** OFFICIAL ** Intel DX38BT & DX48BT2 BoneTrail Thread

[Manel]

Funny how these machines get on eBay so cheap. Good deal.

Guess I'll take the time someday to understand how the BIOS is organized and deal with it. Still, how do you remove the SLIC table without reflashing it?! Is there also an easy way to remove nVidia's SLI protection for boards older than X58?
But BIOS will die this year so I guess all this will become academic.
Maybe with UEFI things are alike, and the same hacks are possible.

[Manel]

Tomorrow IE9 will be out, can't wait to try it out.
Intel's page is out...

[dukzcry]

Guess I'll take the time someday to understand how the BIOS is organized and deal with it. Still, how do you remove the SLIC table without reflashing it?!
But BIOS will die this year so I guess all this will become academic.

Sorry, that is all my weak English. I didn't mean that you can fully avoid changing of something in EEPROM (if you're interested in activation technique which involves implementation made in software only (which doesn't touch BIOS and so on), consider looking at "SLIC emulation", i recommend "Windows 7 Loader eXtreme Edition" by napalum which offers a dozen of different activation methods (both safe and unsafe)). I just meant that the process of adding SLIC tables to Intel boards is much safer and easier than with motherboards from OEM vendors. You're not required to know much about BIOS internals, i'll just quote meaningful excerpts.

Some BIOSes contain a "SLIC" (Software Licensing Information Code), a digital signature placed inside the BIOS by the manufacturer, for example Dell. This SLIC is inserted in the ACPI table and contains no active code.

News that BIOS will die don't mean that it will not be used in futher x86 OEM products. Intel will try to be trend-maker there, will try to make coexistence of BIOS and logicals of new boards less possible, but who cares. There are a lot of companies, which business is built exclusively or mostly around BIOS (Phoenix Technologies, American Megatrends just to name few) and i think that they will continue to do their work in this field. They have (and will) all required datasheets, both public and internal (NDA) via contract with Intel. In addition, don't forget about AMD, which, as i presume, is not interested in UEFI so much.

What about me, i find both BIOS and EFI ugly. The most elegant, and on the other side expandable firmware interface that i was pleased to play (which doesn't mean that there is no other nice computer firmware interfaces in the world, but i never used something better, maybe they exist only as academical prototypes or non-mainstream general purpose systems) with was an Open Firmware (also known as OpenBoot). You can find realization of this standard in old PowerPC Macintoshes or SPARC64-based machines from Oracle/Sun. In it's pure form, the firmware provides CLI-based interface and embedded Forth language interpreter. Possibility to use serial console or high-resolution framebuffers for outputs was also been there, when BIOS did provide only low-res output, and look now, Intel still tries to jump over BIOS suckness, inviting stupidnesses like AMT technology The Forth nature of OpenBoot also makes possibility of easy migration of the same logicals and periphery to another processor architecture without full firm. rewriting. EFI looks like weak clone of OpenFirmware with much of ugly overhead .

[dukzcry]

Manel: This is an example of what you should usually do for SLIC-based activation on non-Intel boards. Of course, for most of modern OEM boards there are special forums/topics where are people modify and upload BIOS images, so, usually, you don't need to do this work by yourself, for most popular boards, but you're still forced to flash your BIOS with new image. Nobody will be responsible for faults at your side, needless to say.

[Manel]

Thanx dukz. I remember having read some of this at mydigitallife some time back, when I was installing Vista for the first time. They have a lot of info on this subject. Still, modifying the BIOS was not for me at the time so I kinda forgot about it. Anyway, anyone modified Intel's BIOSes already?
You are well into this BIOS/firmware issues I see, do you work on this area or is it just for fun?
I find it great that Intel is trying to let go of BIOS, it's about time we ditch such old tech. I'll believe you if you say UEFI is still far from perfect but I guess at least it should be better than good old BIOS, or not? It has been used in server systems and HPC and even the newer Macs for ages, right?! It must be somewhat good I guess.
I know about Open Firmware, is it still being developed? Yes, I guess it's leaner than UEFI.
What annoys me are those Taiwanese OEMs that now pimp up the UEFI interface to make it look good (if you can call it good, horrible for me), Intel at least still makes it simple, same looks as BIOS (not that I like it either). Couldn't everyone just keep it simple and intuitive?
Again, thanks for the info. I'll take the time to get up to date on this.

[Manel]

It seems Intel already has 2 Z68 boards ready: the all-out DZ68BC and leaner DZ68DB.
More or less same features of the other 6 series chipsets (without the bug), better oc'ing (now proc, mem, graph core), ssd cache as soon as RST v10.5 is released.
Rumours say that Z68E will be the new high-end chipset, more PCIe lanes, no integrated HD graphics, ring bus. Guess it won't be X68 after all. Did Intel drop the X for Xtreme Series? Sandy Bridge-E procs are rated 130W TDP as expected, 4 or 6 cores (wot?! no 8 core?), larger L3 cache, socket 2011.

[dukzcry]

Anyway, anyone modified Intel's BIOSes already?

There is no need to do this. As i mentioned already, that internal utility by Intel will do all the work for you. It will write SLIC table extremely fast, no need to wait for full reflash.

You are well into this BIOS/firmware issues I see, do you work on this area or is it just for fun?

Just ran across of it few times, nothing more.

but I guess at least it should be better than good old BIOS, or not?

Of course, it is a killer thing in comparison to BIOS, so nothing to worry

I know about Open Firmware, is it still being developed?

I think that guys from OpenBIOS project are still working on it. Regarding the open-source realization, take a look on OLPC laptops.

What annoys me are those Taiwanese OEMs that now pimp up the UEFI interface to make it look good (if you can call it good, horrible for me), Intel at least still makes it simple, same looks as BIOS (not that I like it either). Couldn't everyone just keep it simple and intuitive?

They just have a very bad taste in UI designing

[Manel]

LAN drivers v16.1 out but, as usual, only for X58 an 6 series. They should work on all hardware though, only Intel won't update discontinued product pages I guess.

[petr0id]

bye-bye PCI

[Manel]

Thanx petr0id.
Finally the specs for the much antecipated and awaited X79, if this is actually the final name. So much has been speculated about this till date. 7 series after all.
Seems it will be on my next rig. Still, PCIe v2.0, even after PCIe 3.0 has been out since last November? Not even PCIe 2.1?! Either this is fishy or eXtreme Fail. Not that it's actually needed but still... thumbs down Intel.
The other thing here is lack of USB 3.0, Intel was supposed to enable it in their next series of chipsets.
The 10 SATA 6G ports out of 14 is awesome, although here somewhat exagerated in my opinion. Even for enthusiasts. But the more the merrier, not complaining here!!
Yes, PCI should be long dead already, 6 series had already killed it off.
Good thing Intel is going (finally) legacy free. Still, they provided a PCIe-PCI bridge on their DP67BG and other. Make up your mind Intel - let it go for good.

[Manel]

I suspect this is more of the enterprise variant of the socket 2011 chipset than the enthusiast X79, or a mix of both. how reliable is this info? The SAS support and RSTe betray the orientation.
The integrated audio codec seems cool if it's any good. I can already see driver issues or lack of support from Intel.
Those 4 additional PCIe lanes from the proc puzzle me. It seems the pin count is already prepared for the additional width linking the proc and PCH. Wow!
If OC is as good or better than what we've seen latelly, this will blow anything out of the water.
Man, I'm already itching.
For when, does anyone know? It was supposed to come out Q3'11 or so.

[Manel]

djs said before that the Ivy Bridge PCIe links are in fact 3.0, although Cougar Pointwould will only support PCIe 2.0.
Maybe it's the same here, although the procs are "only" Sandy Bridge. Why not Ivy Bridge-E, Intel? That way we could have future GTX6xx ot HD7xxx cards working full speed off the CPU, if they'll be PCIe 3.0 that is.
I guess PCIe 2.0 for any others peripherals is good enough, at the time. Although with that amount of SATA 6G ports, 14 USB ports (more so if some of those come out to be USB 3.0), GbE - it kinda gets crammed.
I guess BIOS will be out of the equations as well, as most of the other legacy crap Intel has been getting rid of. Good work there.
Not even TB is mentioned, as an option to USB 3.0, with USB 3.0 this would have been a kill all solution. 3rd party controllers I prefer to avoid.

[djs357]

Don't confuse the PCH capabilities with the processor capabilities. x79 will target ivy bridge, although m/bs can support both. IB will have 16 PCIe 3.0 ports at 8GBs and optional 4 PCIe 2.0 for speciazed apps. x79 looks to be an enterprise chip rebranded. USB 3.0 yes, SATA 3.0 yes. Don't expect them soon though.

[Manel]

I wasn't, just the info isn't very specific. Now I found the article, source was chinese so I'm not sure how accurate this is. It's slated for Q4'11.
You mention PCIe lanes from the proc but I couldn't find that info. Also, I can't see any reference to it being based on IB rather than SNB, much on the contrary. But hey, you know your way around Intel classified news so I trust you're correct. Hope so!
It seems you can, through UEFI, configure the extra 4 PCIe 2.0 lanes as additional width for the link CPU-PCH. Awesome bandwidth!!
Mem quad channel seems confirmed.
From what you say, socket 2011 will support Ivy Bridge as well? Or are you referring to Sandy Bridge-E instead. Because, at least mainstream, Ivy Bridge will continue to be socket 1155 I believe.
Unless all the info is incorrect and the new enthusiast processor will in fact be IB instead of SNB-E as is beiog reported.
And where did you see USB 3.0 support? Al over it's only 2.0 that is confirmed.

[djs357]

It is IB-E, SNB-E. There will be USB 3.0 on the PCH. The variants are starting to get to be tomany.

[Manel]

Guess we'll have to wait for more reliable info. Been reading some "old" (Jan. 2011) info on semiaccurate too regarding Patsburg and some inside info refers 4 Patsburg SKUs but for the Xeon platform. All is very confusing and differs from the current specs. I guess things changed meanwhile.
Good to see you're so sure regarding USB 3.0 and PCIe 3.0 being there. It's just that it makes the platform more future proof. Cycles of new platforms are getting too short, although not in the enthusiast segment, X58 stood it's ground.

[Manel]

And what about Intel revealing that will ship Virtu with some select desktop boards based on Z68 and H67? Well, that looks to me that they're admiting that some Israeli guys can do something they can't, right? Although I believe Intel has some cut in LucidLogix. But still that makes you wonder why they couldn't do it in hardware in the first place and develop their own app to do it.

[Manel]

Intel has updated Intel Desktop Utilities and Marvell RAID drivers for some boards.
Nuvoton CIR drivers had already been updated recently.
Funny how there have been no BIOS updates recently, maybe they're already perfect?

[Manel]

Intel Xeon E7 series with upto 10 cores + HT... for mission-critical applications.

[Manel]

Last stop for the LGA1366 platform: Intel intros the i7-995X Extreme Edition proc with 6 cores (plus HT) @3.6GHz (Turbo @3.86GHz), Gulftown 32nm, available in Q3'11. After this the void, no more X58 I guess, just in time for the X79 in Q4. As usual, it comes with a $999 price tag.

Also recently introduced was the first DDR4 modules by Hynix @2400Mbps, 2GB ECC-SODIMM modules based on 30nm 2Gb chips, 1.2V. Volume production should start 2H'12, maybe mainstream by 2015. Apparently doubles throughput. Maybe not even for the platform next to the upcoming, which is what? X89?! Maybe X99 will be DDR4, looks darn good at least.

[Manel]

Wow, Intel updated their BITS download page, they're now actually providing the latest version 372, which (as usual) is not common. Just before they were only providing 329 if I recall correctly.

[djs357]

Those new Xeons are quite nice. I really don't have one. Never know, but it would make my dual 55xx server jealous.

[Manel]

Yes, they look great. If you have such a server already, I guess they wouldn't get along well together. If they're like women, they can't stand the competition!! :-)

Funny, Intel has posted already drivers (video, LAN, INF, ME, Audio, BIOS, RAID, Renesas USB 3.0) of the new DZ68DB but the boards isn't listed on their web site still. Good ol' Intel, always the same.
No documentation yet so we can peak at the specs though. BIOS 0010 readme only states that it's the initial production BIOS release.
Well, I guess it shuld be the same as the older P/H67 boards, give or take.

[Manel]

It apperars Z68 will be out on May 8th.

[dukzcry]

http://translate.google.ru/translate...ntel-dp35dp%2F

Anobody tried this already? No profit of doing that, unfortunately.