CakeAlexSHere

azslow3
General PC components are not DAW optimized. Gamers have different requirements.

Name one motherboard, memory module, graphics card, CPU, case or power supply esp optimized for DAWs in mind by design (hardware or software drivers).

Yes. And that was my point. If someone want DAW optimized PC, the only way is to ask people/company which is specialized in building such PCs. They check which components are good for DAWs and which are not, since that information is hard to find otherwise. Particular user has much less possibilities to "try" different components, general/gaming PC manufacturers do not check that aspect.
 

CakeAlexSHere
 I've been building computers professionally since the 90's myself and have worked for some major names. Respect the fact that people can have different opinions to yourself and are qualified to make them. There is no harm in disagreeing JUST STOP DOING THIS. Thx.

Jim has written about hyper-threading and TurboBoost on modern system after careful tests, the same is true for all his other statements in this forum.
 
My English is not good, so I will just quote another post:

melmyers
CakeAlex, you are a legend around here, but not in the way you wish. You are well remembered as the tool who attempted to hijack this forum a few years ago, pretending many of us were unworthy of having a problem addressed or even participating in a discussion without your blessing. You continue to post here, trying to appear to be the smartest person in the room, while in the end you come off completely the opposite.

 

^ The first point is a sensible answer.

^ The second point totally evades my point. Hence you don't fully quote, instead cherry pick to make me look bad. Ultimately you are saying you don't respect different viewpoints, try to twist and then clearly state you are effectively taking sides, when Jim has already clarifield. Unnecessary.

^ The third is just being a dick wanting to stir up trouble. I assume you will want to PM me and threaten to punch me in the face next. It's amazing you got away with that one, on the UK the police would have knocked on your door, but then that's just another reason why these forums have a reputation of being rude rather than having sensible discussion that may contain disagreements.

Well you brought up the past don't expect to get something back. It is clear you are simply posting to be aggressive and it's people like you who serve the forums reputation.

This will probably get deleted or something, but it's the truth.
Have a good day.

Here's something fun to try.  I call it the 'Diva Stress Test'. 
 
You can use the free Diva demo if you don't have it.
 
Load up a dozen tracks or so with Diva inserted with a dual VCO preset, set the accuracy to 'Divine' on each (top-quality zero delay feedback filters), normal poly mode with 8 voices or so each, and route your controller to all tracks so that all Diva instances play in unison.
 
Now play a chord.  If you get breakups right away, turn off MIDI echo on a few tracks and try again.  If you don't hear anything bad, check your CPU meter, then play chords with both hands.  If still good, keep adding tracks until your system maxes out.  That will give you an idea of what your system can handle.
 
I can run about 4 Diva instances with 4 voices at ASIO buffer size of 64.  I start getting dropouts with more voices before my CPU maxes, so increasing buffers lets me add voices without dropouts until I hit my CPU max.  Seems to be a combination of buffer size and CPU capacity determines the overall performance.  I normally run less demanding plugins, and run with a 128-256 buffer.  So what I have works for what I normally do.  This is analogous to bench-marking frame rates on a gaming GPU...all in good fun!!!
 
From the u-he manual:

But what sets DIVA apart from other emulations is the sheer authenticity of the analogue sound.
This comes at the cost of quite a high CPU-hit, but we think it was worth it: Diva is the first native
software synth that applies methods from industrial circuit simulators (e.g. PSpice) in realtime. The
behaviour of our zero-delay-feedback filters when pushed to the limit clearly demonstrates the
advantages of this groundbreaking approach

i'm gonna try that diva test, although not at 64! do you have repro yet?

abacab
Here's something fun to try.  I call it the 'Diva Stress Test'. 
 
You can use the free Diva demo if you don't have it.
 
Load up a dozen tracks or so with Diva inserted with a dual VCO preset, set the accuracy to 'Divine' on each (top-quality zero delay feedback filters), normal poly mode with 8 voices or so each, and route your controller to all tracks so that all Diva instances play in unison.
 
Now play a chord.  If you get breakups right away, turn off MIDI echo on a few tracks and try again.  If you don't hear anything bad, check your CPU meter, then play chords with both hands.  If still good, keep adding tracks until your system maxes out.  That will give you an idea of what your system can handle.
 
I can run about 4 Diva instances with 4 voices at ASIO buffer size of 64.  I start getting dropouts with more voices before my CPU maxes, so increasing buffers lets me add voices without dropouts until I hit my CPU max.  Seems to be a combination of buffer size and CPU capacity determines the overall performance.  I normally run less demanding plugins, and run with a 128-256 buffer.  So what I have works for what I normally do.  This is analogous to bench-marking frame rates on a gaming GPU...all in good fun!!!

 
That is an excellent (quick/easy) way to gauge low-latency performance (under substantial load).
Thanks for posting!

pwalpwal
do you have repro yet?

Not yet, and it really looks interesting since they added the Prophet 5 model!  But I already have the Pro-V synth in IK Syntronik (I know...SAMPLES... ), but I do love the sound of the Prophet!  Guess I need to check out the u-he demo of Repro.
 
I wish I could afford to buy all the u-he synths, since I think they some are the best sounding ones around these days.  I have been dabbling with the Diva demo, so that one would probably be next.  The u-he TyrellN6 V3 is an awesome VA for free!
 
KV331 Audio has updated SynthMaster 2 (since 2.8, I believe) with zero delay feedback filters, and SynthMaster One has had them since introduction, and they seem to be a bit more CPU friendly than the u-he stuff.  Those two have been keeping me away from spending more cash, thankfully! 

CakeAlexSHere

 Also please stop posting as though you have are the last word in PC hardware, your opinions are valuable and right up the top I'm sure.. but not all of us are altogether stupid. I've been building computers professionally since the 90's myself and have worked for some major names. Respect the fact that people can have different opinions to yourself and are qualified to make them. There is no harm in disagreeing JUST STOP DOING THIS. Thx.

well put.  as far as ht ...i leave it on.  I've tried with it off and on and i didn't notice any difference. 

Although apparently there is nothing to stop you from having a pissing contest. Why would you? Frankly, I fail to see how this answers my question. I guess that must not be the point. And to the one who felt it necessary to argue my financial status. What? I'm disabled... I receive $1000 a month! 

@ the OP:
David, if you need help spec'ing out a DAW with "Silent PC", I'd be happy to steer you in the right direction.
 
 
Back in Nov 2002, the Pentium 4 (single core CPU) was released.
That was the first incarnation of Hyper-Threading.
Back then, Cubase (in particular) suffered from audio glitches if Hyper-Threading was enabled.
Fast forward 16 years... and it's a much different situation.
 
If you have a CPU with six physical cores (like the 8700k), open Sonar and look at the Performance Meter.
Notice that with Hyper-Threading enabled, there are twelve processing threads.
Now, disable Hyper-Threading... and have a look at Sonar's Performance Meter. 
Note that there are now six processing threads.
 
Additional CPU cores (especially virtual) don't yield a 1:1 increase in processing speed.
IOW, Enabling Hyper-Threading won't anywhere near double the machine's performance.
BUT... unless your CPU is over-heating, there is *zero* value in disabling Hyper-Threading.
The only thing you achieve by doing so is lower performance.
You may not be running projects where the performance-hit is a factor, but you *are* limiting performance.
It's not opinion, it's physics of a modern CPU.
 
 

∞