Bruce Ennis: SHUTTLE SN25P nVIDIA nFORCE4 CHIPSET
AMD ATHLON 64 4000+

Bruce I've read on the Sound on Sound forum that

There is currently a problem with all PCIe motherboards used for DAWs, though nForce4 may have it worst.

I take it your not haveing any problems ?

Andy

Just read this thread and wanted to post another centrino result - toshiba m30 1.6Ghz 1MB Cache, Echo Indigo I/O, Windows XP Home, 512MB RAM, Sonar PE 4.0.2

at 1.6Ghz

46: 29
23.2: 31
11.6: 34
5.8: 38
2.9: 46
1.5: 65 +/- 5

OC to 1.9Ghz

46: 24
23.2: 26
11.6: 29
5.8: 33
2.9: 39
1.5: 53 +/- 3

Date: 07/17/2005

46.4 23%
23.2 26%
11.6 30%
5.8 37%
2.9 N/A

Figures are average readings from meter variation of +/- 2%


Dell Dimension XPS
P4 3.6 with HT "On"
2 Gig RAM
two Raid-0 74 GB 10,000rpm drives
M Audio Firewire 410 & newest drivers vs. 5.10.0.5034
XP Home SP1

Sonar 4.02
24 bit file depth
Sampling Rate: 44100
Buffers in playback queue: 2

Are we supposed to do this test with the ASIO drivers? I get better results with the WDM drivers.

Dale,

Try ASIO and try changing the buffer size in the ASIO Panel. I think the Firewire410 goes lower than 5.8ms.

-S

ORIGINAL: Scott Reams

Dale,

Try ASIO and try changing the buffer size in the ASIO Panel. I think the Firewire410 goes lower than 5.8ms.

-S

Scott,

What the best I/O buffer setting? I have 2 gig RAM, P4 3.6 ghz Dell. Right now it's set at 512kb.

Also does checking "Enable Read Caching" and "Enable Write Cahing" help latency?

Thanks,

Dale

Thanks, Junski for the expanded information. I will play around with it.

Dale

ORIGINAL: Scott Reams

Okay folks...

There's a new entry on the chart that many should find interesting. I ran some quick tests on a dual-CPU dual-core Opteron 275 system (four cores). The test wasn't 100% ideal, because I was running 32bit Sonar4 in 64bit Windows (which hurts performance a little). Sonar likes all four cores. This system is the new leader on the chart by quite a bit, despite being 400MHz slower than the Opteron 252s which previously led the pack:

http://forum.cakewalk.com/tm.asp?m=34827&mpage=1&key=

-Scott

Scott

did you notice that 4.0.3 changes the result? My testing shows that it averages slightly higher CPU usage, but does not jump up out of bounds as often, so that I can now run at 1.5 when I could not before.

By the way, this sounds like you have a great machine with the 275 system.

yes

www.adkproaudio.com/benchmarks.cfm

Here are some specs for a new laptop running Sonar 4.03 PE.

Sager 5560

Pentium M 2mhz w 2g L2 cache
1 gig DDR2 ram (2x 512 meg)
Intel 915 chipset
120 gig 7200 rpm drive
XP Home
MOTU 896 HD

AC adapter used

Running ASIO (seems to work better on this system)

Samples latency (ms) cpu %

2048 46.4 27%
1024 23.2 29%
512 11.6 32%
256 5.8 39%
128 2.9 48%
96 2.2 58%

I initially had trouble with this test. Only at 256 samples was it stable enough to get a solid reading. CPU meter spiked constantly. But after tweaking and ESPECIALLY after setting the power scheme to ALWAYS ON did it settle down. I was able to get good solid readings.

My mouth is just now able to shut. This machine is faster than the NEO2 platinum with an A64 3500 that I had before. I suspect the DDR2 ram has something to do with it?

I like it!

ORIGINAL: kab

Unintentional bump... are there any plans for this same sort of shootout for Sonar 5 running under 64 bit mode?

I second that motion...................

My mouth is just now able to shut. This machine is faster than the NEO2 platinum with an A64 3500 that I had before. I suspect the DDR2 ram has something to do with it?

Actually, DDR2 , with its higher latencies, has very little to do with it. PentiumM is simply a very efficient CPU.

I do plan on rewriting Sonartest for Sonar5 soon and posting for a new set of results. Stay tuned for that...

-S

ORIGINAL: Scott Reams

My mouth is just now able to shut. This machine is faster than the NEO2 platinum with an A64 3500 that I had before. I suspect the DDR2 ram has something to do with it?

Actually, DDR2 , with its higher latencies, has very little to do with it. PentiumM is simply a very efficient CPU.

I do plan on rewriting Sonartest for Sonar5 soon and posting for a new set of results. Stay tuned for that...

-S

Thank you, Sir.

Here are my results with my system:

Pentium D 820, (dual@2.8ghz)
1Gb RAM
Winders Server 2003, sp1
E-MU 0404 ASIO drivers
(use Multiprocessor Engine checked)
Sonar 4.04

2ms = 70%
4ms = 42%
5ms = 37%
7ms = 33%
10ms = 27%
20ms = 23%
40ms = 20%


---shane

On my yet-to-be-optimized/freshly installed machine, with Internet Explorer opened and a bunch of "useless" process running in the background, I get the following:

1.5 79%
2.9 54%
5.8 46%
11.6 42%
23.2 36%
46.4 32%

5.8 ms being the comfort zone for me, I rarely run into a dropout or out of tracks...


AMD Athlon XP 2800+
A7N8X Board
nVidia GeForce 4MX
512 MB Ram
M Audio Delta 44
XP SP2
Sonar PE4