry 100watts monsters. semprons suck as much as celerons. The advantages of the Pentium M's relating wattage and thus battery running time and noise are huge. it's so pleasant working on a silent computer, ahhhh... i love it. everytime i have to work on a normal-loud desktop/laptop i'm wondering how i could stand that noise over the years.
sven.
Hi Sven,
Selon sven ml.sven@subscience.de:
I have a centrino and in comparison to desktop p4's it's much faster at the
same clockspeed. It's said that a 1.7GHz Pentium M (dothan) is faster than a 3.0GHz P4.
http://www.tecchannel.de/imgserver/bdb/341800/341851/C2E2D1F2D6C75DD830F7D60... 78452_1000x700.jpg
it's ok because I currently have a 3-year old 2.4 GHz P4 desktop and no cpu problem :)
i've never noticed any denormal problems.
cool
the only processors which are a little faster than the PM2.0Ghz are the P4 3.8Gz and the the Athlon 4000+. This small advantage of additional speed is outweighted by the fact that these cpu are electricity-hungry 100watts monsters.
moreover, it looks like 64 bit is not always the best choice today, and the mobile version of the P4 doesn't look good for audio processing (see the discussions about the denormal problem in the pd mailing list, for example)
semprons suck as much as celerons.
ok, so I was wrong and the AMD equivalent to the pentium M is the Turion, but it's a 64 bit processor, and it looks wiser to setup a 32 bit system currently, especially because I will use it with both linux and windows for audio.
The advantages of the Pentium M's relating wattage and thus battery running time and noise are huge. it's so pleasant working on a silent computer, ahhhh... i love it. everytime i have to work on a normal-loud desktop/laptop i'm wondering how i could stand that noise over the years.
well, I have a shuttle desktop and it's ok :) but a M/Turion laptop is better for this of course
thanks for your information, -j
Hallo, julien.breval@tremplin-utc.net hat gesagt: // julien.breval@tremplin-utc.net wrote:
ok, so I was wrong and the AMD equivalent to the pentium M is the Turion, but it's a 64 bit processor, and it looks wiser to setup a 32 bit system currently, especially because I will use it with both linux and windows for audio.
I might add, that running AMD64 processors at 32bit is still very fast. Somehow Semprons are (kind of) AMD64 processors without the 64Bit part (and without Cool'n'Quiet), and AMD took quite some effort to optimize 32bit performance, if alone because Windows64 still has some way to go. I still run my new A64 3000+ in 32bit mode, because I didn't have time to make the switch, and actually I don't really see the need to have it running with 64bit and be the coolest kid on the block immediatly.
Nevertheless for a laptop I would not buy a AMD processor but some Centrino system, which has much lower power consumption while still performing very well. Also the newer Intel GMA graphics chipset seems to perform very well and they even come with Open Source 3D drivers. I don't think, there are any AMD laptops with free (as in software) 3D drivers.
(I think, these Samsung X20 notebooks look mighty attractive.)
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41
notes: I fear 64-bit will be too expensive so I won't focus on it. Also, I
x86_64 is not more expensive than x86, there are a number of Turion notebooks from ASUS, HP, et al. this is my choice, since it allows for binary-sharing and distcc with a shuttle, but if you dont care about 64bit, pentium-M is also great, and runs pd quite a bit faster using less energy than a comparably-mhz'd athlon32 or P4..
didnt try the latest pd CVS. does it fix the 64bit incompatibilty issues?
don't see clearly the difference between a "mobile xxx processor" and its "xxx processor for notebooks equivalent that you can find in both amd and intel ranges :)
thanks for your help, -j
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
x86_64 is not more expensive than x86, there are a number of Turion notebooks from ASUS, HP, et al. this is my choice, since it allows for binary-sharing and distcc with a shuttle, but if you dont care about 64bit, pentium-M is also great, and runs pd quite a bit faster using less energy than a comparably-mhz'd athlon32 or P4..
well, i'd go for a pentium-m unless you like fan noise ;-)
didnt try the latest pd CVS. does it fix the 64bit incompatibilty issues?
unless i really missed somethin, pd is broken on 64bit hardware ... if i'd have some time / get my hands on an 64bit machine, i'd try to do something ... so the only thing to do is waiting ... 1, 10 or 100 years ... hopefully pd will support 64bit machines before 128 bit processors become common ...
t
Hallo, Tim Blechmann hat gesagt: // Tim Blechmann wrote:
unless i really missed somethin, pd is broken on 64bit hardware ... if i'd have some time / get my hands on an 64bit machine, i'd try to do something ... so the only thing to do is waiting ... 1, 10 or 100 years ... hopefully pd will support 64bit machines before 128 bit processors become common
At least it builds - debian-amd64 has 64bit packages of Pd: http://debian-amd64.alioth.debian.org/debian-pure64/pool/main/p/puredata/
;-)
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41