This is a popular link over at KvR to test your audio (native) processing power:
http://perso.wanadoo.fr/vb-audio/us/mm/ ... st_spp.htm
Here's the KvR link where people are posting results:
http://www.kvr-vst.com/ubb/ultimatebb.p ... 1;t=001766
Audio performance test
I´ve had some discussions whith Vincent Burell... He is one of the real hard-hitters when it comes to quality-plugs for the Merging platform, which is one of the very best platforms, I have ever seen... I will try to explain what those numbers mean. Anyway those are just approximations which have a little error rate... One of those factors which is quite mportant is the chipset itself... When testing an Athlon system you need the right chiset (761, 760). When you do use multi-track sequencers like Nuendo, Cubase, ..... there always has to be the right balance between ram access, cpu-power, and pci bandwidth. The differences do show up, whith things like track count and restricted processing power.
Regarding those figures, I think the most important thing is that the memory bus does not stand in the way of processing power, since...
-SPP-M represents the transfer speed of you memory and
-Spp-2 your processing power...
Vincent Burell:
-"i've Created the SPP-M unit to check that the processing is not bound by the speed of the memory BUS."
-"Note that the Important unit is the SPP2 (cause most of audio processing software now use float processing only) ..."
-"MFloat/s is deduced from the SPPM and is the number of Mega Float (1024 *1024*4 Byte) moved per second when computing the SPPM measure."
-"SPP2 > SPPM means that the memory BUS is not fast enough for your processor."
Those are his statements... But nevertheless there is a certain error rate as I mentioned earlier... Here are some results that were gotten whith a new Tualatin-S:
SPP-1 mono 44100Hz : 1942 SPP/smp
SPP-2 mono 44100Hz : 1988 SPP/smp
SPP-M mono 44100Hz : 3179 tracks
Transfer rate : 267 Mf/s
Although those numbers seem to score higher than "Athlon" and "P4", real performance tests show a different result!
Regards,
Sunshine
<font size=-1>[ This Message was edited by: Sunshine on 2002-04-03 01:53 ]</font>
Regarding those figures, I think the most important thing is that the memory bus does not stand in the way of processing power, since...
-SPP-M represents the transfer speed of you memory and
-Spp-2 your processing power...
Vincent Burell:
-"i've Created the SPP-M unit to check that the processing is not bound by the speed of the memory BUS."
-"Note that the Important unit is the SPP2 (cause most of audio processing software now use float processing only) ..."
-"MFloat/s is deduced from the SPPM and is the number of Mega Float (1024 *1024*4 Byte) moved per second when computing the SPPM measure."
-"SPP2 > SPPM means that the memory BUS is not fast enough for your processor."
Those are his statements... But nevertheless there is a certain error rate as I mentioned earlier... Here are some results that were gotten whith a new Tualatin-S:
SPP-1 mono 44100Hz : 1942 SPP/smp
SPP-2 mono 44100Hz : 1988 SPP/smp
SPP-M mono 44100Hz : 3179 tracks
Transfer rate : 267 Mf/s
Although those numbers seem to score higher than "Athlon" and "P4", real performance tests show a different result!
Regards,
Sunshine
<font size=-1>[ This Message was edited by: Sunshine on 2002-04-03 01:53 ]</font>
p4 1.6a @ 2133 ddr @ 167mhz x 2 DDR
SPP-1: 1119 / 1028
SPP-2: 1324 / 1216
SPP-M: 6798 / 6218 / 569 <<!!!!!!!
first 2 numbers are similar to what the high end athlons get, look at the track count though
those first 2 tests really show where the tualatins 2 FPU beat the p4 single one.
<font size=-1>[ This Message was edited by: defex on 2002-04-07 21:47 ]</font>
SPP-1: 1119 / 1028
SPP-2: 1324 / 1216
SPP-M: 6798 / 6218 / 569 <<!!!!!!!
first 2 numbers are similar to what the high end athlons get, look at the track count though
those first 2 tests really show where the tualatins 2 FPU beat the p4 single one.
<font size=-1>[ This Message was edited by: defex on 2002-04-07 21:47 ]</font>