APPLE MAC PRO or The Allenstein "Quadzilla" Machin

[toddm7]

@marspe:


@those wanting to run XP on Mac Pro:
Caution is the word of the day! BootCamp is a BETA. A public one, and one many have had success with. But, just the same, it is a beta.

I'm seeking more info about running LE on a MAC PRO under XP. To clarify, this is being done successfuly? I heard that there are problems with midi in this configuration. I would like to gradualy move my life back over to the MAC side of things. A new MAC pro would be a good place to start. Any links or advice are appreciated.

[Diginerd]

Or 400 D-Verbs on a Dual Clovertown machine..

or Dave-C 14 complete (32+128) with just 62% CPU in use
or Dave-C 15 (I made it up) (48+128) with just 68% in use

By far the fastest thing out there, but as Shan says, even the now "Baby" (Wtf am I typing?) Quadilla pump out more CPU goodness than most people could even dream of using.

What's interesting under OSX on the same machine I get just 220 D-Verbs. :-(

That's not a cheap piece of hardware though...

[nikki-k]

Or 400 D-Verbs on a Dual Clovertown machine..

or Dave-C 14 complete (32+128) with just 62% CPU in use
or Dave-C 15 (I made it up) (48+128) with just 68% in use

By far the fastest thing out there, but as Shan says, even the now "Baby" (Wtf am I typing?) Quadilla pump out more CPU goodness than most people could even dream of using.

What's interesting under OSX on the same machine I get just 220 D-Verbs. :-(

That's not a cheap piece of hardware though...

I had a post written, but remembered I am supposed to behave now...so..

Please do not post something without a disclaimer when it is untrue.

Edited to say- I am sorry to Diginerd for a statement that could be construed as calling him/her a liar. It was not meant to. I would like to take back "untrue" and substitute "invalid" for the word "untrue" please. And since I AM a cripple, I can use the word "invalid" with either pronunciation, so no flack on that please.

I am not saying the poster did not have the expeirences he/she did. I am, however, cautioning anyone using this thread as serious consideration when purchasing a computer and making an OS decision to collect as much info as they can, and wait for other quad core reports to come on.

Both OS's are great, with XP exhibiting (IMO!) more stability with PT right now, but OS X having better memory handling. Ignoring the new Intel quad core CPU's, and sticking to the dual core chips, both camps get you an excellent machine, and equal speeds across platforms exhibit close enough performance to consider either one.

I am not the be-all-end-all on this. Take my words with a grain of salt please; but- above all- PLEASE shop intelligently, and do not be taken by something that offers something unreasonable. If you think you are getting that Ferrari for $10K, you are in for some serious tears and anger....

[Diginerd]

What you mean untrue?

I have it sat under my desk and running EXACTLY as described.

This is in no way a troll. I am 100% genuine in this post.

What would you like to see to prove that it's real?

I'm happy to answer any questions about it too.

[Vox Digital]

That's 100% true. I helped him to choose the motherboard and other components. This machine is using two 2.4GHz quad-core Clovertowns. Should we kindly call it "Octazilla"?

[Diginerd]

Hmm, maybe I did lie or at least not check my facts fully...

Dave C reported results were under XP. This is built as a Windows PC First, but I can also run OSX

Under OSX 10.4.8 it's about 50% give or take 1% fluctuation..

This is at 512K Buffer.

Here's a screen grab straight from "Grab". If you want to poke at the meta data you'll see it's not bee touched by any other app. I'm genuine.. Next it how to work out how to show a video of it..

http://www.free4up.com/ShowImage.asp...219137097.tiff

Enjoy..

[Diginerd]

Questions?

Sorry, I hate being called a liar..

[nikki-k]

OK...that is what I mean- it is not a Mac Pro. You have a "hacked" copy of OS X running on a BIOS'd machine. You will not see accurate results, and the reports you make can be VERY ('scuse me, make that VERY) misleading. So, before even comparing, the results, as presented and compared, are "untrue."

First of all, under OS X, the Dverb test should (and will) produce a few more instances, at the very least. It is due to the memory handling in OS X over XP. So, something in your venture is seriously flawed. Therefor, the results are "untrue." You have results from a test within conditions you set up. True. No problem. But to say that "XP produces 400 Dverbs, OS X with the same processor will produce 220" is very wrong.

You can have 400 Dverbs. Set up a video camera and tape the test being done. THAT is proof. A 2.4 quad core, dual CPU system on XP should be getting about 350-370 Dverbs. A 2.66 version should be getting about 360-380. IF that logic actually came to be true, then on OS X, the count should be 3-5 more than XP.

I was one of a very small group that pioneered hardware for PT and XP. I began the trek to find and build the best, most stable XP machine for HD. So I know the hardware, and I know the OS. I also work with OS X, and due to memory handling issues, it is my choice right now. And it is on a Mac Pro. And if someone can show me actual, unedited video footage of an XP machine with the exact same CPU's as those in a Mac Pro and both having 4G RAM (as is supposed to be properly utilized by any 32-bit OS and it's 32-bit apps) and the XP machine yielding a higher Dverb count, then I will come here and create a thread apologizing. Ask Shane. Ask Matt. They will probably say the same thing I just did.

Now- I never called you a liar. I never said you were intentionally passing off your personal observances as being defacto, 100% accurate truth. I simply am saying that past experiences by MANY users would bring up serious questions as to how you arrived at your results. If I had the time and $$ to waste throwing quad cores in my Mac Pro just to prove this with video footage, I would. But it isnt worth it! All I am saying is...a claim of 400 Dverbs on an XP fuelled machine with dual quad core Xeon 2.4G CPU's and that OS X with the same CPU's only does 220 is seriously flawed, or untrue. It might take seconds to figure out why it is untrue, or it might take months. But, as soon as the quad core dual CPU Macs are out, I guarantee you will see what I am saying as being fact.

Shane? Matt? Anything to add?

(oh- the funny thing? While OS X is excellent with memory heavy apps- such as PT with virtual insturmen ts- I, and many others, agree 100% that XP + PT is more stable. If you know what you are doing, and how to resolve small issues, OS X can be bearable enough for those wanting/needing to use those certain virtual instruments. But, right now, XP is more stable with PT7+. Vista? If MS got it right, and if Leopard does not screw the pooch too badly, we might...just maybe...for the very first time be equal on both sides. It would truly be a matter of which OS one wants to have to stare at)

[Diginerd]

Firstly I never said I had Mac Pro, and never even ((un)intentionally) implied it. I built a PC first and then happened to run OSX on it. I don't want to get into semantics on this. It's OSX, no more different than 10.3 to 10.4, infact much more closely related...

Now, I do disagree with a number of your statements, especially as I feel that you decided to try and go off the deepend with me. You may not think you did, but that is certainly how I and several others I checked in with read it..

So...

+++ You
"You will not see accurate results, and the reports you make can be VERY ('scuse me, make that VERY) misleading. So, before even comparing, the results, as presented and compared, are "untrue."

+ Me
The results are NOT untrue, they are what happen on THIS specific machine. Therefore they are True, not False. They may not be how a Mac Pro Would behave with the Same CPUs in it but that is a DIFFERENT matter, we're talking about THIS machine. I cetainly was...

I'm simply presenting a serious alternative to both the Quadzilla and the MacPro. In addition, I'm not slighting or favoring one OS over the other, I'm just trying the same thing with box, naked and untweaked and comparing results. Personally I want to be running OSX because I'm a longtime Machead on my own time, but I'm not going to let bias my observational data.
+++

+++You
But to say that "XP produces 400 Dverbs, OS X with the same processor will produce 220" is very wrong. =

+Me
Hmm, exact same hardware last time I checked, about 15 minutes between tests. Just long enough to swap out the OS disk on a removable pod. Both OSX and XP are installed on exactly the same type of drive and boot in the same IDE position.

XP SP2 on one test, 10.4.8 on the other. Same version of PT not much more than Pro Tools & VNC installed on either OS. No tweaks to OS by me at all. You might want to "Blame" the kernal of my 10.4.8 but there is some other interesting data which I shall get to in a second...
+++

+++You
"Now- I never called you a liar. I never said you were intentionally passing off your personal observances as being defacto, 100% accurate truth."

+Me
You certainly tried to imply it. And you certainly are trying to take the high ground. Sorry that's the way I read it.

Now my personal observations of this machine are what they are and I stand by them, but I'm open to scientific discourse. Calling me out on my observations if they don't jibe with your own if you ask me is childish an unproductive, but we can resolve this.

Please note, I'm not getting into MacPro discussions at this point. I'm all about this specific machine. A machine which happens to have some pretty sporty hardware. Not least the motherboard bears a very distinct architectual resemblance to the MacPro. Data at the end...

Now where is does lack is that it does only have 2GB configured as 2 x 2 512MB Ram (Dual Channel Interleaved, frequently mislablled as quad channel). I can get more more RAM if you REALLY think it will matter, but I'd like to perform any tests with them machine as it stands right now since that is what I based my statements on.
+++

Now you don't believe any of this of course (can't say I blame you), words are just words, so here's where we try and resolve this..

Firstly, this is the 21st century. No video tapes required...

YOU, YES YOU can perform the tests, hell, if Shane and Allen want to get in on this too I'm open to that too.

I can give you VNC to the machine, when you're done I can swap over to OSX and then you can repeat your tests. If you want I can FRAPs the whole thing from another computer too via a passive VNC session watching the first so that others who are curious can see the results "first hand". Especially if multiple parties certify what they see in the FRAPs is what happened.

Of course you can't hear a thing but given the nature of the tests you wouldn't want to.

I do undertand that VNC consumes some resources, but not that much. My OSX screenshot of the DaveC 14 test was done via VNC.

This brings me to an interesting point. The DaveC performance under OSX looks to be (Unqualified statement here) slightly better under OSX than XP (As you predicted for D-Verbs), but the DVerb performance was much lower, which I too found strange indeed. I'll admit too that I was sorely disappointed with the results I got.

You're not the only one to claim to have a deep understanding of operating systems and hardware architecture. I'll grant that these results indeed appear contradictory, unless there is some emulatated code somewhere.

Now, if you REALLY want to after this first set of tests I can stick the same CPUs into a Mac Pro, but that would take a little more time to get set up. Like I'd have to persuade my friend let me poke at his machine... We'd have to clear this up first.

So, if you play with it and you agree with my initial statements (or at least corroborate them closely) then you back off and not cop such an attitude and accept my words at face value.

Publically.

If I'm wrong post your findings and I'll apologize. Ultimately I feel that this is a very interesting development for high power audio systems.

At least until "The Next Big Thing" comes along...

Deal?

PM me so we can IM and maybe even setup a couple of phonecalls and resolve this sensibly.

Finally here's the details of the machine in question:-


35-114-056 CPU COOLER DYNATRON|H46G
17-153-040 PSU TT|W0128RU 650W Power Suppply
13-151-048 SERVER_MB TYAN|S2696A2NRF (SATA) otherboard
20-161-045 MEM 512M|WINTEC DII667 39C925284C RAM
14-125-043 VGA GIGABYTE GV-RX165P256D-RH Graphics card

CPUS are 2 x 2.4GHz 1066FSB Clovertowns. Yup FSB is only 1066, but they stack up quickly when multi-threading.

Motherboard URL http://www.tyan.com/product_board_detail.aspx?pid=43
Motherboard Block Diagram / Architecture (On P2) ftp://ftp.tyan.com/datasheets/d_s2696_100.pdf
Mac Pro Block Diagram / Architecture http://developer.apple.com/documenta...0906_arch.html

[spkguitar]

1. What you are posting about, running OSX on PC hardware, is illegal according to the Mac OSX EULA. Posting about illegal activities here is against the DUC Terms of Use.
2. OSX was designed to be used on MAC hardware. Nikki-k is right, any results you get using the hacked/illegal version are suspect, as you are not running that software as it was intended to be run. If your results came from a comparable MAC running windows, they might be valid, but until then, it's just bunk. It's not disbelief, it's just not a valid test.