View Single Post
  #19  
Old 21.07.2014, 02:52 PM
MBTC MBTC is offline
This forum member lives here
This forum member lives here
 
Join Date: 16.04.2010
Posts: 1,082
Default

Quote:
Originally Posted by grs View Post
On the cpu thing, I just update when benchmarks show around double performance gain. Like here http://cpu.userbenchmark.com/Compare...770K/617vs1537.
Check the server oriented benchmarks I think they are called Geek Bench.
Their take on the i7-950 "The i7-950 is a first generation Core processor which is now three generations and four years old. The 950 was included in the group test as a reference for how fast processors were back in 2009. "
Synthetic benchmarks like Geekbench or Passmark are good at providing a relative measure of certain types of CPU performance, but usually in real-world application scenarios, one typically does not see anywhere close to the same level of performance difference, because to get similar results an application would have to be written in such a way that it's doing the same thing as the synthetic benchmark (and of course most DAWs and VST execute entirely different types of instructions). Synthetic benchmarks are mostly a best-case scenario and designed to highlight even miniscule differences between processor performance, so with a roughly 50% performance increase (synthetic) in CPU performance, I'm trying to wrap my head around how that could translate to a 300-500% increase in performance when using Spire (keep in mind I'm not doubting you here, just wondering if there was something else going on with your previous configuration that could account for the difference you're seeing). I'm wondering if the Haswell architecture has some sort of multi-threading features that newer plug-ins might be able to take advantage of?

There is a benchmark tool called DAW Bench at http://www.dawbench.com/ that is designed to provide a real world measure of performance, but I have no direct experience with it. The Massive test I posted is far from perfect but it's just a quick way to get some sort of plug-in specific performance number. About Massive itself - yeah the presets aren't great but the synth itself is pretty good (and some third party sound sets are amazing).
Reply With Quote