• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: Distributed build overhead and Mac Mini vs XServe pricing
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Distributed build overhead and Mac Mini vs XServe pricing


  • Subject: Re: Distributed build overhead and Mac Mini vs XServe pricing
  • From: Mark Dawson <email@hidden>
  • Date: Mon, 17 Jan 2005 08:05:16 -0800

It's too bad that RAM disks are not supported on OS X--adding 1 GB of memory to a G5 is less than $150. If I/O disk speed is important, having 1 GB of memory, instead of a drive, should be a lot faster (back in the olden days of OS 7, I remember adding a RAM disk could speed up MPW C++ compiles by 2x, and even with OS 9/G4s, by up to 40%).
It was pointed out that RAM disks ARE supported. However, in a simple test, they gave no benefit.

I set up a 128 MB RAM disk, and copied the "Sketch" example program over, along with the XCode application itself, and the Frameworks that are linked (Cocoa, AppKit, Foundation). I launched Activity Monitor, and at no time during the build process did "Free Memory" get less than 300 MB (I have 1.5 GB installed). In the RAM disk project, I pointed the frameworks (cmd-I, choose) to the RAM disk frameworks. This project creates a precompiled header. This was running on a dual 1 GHz G4 with 1.5 GB memory and 10.3.7, with the XCode 1.5 developer tools installed.

Steps:
(1a) Launch XCode & project from RAM disk
(1b) Launch XCode & project from 7200 RPM UATA 133 drive (boot 10.3.7 drive)
(2) Do a "clean all targets"
(3) Build


Results:
RAM Disk: 16.4 seconds
Hard drive: 15.6 seconds

Multiple runs showed about the same result (hard drive beating the RAM disk by a few tenths of a second).

Maybe this sample project is too simple (too short), but it seems to rule out file I/O as a big factor (at least in relation to already existing RAM). Given that "Free Memory" never got less than 300 MB, I'm assuming that my RAM disk didn't cause any excess page faults, so it was a "fair" test. It would seem that OS X's disk cache negates any negative speed differential of a hard drive.

That could mean that the Mac mini's 4200 RPM drive wouldn't be a factor if only XCode and its tools were running (i.e., there was enough memory for the disk cache to handle all I/O).

Mark

_______________________________________________
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden


References: 
 >Distributed build overhead and Mac Mini vs XServe pricing (From: Andrew Kimpton <email@hidden>)
 >Re: Distributed build overhead and Mac Mini vs XServe pricing (From: Scott Tooker <email@hidden>)
 >Re: Distributed build overhead and Mac Mini vs XServe pricing (From: Andrew Kimpton <email@hidden>)
 >Re: Distributed build overhead and Mac Mini vs XServe pricing (From: Mark Dawson <email@hidden>)

  • Prev by Date: Re: MIDI via X-code
  • Next by Date: Re: XCode, Java, and Eclipse
  • Previous by thread: Re: Distributed build overhead and Mac Mini vs XServe pricing
  • Next by thread: Re: Distributed build overhead and Mac Mini vs XServe pricing
  • Index(es):
    • Date
    • Thread