July's Sound on Sound magazine had a good DAW-based CPU benchmarking article. 
 
http://www.soundonsound.com/sos/jul15/articles/cpu-performance-0715.htm
 
If you're not a subscriber you can buy the PDF for $ 0.99

who won?
 

Well, it was fairly well thought out, so there was a lot of discussion about where you were starting from and what upgrade would give the best value for your money.  So if you have a CPU that's at least 3 years old, it looks like the Core i7 4960X @ 4.0 GHz will give the best bang for the buck.

http://cpubenchmark.net/high_end_cpus.html

That is a good site for benchmarks of all kinds. The current avg cost of components is down the right edge of each section. Bear in mind the benchmarks of components are an average of numerous systems they are contained in.

mettelus
http://cpubenchmark.net/high_end_cpus.html

That is a good site for benchmarks of all kinds. The current avg cost of components is down the right edge of each section. Bear in mind the benchmarks of components are an average of numerous systems they are contained in.

Thanks for that, mettelus  I think the value in the SOS benchmark is that it is measured using a DAW (Reaper).  Different CPU's benchmark differently depending on the task.  Ones that scream at gaming can suck at doing intensive calculation in large macro-laden Excel workbooks or MatLab, and vice versa.  Maybe that's evened out some lately but previously the differences could be surprisingly large.  I used to be a devotee of Tomshardware before they sold. I used to be amazed at some of his/their benchmark results on seemingly similar chips.

mettelus
http://cpubenchmark.net/high_end_cpus.html

That is a good site for benchmarks of all kinds. The current avg cost of components is down the right edge of each section. Bear in mind the benchmarks of components are an average of numerous systems they are contained in.

I like the benchmark/$ further down on the page.  The i7 5820K looks pretty good according to that assessment.  However again I would say that alot of what counts in their test doesn't matter to DAW performance:
 

• CPU tests Mathematical operations, compression, encryption, SSE, 3DNow! instructions and more
• 2D graphics tests Drawing lines, bitmaps, fonts, text, and GUI elements
• 3D graphics tests Simple to complex DirectX 3D graphics and animations
• Disk tests Reading, writing and seeking within disk files
• Memory tests Allocating and accessing memory speed and efficiency

This is from the explanation on their webpage and is a good and typical benchmark.  But the highlighted sections aren't really pertinent to performance using a DAW and so will skew the benchmark results relative to performance in DAW's.  That's why I think SOS's benchmark is so valuable to us.  It's benchmarking specifically on DAW performance.
 

Yeah, the benchmark that they use is pretty comprehensive, but all test are not applicable to a DAW-only machine as you pointed out. Interestingly when I built this machine this article on the i7-990X vs. i7-2600K was a massive eye opener (one of their benchmark tests used SONAR X1). Intel shot themselves in the foot with putting these two CPUs on the market at the same time. Bottom line, the fixed-OC 6-core 990x was evaluated as "'extreme' only in terms of cost and power consumption" (it only performed better for encryption and video rendering basically). The variable-OC 2600K Sandy Bridge outperformed it in all else at 1/3 the cost); but granted, at that time not a lot of programs were written to use all available (6) cores properly either.
 
Passmark's performance software has a free 30-day trial (I install it, benchmark various BIOS/system settings, then uninstall it), and is an interesting test to run. It actually opens running the graphics card at max speed (which is nutty!) - I hear the GPU "squealing," then look in the bottom right and see "4200 frames per second" LOL - no one's eyes are that fast!!! The link in the upper left (full benchmark) is an interesting test to watch.
 
This system still benchmarks 15-20% higher than their site reports (I just ran it after a reformat/rebuild). It is getting long in the tooth for "raw speed" but as a DAW is more than adequate, and I think it could be built rather cheaply these days (even newer SSDs/GPUs are cheaper and faster than mine, and a DAW doesn't need "3D" anyway).