My only point was that it is all going to the same physical drive, and whether or not it is in partition A, B, or C, access times are in realistic terms comparable.
 
His question was not about having on different physical drives, just different partitions on the same physical drive (unless I misread his post badly).
 
Bob Bone
 

I think what we both wrote is entire separate and valid points, just needed to make it clear. You wrote that it makes no difference where you install apps, I wrote about it makes a difference what software you actually install on your DAW OS.

Look, partitioning does not create a new drive. It is possible that by requiring the read heads to jump to a new partition to access a different logical drive , there may be a longer travel time  (which is the biggest source of delay) to get the heads to a new partition/logical drive than if the different point of access are on the same partition. So you may actually decrease performance by partitioning. It depends on what is on what partition and where and  how often it is accessed.
 
The main use of partitioning in my view is to organize your files, and the main utility to that is to create a system partition that can be imaged as a file of a reasonable size. If you have to store a 2 TB file on a backup drive in order to be able to restore from an image file, you are going to need a very large backup drive. I would make a small enough partition on your boot drive to hold your operating system and your important programs. Install the important programs to that partition. Image that partition so that you can restore a working system to a new drive in a few minutes if you have a catastrophic failure. Use any space left over on that drive, as one or more extra partitions to install programs of less importance (you can always re-install from distribution media later if you need them), or to backup data from another drive so you have a copy if that other drive fails.
 
Backup your actual data using a backup program (not a drive/partition image) to an external drive ideally in native format, that is a format that you do not need the original backup program to read.
 
For performance, there is no real argument that splitting the 1) programs/OS, 2) audio data/projects, and 3) samples onto three separate drives is the optimum. But most users can get more than adequate performance from a single drive. In that scenario, I would still partition that single drive into an essential programs +currently updated OS partition, and an everything else partition for the reason that you can easily create an image of the working system to restore from in  an emergency.
 
 

Regarding SSDs, I am also curious if there are performance benefits to having 2 separate SSDs versus one, for example...
 
Currently, I have a dedicated SATA HD each for my C drive (OS + Sonar and other apps) and D drive (all audio data), which of course is the best practice for SATA drives. But I'm thinking of upgrading to an SSD or two - would there be any advantage either way to having one larger SSD drive, divided into two partitions for OS, apps + Data, or to get two SSDs  and keep things on separate physical SSD drives? Basically, I am wondering if the two separate, physical drives still offers a performance advantage with SSDs.
 
Thanks!

I don't know this for certain but I believe if there were any performance benefit, it would be from having the two SSD drives on two physically separate controllers.  This way you could have the full data bandwidth of individual controllers.  Two partitions would not gain you benefit aside from the logical organization of your data, I'd think.  With mechanical disks, data at the outside edge sectors is more rapidly accessed than that in the inside of the drive because a single revolution of the disk places more data under the head out there.  With an SSD, there's no benefit there.
 
With mechanical drives, the performance loss of having OS and data on the same drive (partitioned or not) is because of the physical movement of the heads and that they can't be in two locations at the same time reading different stuff.  There is an actual physics thing involved there.  Since SSD drives work on a random access method and there are no moving parts (except at the microscopic level, I suppose), there's less contention than there would be with the mechanical counterparts. 
 
I'd be interested to know how SSD drives work in that case.  I know for my laptop and desktop office machines, the SSD benefit is dramatic compared to the same systems with mechanical 7200RPM drives.  The laptop I'm using right now has a 100gb SSD - it's a 2007 model with the first gen laptop core2duo chip in it.  It runs Windows 8.1 (64), Office 2013, etc and does it basically like a new machine.  You'd never guess it to be 6+ years old.

Mouser
Regarding SSDs, I am also curious if there are performance benefits to having 2 separate SSDs versus one, for example...
 
Currently, I have a dedicated SATA HD each for my C drive (OS + Sonar and other apps) and D drive (all audio data), which of course is the best practice for SATA drives. But I'm thinking of upgrading to an SSD or two - would there be any advantage either way to having one larger SSD drive, divided into two partitions for OS, apps + Data, or to get two SSDs  and keep things on separate physical SSD drives? Basically, I am wondering if the two separate, physical drives still offers a performance advantage with SSDs.

Blades is right here. The only advantage you will see is with two physical SSDs on two separate SATA III connections (I stress the SATA III (6Gb/s) because an SSD on a SATA II would choke its operational abilities dramatically).
 
To be clear, the reason for an SSD as the OS/Program drive is because when a program is loaded, it loads much more than its exe file... there are associated dll files, ini files, etc. that all must load to make the program start. An example is Adobe Acrobat... just to start the splash screen shows it loading a ton of dlls. I get a chuckle to this day that I can open the X3 reference guide in less than 1/2 second (2094 pages), and that is loading Acrobat and the 35Mb file. Each time I do this, I will immediately close the file (leave Acrobat open) and double click the file again, and the second time is even faster!
 
That said, a data drive does not often meet these "needs," as they typically are accessing files one or two at a time. A SONAR project would be an exception, where the project file points to all the plugs and audio files and SONAR then "assembles" them. Yes, you will see an improvement in this with an SSD, but it is not the "common" use of the data drive. In most applications, this is overkill IMO, but some choose this route, and that is choice. When I talk data storage, I mean MASSIVE data storage (like 3 Tb worth), so to get that with an SSD would be expensive, and actually impossible (since most machines do not have enough SATA III connections anyway). My SSD is older, and have seen it do things that I do not "trust" with data libraries, and for this function, a 7200 rpm HDD is ideal.
 
For these reasons, it is more common to see only one SSD in a machine running the OS/Programs, and the remainder as 7200 rpm HDD. Some sample libraries are big enough to fill (no joke) an entire SSD on their own.
 
As far as partitioning drives... the only reason I have done this is to meet OS restrictions. Otherwise they do not buy you much other than organization. Partitioning an SSD would actually be a waste of precious space (the partition takes up a small amount by itself), and would potentially bite you as you fill it up. An SSD is essentially a massive chunk of RAM... no one thinks about partitioning their RAM... just because you can do something, doesn't mean you should.