KVR Audio

jcschild · Post by **jcschild** » Tue May 31, 2011 3:39 pm

egbert wrote:That doesn't follow at all. Like most companies (eg Apple) who wish to sell their current products, Intel doesn't generally tip it's hand on all forthcoming models - hence all the leaked roadmaps from 3rd parties and the speculation. Of course they need to send engineering samples of CPUs and chipsets to 3rd parties or the products will not be ready for market in a timely fashion so they can't keep too tight a lid on what is coming. Another thing is that the various roadmaps and slides that are all over the hardware sites are from different dates so they don't necessarily reflect the current state of play.

I don't claim any certain knowledge about this - geez - Intel's marketing people may not have even decided on the what and when of forthcoming releases. A quick search turns up lots of discussion of what is likely to be released. You seem quite categorical about this - if you are wrong, what will you do? Drop your virtual pants and do three laps of KVR doing your impression of Vincent Donofrio on the parade ground in Full Metal Jacket? (j/k)
(PMing me an abject apology would do fine)

Let's wait and see.

sorry but you are wrong.

there will be NO new processors for the desktop/workstation market from Intel until Late Nov/Dec at the very earliest.
those are supposed to be server (Xeons E5 EP C600 chipset) (patsburg) and Desktop X79 chipset. (sandy Bridge E socket 2011)

now as we all know Xeons and desktop processors are the same exact processors with different boxes.

these are ready to go now so depending on how well AMD does (i dont expect much from AMD) Intel could jump the release date up.

on a side note we could see some Ivy bridge socket 1155 (nothing more than a die shrink) sooner.

Intel is keeping releases very close lately. normally i can get a lot of info from my engineer contacts within Intel way beyond road maps i get emailed all tha time. last 6 months has been very difficult.

as to if there will be Quad Memory support in a desktop (single processor board) remains to be seen, this is one of those things i cant get a straight answer about.
every answer i have recieved seem to be that only dual Xeons will have this.

in all the yrs i have been doing this i have never seen so much speculation as to what is really being released and when..

while 8 and 10 core processors are already out (June) from Intel they are for 4P servers (blade) E7 48xx series. and X75xx were released oct last yr (8 core) these are all on FCLGA8 socket.
again not for workstation.

i dont see any 8 core for desktop being released untel next yr.

Scott
ADK

PS: to give you a small idea of how intel works presnet Sandy bridge we have now were designed 5 yrs ago in Israel.

pandashake · Post by **pandashake** » Tue May 31, 2011 4:01 pm

Once you get that ASUS P8Z68-V PRO up and running, manx, post back here to let us know how it's going.

And you're right, this did get off topic.

egbert · Post by **egbert** » Tue May 31, 2011 4:04 pm

@ Scott

This is where I came in - in response to Ashe37:

Ashe37 wrote:

The next gen chips aren't going to be out until the end of the year.

I wrote:

What? Surely you jest. Bulldozer is weeks away and Intel will launch their 8 core CPUs well before the end of NEXT year.

I expect faster Sandy Bridge this year and then more next year.

I have read all year that faster Sandy Bridge chips were coming and I expected them to be 8 core since Gulftown is already 6 core and the current server chips are 8 core and 10 core. (I saw a demo of an Intel Westmere based graphics workstation doing real-time ray tracing on 40 cores/80 threads in 4 sockets earlier this year.) I had another look around and the hardware sites seem to be a little short on definite info but there is a lot of talk about 8 cores before Ivy Bridge.

I think what you are reporting about Intel playing this very close makes perfect sense. Bulldozer is potentially the first threat to complete dominance by Intel at the high end in quite some time so I would expect Intel to wait and see what their competitor does. Intel will pull the trigger on desktop versions of their Xeons - which as you point out is purely a marketing decision as they are the same product underneath the marketing hoopla - if they feel the need to compete favorably with Bulldozer.

Time was, the marketing bullsh@t used to be all about "MHz". To some extent that same spurious measure of CPU power could now be "# of cores" - AMD might get mindshare simply because it has an 8 core chip out there. If Intel feels the need to release 8 core desktops to assert itself why wouldn't it?

The other interesting possibility that you mention - bringing Ivy Bridge forward seems quite possible. I just read on the Inquirer that Ivy Bridge Laptop chips are due in 1H of 2012. If Intel has their new 3D trigate process on 22nm ready to roll there is every reason to ramp it up sooner because it will make chips with more transistors and higher clocks cheaper to make.

egbert · Post by **egbert** » Tue May 31, 2011 7:01 pm

Here's Tom's Hardware on early arrival of Ivy Bridge and projected SBEs - 6 core is how they are calling it for SBE. Exactly what models of IB will be out they ain't saying.

monkeymanx · Post by **monkeymanx** » Wed Jun 01, 2011 12:12 pm

pandashake wrote:Once you get that ASUS P8Z68-V PRO up and running, manx, post back here to let us know how it's going. And you're right, this did get off topic.

It will be a bit. I'm working on the road now which is one of the reasons I have money now for a new system. I can't wait to get it all set up and get a healthy stable over clock.

NTR · Post by **NTR** » Thu Jun 02, 2011 5:48 am

monkeymanx wrote:I'm pretty much set on the ASUS P8Z68-V PRO now. Seems like a good overall board and after reading reviews is pretty much the best board for overclocking out there at this time. Just have to wait for newegg to get them back in stock. mwave.com has them but I haven't used them before.

Not sure where you live, but I have used NCIX Canada in the past and found their prices overall better than newegg and fast shipping.

Please let us know how that mb works out for you, its the one I want as well, I'm just holding off till fall to start my new build. Till then, there is just not enough time in the day to bother.

leonski · Post by **leonski** » Thu Jun 02, 2011 11:35 am

monkeymanx wrote:I was curious if there was a certain board that was known to be good for DAW use that is also good for overclocking? Also one with low DPC latency? Boards range from anywhere from $120 to $350 pending features. I just want to get something that is tried and true for DAW use with the 2600k.

There's some information regarding 2600K motherboards and DPC latency at the bottom of this page - http://www.anandtech.com/Show/Index/433 ... 68v-review

egbert · Post by **egbert** » Thu Jun 02, 2011 11:57 am

It's great that this info is now being published. What you might find is that there are a couple of settings in the BIOS that could turn that whole graph upside down - some of those boards might get right down into double figures with appropriate settings. Nevertheless they should all do that out of the box and hardware sites publishing this data might help make that happen.

Ashe37 · Post by **Ashe37** » Thu Jun 02, 2011 12:15 pm

egbert wrote: It's great that this info is now being published. What you might find is that there are a couple of settings in the BIOS that could turn that whole graph upside down - some of those boards might get right down into double figures with appropriate settings. Nevertheless they should all do that out of the box and hardware sites publishing this data might help make that happen.

yes, thats neat but its DPC latency to the onboard realtek audio that we're going to turn off anyway

egbert · Post by **egbert** » Thu Jun 02, 2011 12:24 pm

Ashe37 wrote:
egbert wrote: It's great that this info is now being published. What you might find is that there are a couple of settings in the BIOS that could turn that whole graph upside down - some of those boards might get right down into double figures with appropriate settings. Nevertheless they should all do that out of the box and hardware sites publishing this data might help make that happen.
yes, thats neat but its DPC latency to the onboard realtek audio that we're going to turn off anyway

Nope - all those boards have Realtek audio and the DPC latencies measured are due to higher priority processes monopolizing the CPU:

Critical interrupts will be handled as soon as possible, whereas lesser priority requests, such as audio, will be further down the line. So if the audio device requires data, it will have to wait until the request is processed before the buffer is filled. If the device drivers of higher priority components in a system are poorly implemented, this can cause delays in request scheduling and process time, resulting in an empty audio buffer - this leads to characteristic audible pauses, pops and clicks. Having a bigger buffer and correctly implemented system drivers obviously helps in this regard. The DPC latency checker measures how much time is processing DPCs from driver invocation - the lower the value will result in better audio transfer at smaller buffer sizes.

Ashe37 · Post by **Ashe37** » Thu Jun 02, 2011 1:08 pm

egbert wrote:
Ashe37 wrote:
egbert wrote: It's great that this info is now being published. What you might find is that there are a couple of settings in the BIOS that could turn that whole graph upside down - some of those boards might get right down into double figures with appropriate settings. Nevertheless they should all do that out of the box and hardware sites publishing this data might help make that happen.
yes, thats neat but its DPC latency to the onboard realtek audio that we're going to turn off anyway
Nope - all those boards have Realtek audio and the DPC latencies measured are due to higher priority processes monopolizing the CPU:

Yes. Ae they using the current driver from Realtek, or are they using the drivers as supplied by Asus? in each case? (they don't supply audio driver revision information at all...) Are the drivers on every one of the boards listen the same revision? or are they different every time? has anything else changed since they tested the other boards?

jcschild · Post by **jcschild** » Thu Jun 02, 2011 1:17 pm

oh great now the bone heads at anantech are trying to show DPC latency on what motherboards without having a CLUE as to how to make DPC change...
more disinformation....

bad enough they were using Sonar and a crappy test.

Scott
ADK

egbert · Post by **egbert** » Thu Jun 02, 2011 1:20 pm

Audio devices experience the symptoms of DPC problems - they don't usually cause them. Realtek isn't the issue.

jcschild · Post by **jcschild** » Thu Jun 02, 2011 2:47 pm

actually the realtek could very well be the issue.
you can change the drive on the realtek and see a vast difference.
on one of our previous laptops we used a driver for lan that was 4-5 versions old.. anything other than that one cuased high DPC..

Scott
ADK

egbert · Post by **egbert** » Thu Jun 02, 2011 3:01 pm

jcschild wrote:actually the realtek could very well be the issue.
you can change the drive on the realtek and see a vast difference.
on one of our previous laptops we used a driver for lan that was 4-5 versions old.. anything other than that one cuased high DPC..

Scott
ADK

Jeez- what is eating you? How do you know that is the fault of the Audio driver? Possibly some interaction between that driver and more recent LAN drivers.

Sure, it is possible that a bad audio driver would cause problems but, given the fact that audio (unlike many other functions) is exactly the sort of function which manifests DPC problems (audible clicks and pops) you would expect that the audio drivers would tend to be sorted and it tends to be things like wifi drivers and the like that won't let go of the CPU in a timely fashion.

Anyway, this is irrelevant. The situation in that article is one where DPC latency of the whole box and dice is being measured - the DPC number is not a specific measure of the performance of the Realtek driver in each installation. If you wanted to make inferences you would need the DPC latency measured with the Realtek on board audio enabled in each case and disabled in each case.

Sandy Bridge Motherboard Advice P67 or Z68 DPC Latency