Laptop performance questions - I'm considering using an HP Athelon 64 3400+ 1 meg cache (model zv5255us)

How much slower will the laptop performance be than a similar athelon 64 3400 desktop? Have any results been submitted to help make a good guess?
The 512meg ram is slower - only PC2700 333 mhz, and I plan to use an external firewire drive and firewire M-audio 410. Can 1.5ms latency be achieved? It would be great to have the portability and the performance if it can be done...

ORIGINAL: rgfors

Okay, I've now posted three different results with three different computers. My first post was with a 3.2ghz P4, the next was with a 3.2ghz P4 Extreme Edition and my new computer is as follows:

2) Opteron 250s
(2) Zalman 7000A coolers
Tyan S2885 dual-Opteron motherboard
4GB OCZ CAS2 PC3200 registered DDR (8x512MB)
ATI Radeon 64MB DH AGP Video
HighPoint 8-way SATA RAID controller
Plextor 712 series DVDRW
120GB 7200rpm 8MB system drive
(2) 160GB 7200rpm 8MB SATA samples/loops drives
(3) 160GB 7200rpm 8MB SATA audio drives in RAID5 array
(3) SATA drive sleds for audio drives
Antec 550W quiet power supply
4-space rack case (26” deep)
WinXP Pro

You left out one of the most important items, the sound card you're using.

Hi All, and Scott...

System:
Laptop HP Pavilion zt3340EA
Centrino 1.6 ghz
512 MB DDR
1 60 GB HD (same hd for win and audio)
M-Audio Firewire 1814
Mobility Radeon 9600

results (WDM) :

46.4 ms 65-66%

34.8 ms 68-72%

23.2 ms 52-79%

11.6 ms 56-60%

5.8 ms 69-71%

2.9 ms 68-79%

1.5 ms DROPOUT

results (Asio) :

46.4 ms 63%

34.8 ms 64-65%

23.2 ms 67%

11.6 ms 56-57%

5.8 ms 55-70%

2.9 ms DROPOUT

1.5 ms DROPOUT

Djoni

tomorrow will send the results of SonarTest on my new Atlhon64 daw.

hi Scott ,

watching your graph I can't understand which is the fastest between Athlon FX 51 and Athlon 64 3400 , they have the same color and the same picture

thanks

Not sure if Im doing mine right, but its hovering at about 50%. Im using ASIO drivers. and my soundcard is the Maudio 2496

Im running a P4 2.5Ghz
1 gig ram
win2k
80gig HD
Sonar 3 PE

sampling rate in S3 is 44100 at 16bit. My latency says 8.7msec in sonar

If I click on the ASIO tab..i get to my Maudio software, and then gives me the option to change DMA buffer size (set to 384 right now)
lets me pick from:
64
128
256
384
512
768 etc...etc..

anyone have any ideas they can help me with for proper settings?

thank you
trybal

what do i want my dma buffer set yo? the lowest number possible without drop-outs and/or pops/clicks?

getting very confused

After giving up on the laptop idea (adding memory was going to be too expensive), I went ahead and set up a desktop with a 19" lcd for a lot less:

Athelon 64 3400+ 1 meg 2.2ghz
ASUS K8V-SE Deluxe (MSI K8N Neo Platinum was sold out)
Seagate 160 UATA boot/system drive
Seagate 160gig SATA audio date drive
MadDog 64 meg 4x agp video card
Memory: 1 gig (2 x 512 mB) OCZ Value 3200 (after returning OCZ platinum 3200)

This test also compares OCZ Value memory with speeds of 3-4-4-8 against OCZ platinum with speeds of 2-3-2-5

The conclusion first: The results were almost identicle, except that with the value memory I noticed the CPU meter would occasionally jump to a higher level, though it wouldn't stay there. There's a subjective element to this test as to whether to report these high spikes or not. Since there wasn't a significant difference I went with the value memory and saved about $150 for better or worse.

Latency __ OCZ value _ OCZ platinum
1.5 _____ 66-72 _____ 66-71 (both types had high spikes of 79 and 78)
2.9 _____ 43-55 _____ 44-53
5.8 _____ 37-40 _____ 36-40
11.6 _____ 32-33 _____ 31-33
23.2 _____ 28-29 _____ 28-31
46.4 _____ 26-28 _____ 26-27
92.9 _____ 25-26 _____ 25-26
349 _____ 25-26 _____ 25-26

I added the high latency test of 349 to show that there appears to be a base level of cpu activity regardless of latency settings.

The spikes were present in the very lowest latency settings, above 5.8mS I didn't see any

I was just thinking that it would be nice to compare S3 vs. S4 with the SonarTest to see how the new audio engine does.

ALL YOU S3 AND S4 USERS POST NUMVERS FROM BOTH VERSIONS.

Ahem, I would post my 1.33GHz Thunderbird results once I redo my system. But I don't think anyone really cares about my relic

Scott , don't you have the chance to update this test with the new processors ( Athlon 64 3500 , 3800 , Pentium 4 3,4 ...) ?

That's a great idea. Scott, any chance you'd consider setting up a SONAR 4 graph?

As someone on the fence about the S4 upgrade, I'd be very interested in seeing this comparison.

Jon

I too am looking at an Athlon64 platform after the new year and was curious myself about 754 vs.939. I will be upgrading after the new year when prices drop after the holidays.

There was mention of getting the Clawhammer for its larger 1MB L2 cache and NOT the Newcastle with a 512KB cache.

According to an article on Tom's Hardware Guide http://www.tomshardware.com/cpu/20040601/index.html, socket 939 offers on chip memory management AND the HyperTransport buss speed goes to 1Ghz as opposed to socket 754 with a HyperTransport buss speed of 800MHz. This is supposed to make up for the smaller 512KB cache of the 3000, 3200, 3500 and 3800. With the addition of the 4000 and FX-55 the cache seems to go back to 1MB http://www.tomshardware.com/cpu/20041019/index.html. The FX-53 is also getting phased out. I'm surethe new 4000 and FX-55 will drive the price of other chips down. The question is about price performance:

Is socket 939 indeed better than socket 754?
With the discontinuance of the FX-53 does that appear to be a bargain?

What say you DAW builders.

Is socket 939 indeed better than socket 754?

In general benchmarking... yes. In Sonar... no, unless you have 1MB of L2.

-S

System specs below!!

46.4 ms 27%
23.2 ms 29%
11.6 ms 33%
5.8 ms 46%
2.9 ms 56%
1.5 ms 0%

dave,
if you know anything about home studio i could use some help seeing as you are in th T.O. area.
check out my posts under the member name "ejmac".
thnx,E.