SSD's in a Sonar DAW: One vs. Two, and Partitioning Impacts

Max Output Level: -75 dBFS
  • Total Posts : 775
  • Joined: 2004/02/26 16:44:40
  • Status: offline
2017/08/14 19:54:20 (permalink)

SSD's in a Sonar DAW: One vs. Two, and Partitioning Impacts

Am rebuilding a new XP Sonar DAW and am considering going the SSD path.  My prior approach was to always go with 2 HD's - one for OS & 1-time-load samples (like drums), and the other drive for audio projects/streaming.  Also, I don't need massive space.  
So I have a couple very specific SSD questions that are key in my mind and am hoping to get some guidance.
- With SSD's, should I still go with two separate drives or would one be ok?  BTW with XP I will be operating in legacy IDE mode, not native SATA. 
- In order for the SSD to properly wear-level for longer life span, I've read you shouldn't take up too much of the free space.  Otherwise the unused area is what gets most abused.  The recommended capacity is apparently 25%-50%, lower the better.  That way it leaves remaining free space for the SSD to work with and cycle new writes across.  Do I have this right?
- And along those lines comes my final question - partitioning.  If you separate an SSD into multiple partitions, does that physically force the SSD to use only a certain area of the memory the way a hard disk would behave, or does the SSD still make use of the entire free memory within the SSD for wear leveling distribution?  
Answers to the above will help me determine how many SSD's, what capacity and partitioning configuration to target.
Many thanks,
*** UPDATES ***
As to the last question, after researching it seems partitioning doesn't have any impact on wear-level algorthms in an SSD. They treat all free space as a common pool regardless of how many partitions the free space may span.
post edited by SonicExplorer - 2017/08/15 01:35:03

3 Replies Related Threads

    Max Output Level: -87 dBFS
    • Total Posts : 200
    • Joined: 2004/04/16 16:33:07
    • Location: UK
    • Status: offline
    Re: SSD's in a Sonar DAW: One vs. Two, and Partitioning Impacts 2017/08/14 20:49:38 (permalink)
    I'd update your research. A lot of these SSD myths have long since been debunked and there's no real reason for not treating an SSD like a regular drive. If you prefer a two up configuration then go with that. Certainly that is my preference but not for any technical reason, I just prefer the separation of OS and Programs from data. I was an early adopter of SSDs after a catastrophic loss of a RAID 0 array based on conventional drives. I have had a couple of SSD early failures (not recent) but that has not put me off using them. I would always recommend a robust backup routine but I wouldn't overthink your use of SSDs - the performance improvement over conventional drives is definitely worth having. Hope that's some help but ultimately it's just my perspective - others may have a different view. Andy

    Sonar Platinum, Win10 Pro 64-bit, 16Gb RAM, Six-core AMD, Twin SSD. Instruments: Roland A88 and Taylor 314CE all through a Focusrite Scarlett 2i4. More ambition than talent.

    Max Output Level: -49.5 dBFS
    • Total Posts : 2567
    • Joined: 2008/10/11 22:06:32
    • Location: West Midlands, UK
    • Status: offline
    Re: SSD's in a Sonar DAW: One vs. Two, and Partitioning Impacts 2017/08/15 03:14:22 (permalink)
    I went down the SSD road after losing two RAID0 arrays as well, in my case one Barracuda went bad after six months use and its warranty replacement lasted five months. Then I gave up and bought a couple of Intel SSDs (in sig), one for the C drive and one to spool audio to.

    The audio drive's been pretty heavily used over the last three, nearly four, years and is still going strong. I've recorded to it, taken stuff off it to free up space sometimes, put stuff I wanted to re-work back on it from the external drives I use for mass storage and it still reports everything is fine.

    Modern generation drives should be even better. MLC SSDs better again, from a wear point of view at least, and if you use one that connects directly to the PCIe bus the performance gets even better. SSDs, even earlier ones, don't seem to suffer from the high failure/wear-out rate that was feared a few years ago.

    Most laptops, all Macbooks and many Macs have one as their only internal drive and if rapid failure and warranty repairs were an issue the computer manufacturers wouldn't be using them in the millions.

    What, if any, advantage they offer over spinning platters in legacy IDE mode however I don't know. It might be worth checking what happens to the data bandwidth in that mode - if an HDD can pretty much use all of it there's little point in going down the SSD route other than for faster access times and less (zero) acoustic noise.

    Another thing to check is whether it's possible to send TRIM commands to SSDs in legacy IDE. A drive that doesn't get sent TRIM will still work faster than an HDD, but nothing like as fast as a TRIMed one, and will show slow-down tendencies over time. Though deleting all its contents, re-formating and giving the drive's built in garbage collection every chance to do its thing might help. Garbage collection in the absence of TRIM is one area a less than full drive might perform better. I'm not sure about that though, I've seen conflicting reports. You can't send TRIM over USB either, in case that might be an issue.

    Finally, I'm sure you know this but I'll say it anyway - never, ever defrag an SSD. More modern versions of Windows know what an SSD is and won't defrag one automatically or even let you defrag one. Older Windows that pre-dates SSDs may not behave so nicely.

    Sonar Platinum 64bit, Windows 8.1 Pro 64bit, I7 3770K Ivybridge, 16GB Ram, Gigabyte Z77-D3H m/board,
    ATI 7750 graphics+ 1GB RAM, 2xIntel 520 series 220GB SSDs, 1 TB Samsung F3 + 1 TB WD HDDs, Seasonic fanless 460W psu, RME Fireface UFX, Focusrite Octopre.
    Assorted real synths, guitars, mandolins, diatonic accordions, percussion, fx and other stuff.
    Max Output Level: -75 dBFS
    • Total Posts : 775
    • Joined: 2004/02/26 16:44:40
    • Status: offline
    Re: SSD's in a Sonar DAW: One vs. Two, and Partitioning Impacts 2017/08/15 04:34:10 (permalink)
    Besides reduced noise & seek time, I see the main advantages of an SSD in IDE mode being increased reliability & longevity when compared to the junk hard drives being made now days.  The more I research, and the more people i talk to, the scarier modern production HDD's are sounding for purposes other than mass storage backup.
    Here is what I've gathered so far about SSD's:  Trim and Wear-leveling are two separate things.  Trim mostly impacts performance, but primarily write performance on non-trim-aware-OS's.  And Wear-leveling is more about longevity, but now days the drives handle it well internally - with the one important caveat that more free space is better for the drive to spread out the write demands. my case I'm only regularly writing in two scenarios:  One would be the OS pagefile, the other would be when tracking.  I only record one track at a time, usually mono.  Even if we call it one stereo track I would be astonished if a drive that has not been trimmed in eons would suffer anything perceptible from the demands of stream-writing a single stereo audio track.  Even when residing with an OS pagefile.  Am I wrong about this???
    Jump to:
    © 2022 APG vNext Commercial Version 5.1