Nuendo / SX Benchmarks :.

Blofelds DSP 40 - Page 2 : Audio Hardware Comparison :.

It became evident very early on during the development stages that there was enormous variance in results at lower latencies using numerous audio cards that I and others had at our disposal , in that respect , the decision was originally made to restrict the official qualifying to 1 -2 high end professional audio cards.

The 2 cards that chosen were the RME HDSP Series and the Lynx 2 Series, it was expected that the variances between the 2 top end products would be negligible at best, and would be a good reference point as a lot of end users could qualify the results easily due to the saturation of RME products within the Nuendo community.

Well that was the plan..


Unfortunately the variances were far more than expected, even to the point that on the Intel DualCore systems, the Session was unable to be run with any consistency at any settings below 512 Samples on the RME cards. What initially started out as a benchmark to test outright system performance quickly became an exercise in testing ASIO driver efficiency, or lack there of.

The ongoing testing over the following months produced a wealth of information regarding the variances that were also being experienced with not only the ASIO driver efficiency of the different audio hardware, but also inter- relationships with chipsets, memory bandwidth as well as the respective driver / buffer implementation being employed.

  One of the most surprising and widest variances reported was when simply saving and re-opening the session at the loaded setting would result in some cases the need to dis-able a large number of plugins for the session to successfully be able to playback cleanly again.

This was more pronounced at Buffer settings below 512 Samples, and is clearly evident in the following graphs.

The other thing to take note is the dramatic and over proportional improvement with certain audio cards at the overclocked settings, more so on the Intel Systems.

AMD X2 - Audio Hardware Comparison    

The selection of Audio devices include both PCI and Firewire units, the variance in the results at the lower latency settings give a reasonably clear indication of the performance deltas that can be experienced on this particular test with the respective units.


Its evident that the PCI cards collectively perform a little better than the Firewire units on the initial test, despite the Firewire devices having an added sample buffer to compensate for the inherent latency of the Firewire protocol.


The most pronounced variance displayed is in saving and re-opening the session. In this case the results are more reflective of the subtleties of the respective ASIO driver / buffer implementation being employed.


AMD X2 OC - Audio Hardware Comparison    
The results for the Overclocked settings on the AMD systems scaled proportionally for most of the audio cards listed, the notible exception being the RME Fireface, which performs substantially better.   This is hinting at other variables at the subsystem level which can effect overall performance of this test.

The pronounced difference in performance when running 1T or 2T memory timings on the AMD systems
  with certain audio cards is well documented, where as it doesn't effect other cards. Again that is reflective of the subtleties of the respective ASIO driver / buffer implementation being employed.
Page 1 | 2 | 3 | 4 |5 |6 |7|8 |9 |10 |11 |

© AAVIMT 2006