AMD vs Intel for music program plugins

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Figured you were watching all this from the sidelines!! Something else to keep in mind ... do you think Cubase (Steinberg) isn't thinking about 64-bit processing and dual processor support? Of course they are!! I just read from Steinbergs site "Cubase SX3 is the first version of Cubase to support a 64-bit operating system". Do a Google search on "Cubase 64bit" and see for yourself.

Sonar just dropped Version 5 which now (in addition to the dual processor support) has 64bit OS support. There is NO WAY Cubase (or any other Audio Processing App) isn't moving towards supporting dual processor support and 64bit OS support. Consider purchasing your rig to support the direction Cubase is going, not where they are now. No bashing, just some advice. Again, good luck!!!
 
do you think Cubase (Steinberg) isn't thinking about 64-bit processing
Click to expand...
Gee, it's too bad the A64s cant do 64 bit.
I hear that Intel is planning on using all of the A64s 64 bit enhancements, in thier upcoming chips. Maybe he should wait. I also hear that the next EE will run @ 3.46ghz, so it will be as fast as an X2 4800. Maybe he should wait for that to.
Listen Intel fanboy idiot, stick around, we do get some of your kind here. They would love to have some backup. In the mean time, read and learn. In the end, like a lot who have gone before you, you may end up a performance fanboy instead.
 
I know it's hard not to bash on anyone who would even consider the strengths and weaknesses of both platforms, let alone hear someone suggest something other then AMD, so it doesn't surprise me that in the same way AMD outruns the competition, you (in your haste) flew right past what was said in my post. If you go back and re-read it, you'll see that there is no mention of Intel or AMD. The whole point of the post was for the guy to consider a platform that would handle a multi-threaded App vs. a single threaded one because the App he uses (Cubase) was headed in that direction.

Running in an overclocked, "bash the Intel fanboy" mode often causes one to get an "f" in reading comprehension.

Actually, your "over aggression" here isn't really a bad thing. Everybody needs an area where they can let off a little steam and it's better to play "macho man" here then to abuse your kids, beat your wife, etc. Overall, I'm sure NOBODY takes that kind of stuff to seriously anyway...
 
You shove AMD anyones throat no matter what the scenario is.
Click to expand...
I'm sorry, but I just <i>had</i> to quote that one. Classic, coming from the almighty fugger, accusing others of exactly what he himself does, only for a different company.

---
<pre> (\_/)
|~~~~~|======
|_____| This was bunny. He was tasty.
/\/\/\/\</pre><p>
 
Actually, Cubase SX3 with 64bit support has been out for a year already. It's just a shame the newer version of Cubase SL, which is what I use, does not support 64bit. Basically, it's looks like over a grand to go full 64bit for me right now. However, I do think I want to move in the right direction, and get a 939 mobo for now, and see what I can afford later.
 
Well my friend, he has shown a good deal of forward looking capability in his s939 choice. If you were questioning his choice, it shows that you can only be an intel fanboy, with his head up his butt in relation to what does what better.
Untill recently, I have promoted Intel for heavy video encoding. For those on a low budget, that is still the way to go. ( A P4c 3.2 if you can find one at a resonable price, if not a prescott @ 3 ghz wont melt too much of your system) For those who only want to play the OCing game, like our own Mozz, the prescotts are great, while offering a real chalenge. There are a few other instances where Intel will suffice, and when they arise, I will recommend the best setup for the situation. I am also looking forward to late 06, when Intel will solve thier major heat problems. In other words, the best tech for the user. Try it, you might like it.
 
Hey,

If you do build around XP-64, keep in mind that there aren't that many 64-bit device drivers out there!! I think I remember you mentioning you used an M-Audio hardware interface, well they do have some 64-bit (beta) drivers for their firewire interfaces (I have a FW410) and the beta driver works fine, but my USB controller (another M-Audio product) won't work cause the drivers don't exist yet (for some reason M-Audio baked 64 bit beta drivers for some devices, but not for others). Neither will my CD/DVD burner. I think I'll wait it out rather then roll back because after all, the primary function of my machine is music, so I can Limp until the drivers appear...
 
Indeed there is no 64bit driver for my USB Audiophile. But then again, neither is my current music program: Cubase SL2. I think I still want to look ahead and upgrade to a 64bit processor, which still runs 32bit programs just fine. Then I can upgrade the others as cash flow allows.
 
Robles!!

There is an interesting article in this months "Sound on Sound" magazine which described PCI-Express related problems for musicians when running a system with nVidias nForce4 chipset. They mention that users have experienced crazy processor overhead and (in order to remove sound glitches) had to boost their audio buffer size up to 2048 which introduces latency. It says its been reported by users running Cubase SX, Nuendo, and Sonar with users using a soundcard plugged into the PCI slot (Although RME has reported similar problems with their firewire interface). They mention that it has something to do with the PCI-Express x16 vidio card interferring with bandwith. PLEASE if you can get the issue and look at it. Here's a link to the webpage, but unfortunately you'll have to either subscribe online or purchase the issue cause they don't let you read it unless you're a subscriber.

http://www.soundonsound.com/index.php?section=/&url=/search&Keyword=PCI+Express

They mention only the AMD systems having this problem, but thats probably due to the fact that most all of the AMD boards use that chipset.

Hope you haven't purchased anything yet...
 
Well, I could see why you might have problems. The soundcard driver (in memory, which is off the North Bridge/SPP) has to talk to the physical soundcard (off the PCI Bus, which is connected to the South Bridge/MCP). What they are saying is that the bandwidth/cycles being used by an x16 graphics adapter (theoretically 80Gbps but “real-world” is more like 60) is interfering with the timing going on between the sound card driver in memory and the actual card itself. That’s why they suggest increasing the buffer, which allows the timing between the sound card and the driver to be a little “more loose” (like streaming audio from the internet), but that causes latency, especially when you’re working with soft synths/midi controllers. Playback will be somewhat smooth, but you’ll have to wait ½ a second from the time you hit a key on you controller until you hear the sound. One solution would be to NOT get a PCI-Express video card cause it isn’t really necessary for music production, but I wonder if you can make the video card step down to x1 or x4?

<P ID="edit"><FONT SIZE=-1><EM>Edited by StarFoxTB on 09/17/05 04:56 AM.</EM></FONT></P>
 
Intel is still superior in multimedia and threading over the X2, be sure to read between the lines to get the damn thing to work correctly.
Click to expand...
Once again, the All Mighty "FUGGER" knows what he's talking about.

1) AMD's dual core are superior when it comes to multitasking and executing multithreaded apps becasue the K8 architecture was developed with daul core in mind and it's not a "HACK" like what Intel did. The EE840 is good at handling 4 tasks, but whenever you try to do more than just 4 tasks, the EE840 struggles (look at Tommy's "stressed" test, that's not the case with the X2 4800 thanks to it's superior desing. :)

2) You're not providing ANY links to backup your bullshit. Until now, all your facts are just a pile of bullshit; and please don't kickback to me saying that I didn't provide any links. If you want them, just search my name in this forum and you'll have ALL the info you need (if you still want more proof or links of what I'm saying, just say the magic word. 😀

3) Intel's dual core offerings are a joke until they release their "true" dual core implementation (i.e. Yonah or whatever).

4) You like call people fanboys when you're one of the biggest of all Intel fanboys. I admit being an AMD fanboy since I found that the Athlon 700MHz was a lot better than a P3 800MHz and I don't have a problem with that. 😀

Anyhow, this topic is a good example of how flaming can be achieved thorough polite ways like comparing which app benefits more on which processor, etc.

Cheers. 😉

My Beloved Rig:

ATHLON 64 FX 55
2X1024 CORSAIR XMX XPERT MODULES
MSI K8N DIAMOND (SLI)
2 MSI 6800 ULTRA (SLI MODE)
OCZ POWERSTREAM 600W PSU<P ID="edit"><FONT SIZE=-1><EM>Edited by Bullshitter on 09/17/05 12:00 PM.</EM></FONT></P>
 
While other are debating over AMD vs Intel bullshit again, I just want to let you know that, any modern CPU will do better than your old Athlon 1.2 GHz. Be it Intel or AMD. In your case, it is not only the CPU that old you back, but probably all of the system itself. So, if you get a new board with fast DDR memory, fast CPU and faster HDD controller and HDD, you'll be fine.

Unless apps are optimized for dual core or HT, there is no t much advantage going right now for them. Intel HT might give advantage for multitasking, but not with plug-in. Plug-in actually stop the main program to do their task. See, you cannot have a 2 plug in oe the plug-in and the main app working on the same data. thus, HT or dual core is no use there. Even if you want 2 apps working on the same data or files, it cannot. one will lock the file for itself leaving the other waiting.

In your case, a single core CPU with a fast HDD, like the Raptor will help you. For the price of the fast dual core, you can have a fast single core with a fast HDD. get lot of RAM. 2 gigs will improve thing, as data is moved faster fro and to memory than from and to hdd. I dont know how much memory your old system has, but anything under 1 gigs for the kind of work your are doing is going to slow down the whole thing because of the system might have to use virtual memory.

So, no matter which system you'll go, just get faster system, not just faster CPU. Since you are doing music, I bet that you'd like to have somewhat a silent system. AMDs are pretty silent and I recommend them for MM stuff. I know that intel are sometime faster at some MM task, but depending of the apps, codecs, .. you use, both CPU will come to par for peformance at the end of the day. Fact is, CPU are way faster than the rest of the system that they are constantly waiting for user input, data to come,... that I dont recommand system based on performace of the CPU alone. Right now, I recommend more based on noise, heat and power required to make the whole system to run. And now, AMD has the advantage.

So, just get something that is modern, whenever AMD or Intel and I'm sure that you'll be more than happy!

(\__/)
(='.'=)
(")<b>ώ</b>(")
 
Well, you're 50% right!! (But you are right...).

nVidia has two flavors of the nForce4: The Single-Chip, AMD supported version, and the Intel (2 chip) version. The Single-Chip setup (which is by the way a totally better design then the "2-chip" seeing that AMD has the built-in menory controller), is the platform where the problem is being reported. Not sure why, but based on the number of people and hardware soundcard vendors who have reported the problem (check the article), there is SOME type of problem. Although I do have the D840/955 chipset, I have no problem admitting that the AMD dual core complemented by the nForce4 looks like a way better design.

And it will be (for music production) once they figure out this little problem. Music production needs to be seemless, flawless, glitchless, jitterless ... all that. Either it works perfect, or it doesn't work at all.

Like I said a few posts ago, maybe the temp solution is to get a non-x16 video card because the you really don't need it for music...

Once they get the problem resolved, that solution will look way-more attractive and I wouldn't doubt that being my next build, but for now ... My D-840/XP64/Sonar/M-Audio bundle works smooth so I'm happy...
 
If someone would have described to me the problem and then asked me which chipset would mostlikely produce it (a 1 chip vs. 2), I would have picked the 2 chip because of what you described (Hypertransport latency between the two chips).

But even with the 2 chip package, I think they (nVidia) allows 1.6GBps thru that channel (The Intel chipset offers alittle more, but they have to because they have some PCI-x lanes connected to the south chip). A 2 channel audio stream (stereo) from the audio driver in memory to the audio hardware device (PCI card, firewire interface, etc.) isn't going to consume anything close to that. That's what makes the whole problem hard to understand.

<P ID="edit"><FONT SIZE=-1><EM>Edited by StarFoxTB on 09/17/05 07:45 PM.</EM></FONT></P>
 
(theoretically 80Gbps but “real-world” is more like 60)
Click to expand...
Check your numbers, you are way off base.
BTW, if 1 16Xpci-express is eating up all the bandwidth, and there are a total of 32X lanes, pci-express would be a total no go. Think about it.
 
There is an interesting article in this months "Sound on Sound" magazine which described PCI-Express related problems for musicians when running a system with nVidias nForce4 chipset. They mention that users have experienced crazy processor overhead and (in order to remove sound glitches) had to boost their audio buffer size up to 2048 which introduces latency. It says its been reported by users running Cubase SX, Nuendo, and Sonar with users using a soundcard plugged into the PCI slot (Although RME has reported similar problems with their firewire interface). They mention that it has something to do with the PCI-Express x16 vidio card interferring with bandwith. PLEASE if you can get the issue and look at it. Here's a link to the webpage, but unfortunately you'll have to either subscribe online or purchase the issue cause they don't let you read it unless you're a subscriber.

http://www.soundonsound.com/index.php?section=/&url=/search&Keyword=PCI+Express

They mention only the AMD systems having this problem, but thats probably due to the fact that most all of the AMD boards use that chipset.

Hope you haven't purchased anything yet...
Click to expand...
Thanks kindly for the tip. No I haven't bought anything yet, fortunately. It seems like the way to go if I want an nForce4 is to get a standard PCI video card. I don't need super flashy video for my music computer.

Further reading on the problem can be found:

<A HREF="http://www.cubase.net/phpbb2/viewtopic.php?t=13009&highlight=usb+audiophile+latency" target="_new">http://www.cubase.net/phpbb2/viewtopic.php?t=13009&highlight=usb+audiophile+latency</A>
<A HREF="http://www.rme-audio.com/english/techinfo/nforce4_tests.htm" target="_new">http://www.rme-audio.com/english/techinfo/nforce4_tests.htm</A>
<A HREF="http://www.recording.org/ftopic-26210-30.html" target="_new">http://www.recording.org/ftopic-26210-30.html</A>

It would seem not only AMD but Intel 915/925 have this problem, though the Intel apparently works better with a low end PCI-E video card.

I definitely want low latency, and I'm going to be pushing my CPU heavily with a couple of FM softsynths I designed which chew up processing.
 
Those links you've posted ... It's official, there’s a big problem.

The good news is that it's a big enough problem that it's going to get a lot of attention so I'm sure it will be resolved within 6 months or so. Until then, get a cheap but functional PCI video card, and life is good!

I hope everyone here (ENDYEN/WUSY) goes and takes a look at those links.

Both of you two have a few things in common and it's unfortunate to see any of these things in Grown Men.

(1) You both declare war against anyone who would attempt to be so kind as to offer information that may be helpful to anyone else here, as if you have to protect your "throne of knowledge".

(2) When you feel someone is incorrect, instead of expressing your point of view in a helpful way, you immediately throw an insult, state your disgust, then back away to see who would "Dare" to engage your invitation to a verbal challenge.

It is so sad to see two grown men being so immature. Both of you (and probably more on the sidelines watching) have the ability to really help people if you just "freely" share what you've been "Blessed" to have learned. You don't have to challenge people with your knowledge to get respect. People will respect you because of how you unselfeshly give. So many people will never, ever gain the knowledge that you guys have. It’s a blessing and a gift, so be thankful and help someone who is unable to figure it all out! I'm not mad at anybody, I just see your potential...
 
I suppose that meager attempt at slander is the best you can muster.
Why are you offended that I asked you to use your brain?
Like most sites of this type, we often find a need for deductive reasoning. If you look at what you have said about this "problem" in a realitic fashion, you will see that there is no way that it could possibly be a bandwidth problem. Think about it. When a high end pci-exp card is involved in a heavy 3D app, it will use as much as 1/4 of the bandwidth available to pci-express. Do you know anyone who runs cubase while playing fartcry? The pci bus is a seperate entity, with it's own bandwidth. The only common ground would be the FSB, or HTT. If a simple pci device could so cripple those buses, in conjunction with a pci-express graphics card,just think of the result if you used more bandwidth sucking devices like gigabit lan cards or SCSI interfaces.
I guess for some people, you just cant expect them to use thier own wit.
 
Why not just get an older AGP-based board and avoid the problem entirely? :O

:evil: یί∫υєг ρђœŋίχ :evil:
<i>The <font color=red>Devil</font color=red> is in the details.</i>
@ 198K of 200K!
 
I only know of one mobo that supports AGP and the new X2 CPUs and it has a ULI chipset. It was reviewed here at THG as being a bit slow. I'd rather get the faster board and use a standard PCI video card.

Note to wusy: Cubase is a multi-track sequencing/recording program, not a soundcard. For the work I do, latency is crucial. They say the problem doesn't occur at latencies 500ms. I'm using an Athlon 1.2Ghz and I have complete stability at 30ms while using 6 or more software synths. No way am I downgrading to half a second latency.
 
Slvr, I hade hoped for a little better from you.
It sounds like one of those coding for dual core problems you are often talking about.
So, is cubase having trouble finding the cache, so it's using a pagefile instead?
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom