Asus Z87 w/ 3 x PCIs

[dawman]

Might be nifty, and can expand to the XITE-1 later on the PCI-e 1X slot.

[siriusbliss]

Looks like it has some potential.
Wondering about shared IRQ's, but otherwise...

Thanks,

G

[garyb]

it would be amazing if it worked, but it's a beautiful motherboard.

those things don't have real PCI slots do they?

[dawman]

Sure look real to me, but personally I would wait until later and then use the MSI Military version, not the gamer style ones.
But I have no proof if the PCIs work, but I can't imagaine the outcry for PCI slots being cush a factor on recent designs, and then have them not working...

[garyb]

very few cards that would be used in the PCI slot are realtime high performance audio cards. the realtime part is the deal breaker.

every single socket 1155 and newer motherboard has been shit so far. every. single. one.

[dante]

The gamer style extreme look mean though. But I think this ver is PCIe slots rather than 3 x PCI.

extreme

extreme.JPG (117.49 KiB) Viewed 4312 times

[frokka]

Gary, what does "realtime PCI" mean in actual hardware specifications?
Is it a PCI latency timer issue?

[garyb]

mo, it's that Scope is a realtime device. zero latency=realtime, like hardware. Scope IS hardware. it's a digital system that's controlled by a computer, NOT a windows soundcard. through a driver, it can pass audio and midi back and forth to computer apps, but it's operation is independant. however, for many things, Scope cards DO access and use the computer's resources, which must be available exactly when expected. this is why Scope cards don't like to share IRQs. when the card needs to communicate with the cpu, it needs to do so NOW, when it's turn comes up. when it needs to access sytem ram, that needs to happen NOW. total theoretical bandwidth means nothing if it is late. that's the cause of most PCI overflow messages, or that's something very close to the proper layman's explanation....

computers are not realtime devices. what happens when you mate a realtime system to a non realtime system?
well, mayhem possibly...

a musician can understand. timing is everything.

[dante]

very few cards that would be used in the PCI slot are realtime high performance audio cards. the realtime part is the deal breaker. every single socket 1155 and newer motherboard has been shit so far. every. single. one.

The Z87 is slot 1150 (not 1155) - does that mean theres a better chance the PCI slots will work well ?

I'm interested in this as my next mainboard and having 3 cards still.

[garyb]

since it's the same architecture as 1155, probably not, but someone will need to be a guinea pig and try it....

[dante]

Well I was thinking of buying the board and keeping it till I could afford rest of system in which the meantime a guinea pig might emerge with a report.

[garyb]

it's likely that just like the ISA cards, that the PCI card is done. this doesn't mean that there will be no more motherboards with PCI cards, just that they will not be common, and they will not be cutting edge, they will be old server designs. we are almost there.

that doesn't mean that the PCI cards are finished, many people still use ISA tripleDAT systems because they work and sound great. never mind that we can't provide keyfiles for them if the keys are lost...

i'm always hopeful, but at this point, i strongly recommend 1366 or used 1156 system. if you must have new, a Supermicro Dual 1366 XEON system won't be obsolete or underpowered for a long time. you can find a very powerful 1156 system for the price of a current processor as people dump them for newer computers that they don't need.

as usual, it's just my own opinion. i just hate to see money wasted. it's not that i don't think that the new designs can perfom well, it's just that i don't believe that thye manufacturers care, especially when the only people who are really hurt by shoddy design are people with realtime high performance PCI cards. the idea of using a card for 15+ years is anathema to those that run the industry and there aren't enough Scope owners to put enough pressure to demand that these things work correctly.

[dante]

Thanks Ill keep an eye out for supermicro as well. Is there anything on here like that :

http://www.msy.com.au/Parts/PARTS.pdf

page 1 left hand column is motherboards. I dont see supermicro there though

[pwhitmaker]

I bought this one, ASUS Z87-K.
I found it good for my 2 Pulsar cards, with the Graphic card at the yellow PCIe place at the top, and 2 Pulsar cards at the lowest 2 PCI.
It seems like the old PCI contacts are going to disappear on motherboards more and more.

[garyb]

that's great that it's working. a lot of people who want new computers are really bummed out that they don't have the latest choices.

dante, you won't find Supermicro in too many bargain sellers' lists. it's a premium motherboard for server applications. they're very stable and made to last, but not cheap. the have real tech support with people that understand how inportant your uptime is.

but heck, is the z87 is working, that's way easier.

[dante]

Coolness thanks all I guess I might bite the bullet

[frokka]

pwhitmaker, could you do the Masterverb load test on your new PC?

gary, thanks for the reply.
I basically understand the difference from PCI cards with DSPs to those with only a controller and some IOs.
What I'm still somewhat unsure of is what exactly causes a certain motherboard to handle PCI traffic disfavourable to Pulsar / Scope? The reason I mentioned PCI latency is because that is one of the things that will affect when and how often the card can access the other resources, as you stated: "when it needs to access sytem ram, that needs to happen NOW. total theoretical bandwidth means nothing if it is late. When the pulsar cards were new I remember people used to tweak this to improve performance, but nowadays I'm not sure if all BIOSes have this option available.

[pwhitmaker]

pwhitmaker, could you do the Masterverb load test on your new PC?

gary, thanks for the reply.
I basically understand the difference from PCI cards with DSPs to those with only a controller and some IOs.
What I'm still somewhat unsure of is what exactly causes a certain motherboard to handle PCI traffic disfavourable to Pulsar / Scope? The reason I mentioned PCI latency is because that is one of the things that will affect when and how often the card can access the other resources, as you stated: "when it needs to access sytem ram, that needs to happen NOW. total theoretical bandwidth means nothing if it is late. When the pulsar cards were new I remember people used to tweak this to improve performance, but nowadays I'm not sure if all BIOSes have this option available.

Frokka, I have done the masterverbtest with this motherboard and I wrote the result in that thread.
It wasnt any better result than my old PC but according to garyb its because one of my two cards are Pulsar1.

[garyb]

well, i said that the number was normal for a P1. it remains to be seen if it gets normal results with only a second gen card, but i would expect it to work normally, if it's normal with a P1.

the PCI latency timer might make a small difference, but it won't be the difference between works or doesn't work.

[guppy]

I found that Supermicro have a solution for scsope boards and 1150
http://www.supermicro.com/products/moth ... 0SLA-F.cfm