Monday, May 2nd 2011

Computex 2011 Launchpad for Intel X79 Motherboards

At Computex 2010, motherboard vendors across the board displayed socket LGA1155 motherboards that support Intel "Sandy Bridge" processors, based on P67, H67 chipsets, months in advance of the platform actually making it to the market. This year, the motherboard industry will do something similar and show off socket LGA2011 motherboards based on the Intel X79 "Patsburg" chipset.

Detailed to much length in older articles, the platform schematic of Sandy Bridge-E surfaced, confirming X79's feature-set, including a PCI-Express 3.0 based supplementary interconnect between the processor and chipset to bolster enough bandwidth for the massive 10-port SATA 6 Gb/s controller, and a 8-port PCI-Express 2.0 hub. Sandy Bridge-E processor itself comes in three main variants, an all-enabled 6-core Extreme Edition, a 6-core unlocked variant, and a limited-OC 4-core variant. The platform is slated for late 2011.

Source: VR-Zone
Add your own comment

25 Comments on Computex 2011 Launchpad for Intel X79 Motherboards

#1
csendesmark
No native USB 3.0? :(

Intel wants USB dead.....
Posted on Reply
#2
btarunr
Editor & Senior Moderator
by: csendesmark
No native USB 3.0? :(

Intel wants USB dead.....
That 10+2 port SATA controller won't leave any need for third-party SATA controllers for eSATA or additional internal ports. That leaves the 8-lane PCI-Express 2.0 hub of the PCH to accommodate several third-party USB 3.0 controllers. Who knows, there could even be a massive 8 port USB 3.0 controller being developed by someone that takes up a PCI-E x4 port.
Posted on Reply
#3
csendesmark
by: btarunr
That 10+2 port SATA controller won't leave any need for third-party SATA controllers for eSATA or additional internal ports. That leaves the 8-lane PCI-Express 2.0 hub of the PCH to accommodate several third-party USB 3.0 controllers. Who knows, there could even be a massive 8 port USB 3.0 controller being developed by someone that takes up a PCI-E x4 port.
I know that! But AMD will integrate it's USB controller into the SB :)
Posted on Reply
#4
Trackr
by: btarunr
That 10+2 port SATA controller won't leave any need for third-party SATA controllers for eSATA or additional internal ports. That leaves the 8-lane PCI-Express 2.0 hub of the PCH to accommodate several third-party USB 3.0 controllers. Who knows, there could even be a massive 8 port USB 3.0 controller being developed by someone that takes up a PCI-E x4 port.
By 'third party' you mean what - an expansion card?

If there are no USB ports on the motherboard, it will be a bad day for inter-connectivity.
Posted on Reply
#5
slyfox2151
by: Trackr
By 'third party' you mean what - an expansion card?

If there are no USB ports on the motherboard, it will be a bad day for inter-connectivity.
no, it will still be on board.
Posted on Reply
#6
csendesmark
by: Trackr
By 'third party' you mean what - an expansion card?

If there are no USB ports on the motherboard, it will be a bad day for inter-connectivity.
It can be an extra chip on the mobo, or as a PCIe expansion card.
As the very first gen NEC USB3 chips were presented :)
Posted on Reply
#7
btarunr
Editor & Senior Moderator
by: csendesmark
I know that! But AMD will integrate it's USB controller into the SB :)
Makes no difference if USB 3.0 is integrated or not. With SATA 6 Gb/s, chipset integration mattered, because a single large controller would let you set up RAID configurations with over two disks (as opposed to using many third-party 2-port controllers), with USB it's not the case. You can have external multi-port controllers, or a number of 2-port controllers, or even 2-port controllers with multiplex chips. Doesn't matter. Cramming those into the chipset won't make it any faster or better.
Posted on Reply
#8
Fourstaff
by: btarunr
Cramming those into the chipset won't make it any faster or better.
Makes it slightly cheaper though. I heard Panther Point will have USB3 integration, is that true?
Posted on Reply
#9
csendesmark
by: Fourstaff
Makes it slightly cheaper though. I heard Panther Point will have USB3 integration, is that true?
When you are buying a motherboard for $700 it really doesn't matter.

@btarunr
I just noticed it, and would be better from software side, because Intel driver usually great!

I know, that is no speed difference between 3rd party and built in controllers.
Posted on Reply
#10
btarunr
Editor & Senior Moderator
by: Fourstaff
Makes it slightly cheaper though. I heard Panther Point will have USB3 integration, is that true?
Nope, it will just increase the die area and TDP of the already big PCH, and Intel will add it to its costs.

Panther Point has a 2-port USB 3.0 controller that is multiplexed between two hubs of the USB 2.0 controller to give four USB 3.0 ports. Speeds will suck, if there are more than two bandwidth-heavy devices connected.

On the other hand, using a number of 2-port controllers ends up being bandwidth-efficient, and those NEC/Renesas 2-port controllers have become dirt-cheap after getting competition from ASUS (ASMedia).

by: csendesmark
When you are buying a motherboard for $700 it really doesn't matter.
Yup.

by: csendesmark
I just noticed it, and would be better from software side, because Intel driver usually great!

I know, that is no speed difference between 3rd party and built in controllers.
Intel's USB 2.0 controller "driver" isn't much more than OHCI driver with name strings for the controllers.
Posted on Reply
#11
de.das.dude
Pro Indian Modder
bravo intel. please leave out the USB. i want you to crash and burn LOL.
Posted on Reply
#12
DarkMatter75
What about native trim support when using RAID?:confused:
Posted on Reply
#13
TheLostSwede
by: DarkMatter75
What about native trim support when using RAID?:confused:
Not yet afaik, although it's meant to work when you're using smart response, but I guess that's not useful for everyone...
Posted on Reply
#14
Fourstaff
by: btarunr
Nope, it will just increase the die area and TDP of the already big PCH, and Intel will add it to its costs.
You can always remove some of the 10+ USB2 ports and swap it out with some USB3, surely the die area will not be increased by a significant amount?
Posted on Reply
#15
TheLostSwede
by: Fourstaff
You can always remove some of the 10+ USB2 ports and swap it out with some USB3, surely the die area will not be increased by a significant amount?
It's not that easy and because of the speed of USB 3.0, the silicon dedicated to it will have to run very fast which also means it it produces more heat. Of course this is still fairly nominal, but compared to USB 2.0 it's a huge difference, least not the space inside the chipset that's being taken up by the USB 3.0 part. Of course some of this depends on the manufacturing technology used to make the chipset and it's at least more advanced that that used to make the host controllers so the actual die size wouldn't be nearly as big as say one of the Renesas chips which are made using 90nm tech if I remember right.
Posted on Reply
#16
erocker
Why not put USB 3.0 on a separate chip? Takes stress off of the South Bridge, PCH, North Bridge, whatever. Doesn't make a difference at all to the people using it.
Posted on Reply
#17
Fourstaff
by: erocker
Why not put USB 3.0 on a separate chip? Takes stress off of the South Bridge, PCH, North Bridge, whatever. Doesn't make a difference at all to the people using it.
That is what they are doing right now :slap: The point of having USB 3.0 in the SB/PCH/NB/etc. is to "officially" make it a mainstream connector. Purely symbolic, but the gesture is reassuring.
Posted on Reply
#18
zads
by: btarunr
...Who knows, there could even be a massive 8 port USB 3.0 controller being developed by someone that takes up a PCI-E x4 port.
Going without any USB3.0 is a bit disappointing, but Intel isn't invested in USB3.0 so they might as well let it be someone else's problem/investment/development in the add-on chip space.
I'm sure the USB3 bridge companies aren't complaining :p

I don't see much usage for USB3 at the current point in time for any of my needs,
I'd rather hot swap a faster, slightly larger, 2.5" SATA2 SSD drive than fumble around with the (fat) USB3 PEZ-dispensers.
The USB3 UFD (circuit/bridge) designs I've seen are all pretty lame too.
Posted on Reply
#19
AsRock
TPU addict
by: erocker
Why not put USB 3.0 on a separate chip? Takes stress off of the South Bridge, PCH, North Bridge, whatever. Doesn't make a difference at all to the people using it.
+2,

Only difference i have noticed is that windows boots a extra app.
Posted on Reply
#20
[H]@RD5TUFF
by: de.das.dude
bravo intel. please leave out the USB. i want you to crash and burn LOL.
Do you really believe USB not being integrated on the chip will kill intel ?:wtf:

Has anyone considered that this is not the full feature list ?

Also AMD please keep making slow chips and horrible drivers for your video cards I want you to continue to crash and burn LOL.
Posted on Reply
#21
btarunr
Editor & Senior Moderator
by: zads
Going without any USB3.0 is a bit disappointing, but Intel isn't invested in USB3.0
It will be disappointing if a motherboard vendor is dumb enough to sell a LGA2011 motherboard without any USB 3.0 ports. Until then, reserve that conclusion.

To the end-user where those USB 3.0 ports are coming from doesn't matter. If it's integrated with the PCH, its cost is factored into chipset cost, and if not, 3rd party controllers/hub chips.
Posted on Reply
#22
Funtoss
lol this doesnt look that great but still its good :)
Posted on Reply
#23
Omnisome
Fuck Intel, fuck technological progress.

A new socket every year?!
Posted on Reply
#24
[H]@RD5TUFF
by: Omnisome
Fuck Intel, fuck technological progress.

A new socket every year?!
I prefer a new socket every 2 years, with nearly 2x the performance of the previous gen, than buying a mobo and using it for 3-4 years, and buying a new processor every year with marginal performance gains . . .
Posted on Reply
#25
Animalpak
wich are the real improvements with this new giant socket ?
Posted on Reply
Add your own comment