Alegria



My thoughts exactly and the reason why I did not agree with Scott to begin with. Also, the recent research I've done on this subject simply confirms it without a doubt. HPET OFF is not a tweak nor is it a "one size fits-all" fix for systems having trouble with a high precision external clock. And as mentioned by Goddard...,
"

 
[size=4 font="times new roman"]What research? Please show me anywhere with data that shows the HPET increases performance with anything. I have seen the opposite every time with Pro Audio. Boards that were unusable for audio below a 256 Buffer could now handle a 64 Buffer with the same interfaces. Laptops by default do not have the option to disable HPET in the bios so that is something we still have to convince the manufacturer to do when needed. What I can say is when I have shown the systemboard engineers the DPC screen pics and the links to the programs we use to get those, the first bios I get back is the option to turn off the HPET. I wonder why that is?
 
BTW please dont look at White Papers from 2006 to 2008 and think the evolution of that technology ends there. How do you think they are implementing live bios changes to EFI bios's without using the IME/AMT. They are integral to those as well as updating Firmware level power management that comes with laptops on the keyboard firmware as well as system bios. The Intel Management Engine was actually introduced first on laptops in the Core2 days. Intel and the laptop manufacturers stated the IME driver used was modified and specific to the laptop manufacturers as they required. Intel creates guidelines with the IME and then allows the manufacturers to change as required. That is definitely how it's done currently as EFI bios features expand. Dont assume 1 white paper from years before the platform tells you exactly how that technology is implemented now.
 
Eric
ADK

ok now that we got all this technical crap posted and out of the way..

Fact: accross MULTIPLE motherboards from varied manufacturers, and with both socket 2011 and 1155
and every processor that can go into them..

HPET OFF = less DPC
HPET ON= more DPC, and in fact can be very detrimental with certain video cards.. also has much to do with PCI cards working particularly Lynx on P67

period end of discussion, on each and every combo.
unless someone here who has tested this accross mulitple platforms like we have and in the thousands of systems count..
time to put this to bed.

Core parking: please with this crap.. Sonar was and is the only software this ever effected (well ok Logic but that totally out of topic in this forum)
it was extremely user dependant and rarely did it ever fix anything for the majority of our clients or in our testing.
that was 2 yrs ago not now.. its long done and gone and over.

"jcschild"
ok now that we got all this technical crap posted and out of the way..

Fact: accross MULTIPLE motherboards from varied manufacturers, and with both socket 2011 and 1155
and every processor that can go into them..

HPET OFF = less DPC
HPET ON = more DPC, and in fact can be very detrimental with certain video cards.. also has much to do with PCI cards working particularly Lynx on P67 

period end of discussion, on each and every combo.

No, this is not the end of the discussion because you say so. And you certainly don't take into account all system configurations such as dedicated DAWS (audio only). HPET adds a minute amount of latency simply because it's an external clock. This is well documented and also applicable to LAPIC. But on the order of what magnitude? This is something that I can answer confidently. Can you?

such as dedicated DAWS (audio only)

Is there a definition for this?

What is a 'dedicated DAW (audio only)'?

HPET adds a minute amount of latency simply because it's an external clock. This is well documented and also applicable to LAPIC. But on the order of what magnitude?

 
 
There is a inherent increase in DPC activity just from the HPET. That latency fluctuates on average between 350us and 150us higher than it is when turned off. This also varies by board, platform, and desktop versus mobile because of further firmware implemented on mobile. Normally this activity is not enough to cause issues with audio accept at a buffer below 64 on most interfaces or 128 on some. The real problems are any spikes that increase the activity more then 350us and often those are hardware effected by the HPET. The boards that show these issues have spikes upwards of 21,000us often replicating every second to 5 seconds. Disabling the HPET reduces these either to sub 5000us or often eliminates them completely depending on the driver/bios. That is how much activity HPET is initiating with those devices while it's active. On those same boards, the DPC activity would also completely go away down to sub 20us if the C-States and Turbo was disabled. All of this is far more than the original DPC activity HPET initiated alone.
 
Eric
ADK

""And you certainly don't take into account all system configurations such as dedicated DAWS (audio only). ""

what the heck do you think i am talking about goober? for audio only. dpc is not much of an issue for video editing as its technically not real time processing. or any other uses.

DPC only matters for audio
the lower you set buffer for audio the closer you are to real time (system time)

again unless you have the database we do which i know you dont, subject closed.. feel free to debate with yourself no one trust me no one will be listening.

pretty sure i can guess as to who people will listen to.

Ah, I just found the definition of a dedicated DAW for professional Audio use.

They seem to have a wide range, I wonder if they come with HPET on or off.  I also wonder if they have core parking enabled.

http://adkproaudio.com/

ECBowen

HPET adds a minute amount of latency simply because it's an external clock. This is well documented and also applicable to LAPIC. But on the order of what magnitude?

 
 
There is a inherent increase in DPC activity just from the HPET. That latency fluctuates on average between 350us and 150us higher than it is when turned off. This also varies by board, platform, and desktop versus mobile because of further firmware implemented on mobile. Normally this activity is not enough to cause issues with audio accept at a buffer below 64 on most interfaces or 128 on some. The real problems are any spikes that increase the activity more then 350us and often those are hardware effected by the HPET. The boards that show these issues have spikes upwards of 21,000us often replicating every second to 5 seconds. Disabling the HPET reduces these either to sub 5000us or often eliminates them completely depending on the driver/bios. That is how much activity HPET is initiating with those devices while it's active. On those same boards, the DPC activity would also completely go away down to sub 20us if the C-States and Turbo was disabled. All of this is far more than the original DPC activity HPET initiated alone.
 
Eric
ADK

In plain terms it enables me to use a 48 sample buffer size on my interface whereas with HPET enabled I can't without that regular 5 second 'click'.

I normally have my buffer size set to 64 anyway but there are times I enjoy the slightly beter RTL.

Funnily enough HPET on or off doesn't affect my baseline background DPC latency at all it's when things start rolling that it shows up, which is why I'm always suspicious of those claiming they've got their baseline figure down to ridiculously low levels when nothing is happening.  I want to run my audio software without glitches not have a machine that's optimized to run DPC latency checker like some here enjoy doing.

Thanks for the insight on IME as well btw.  I've often wondered if there is any benefit from installing the IME software package for my board but it seems to install all manner of components so currently I just install the driver from that package to silence the 'unknown device' in device manage rather than install all the other stuff.  So is there any benefit to installing any of the other components or is just the HECI driver adequate?

"ECBowen"
There is a inherent increase in DPC activity just from the HPET. That latency fluctuates on average between 350us and 150us higher than it is when turned off.

In my case, it's no more than approx. 15us. Quite the discrepancy. 

"Jonbouy"
I want to run my audio software without glitches not have a machine that's optimized to run DPC latency checker like some here enjoy doing.

I run my audio software without glitches and HPET on. I actually run demanding libraries (24 bit) in real time with HPET on. And the ensued smoothness is..., well it needs to be experienced. But as I've mentioned before, this is not something you can do..., simply because of the nature of your multi-purpose machine and the insane amount of overhead created by the greenness of it. To each his own.

The "DPC latency checker" btw. is not a benchmarking tool but rather a diagnostic tool that display spikes in real time. Along with the "Latency Monitor" which goes many steps further in identifying processes that may be causing these spikes, they are tools to help you in this regard, no more no less.

Good day gentlemen. 

We normally just use the standard IME install for the specific model. We only change it to driver only if we run into a problem. Lately I have not seen issues with the installed IME. However the manufacturers dont include information on what is implemented with the install so you never know. The safest method is just manually update and select the driver.
 
Eric
ADK