More to come...
Printable View
More to come...
Note that the HS on the X38 is the same size as the 975X but uses a screw type retention mechanism. The bolts are in the same place the wire loops were. Also notice that the X38 has an IHS.
Where's the skull SB cover? :lol2:
LoL. When ya gonna start clocking?:stick: got a B3?
Good luck on this thing, it looks great :up: :up: :up:
I watched your posts alot in XBX and XBX2 threads, so I'm eagerly awaiting your thoughts on this board. GL. :up:
subscribed :up:
my board as soon as they get stock on internal intel website.
I thew together a quick edit of FCG's CAS latency window optimizer spreadsheet that has the strap/memory frequencies worked out for the bonetrial for everyone to use.
check out FCG's thread here
Leak testing...
Had to rig a mount for my MCW30 chipset block but I used spare Swiftech parts so I guess it's OK. :)
If you're thinking about upgrading from an XBX2, don't throw away the extension cable you made for the BIOS jumper. It will find a new home with this board.
Also, I think I'll do a wiki over Thanksgiving so that more people will contribute and things will be less likely to get out of date.
Oh yeah, this board looks like it uses the same aSC7621 sensor chip as the XBX2 so I'm hoping Everest, Speedfan and the others will work.
Got a custom (thicker) backing plate for that GTX? If so, soo sexy... 1 other observation, the DIMM slot closest to the PCIx16 seems real close like my P5WDG2-Pro. I had real interference issues with the GTS320MB video and switching ram out...
Swiftech is shipping the GTXs with backplates so it's nice and snug with little stress on the MB. The copper top for the GTX is actually an add-on. Stock one is anodized aluminum.
The DIMM retainers are close, but I just tried and was able to get the stick out and back in again without much fuss.
30 minutes to go on the leak test...
After a bit can you load this up with 4 gig of RAM and report impact on dual core max fsb? Also, can you run a few prime95 sessions with the higher fsb to see what's actually stable as opposed to just benchable? :)
Stupid Corsair. I've always had great luck with OCZ but this time I went with Corsair. When I opened the package, one of the heat spreaders just fell out with one of the chips stuck to it. So I've got some OCZ on the way and in the mean time, I'm operating with only 1 stick.
keep that hw-pron coming!
already ordered the cheapest ddr3 i could find (g.skill F3-8500CL7D) and waiting for the board to get in stock over here. :rolleyes:
$579 for 2x1024mb. Already got an RMA and ordered the OCZs. They should be here tomorrow. Till then, I'm limping along on 1 stick.
Subscribed.
I am having an issue with my DX38BT. I'm a newb at this so don't laugh. I have Corsair XMS3 that is rated @ 1333, 9-9-9-24. I set it in the bios, and the settings stay there on the setting page after reboot, but the memory only runs at 1066. The morons at Corsair aren't any help. Why won't the settings stick ?
Memory Configuration = Non-ECC
Performance Memory Profiles = Manual - User Defined
Reference Frequency = 333
Memory Frequency = 1333
tCL = 9
tRCD = 9
tRP = 9
tRASmin = 24
Memory Voltage = 1.70v
The main bios screen says the CPU is @ 1333FSB and the memory is at 1066.
Everest and the Intel Desktop Utility also says 1066 on the memory. CPU-Z doesn't support the X38 yet, so I get no report other than SPD values.
In the memory bios screen check what the SPD is reporting. Initially my sticks were actually reporting DDR3 1066 and that was goofing things up. unfortunately I don't remember what I did to fix it. Also try going back to Automatic and see what happens
More observations while I'm waiting for my OCZ to arrive (9am tomorrow hopefully).
The BIOS Watchdog is as useless as mammaries on a male Sus Scrofa. Same as the XBX2. Thankfully the default is "disabled".
If you put the Intel controller in RAID mode, and you have a SATA DVD drive, USB keyboards don't work.
The Marvell SATA controller also handles IDE port so disabling the Secondary SATA Controller also disables the internal IDE port as well as the external SATA ports.
Vista will install cleanly on an Intel RAID array without any additional driver stupidity.
Vista doesn't include the driver for the new LAN controller so you have to install it separately from the driver CD.
The mosfet heatsinks have a much lower profile than the XBX2.
The vCore vMCH and vDIMM measurement points seem to be close to, if not exactly in the same place as the XBX2. I'll post pics later.
The BIOS screens are pretty close to the XBX2. Pics later.
definitely subscribed, but there's no way I'm upgrading until next year at the earliest!!
Looking good so far, I just hate the thought of paying 500+ for DDR3...
Been there. I worked with a tech for about an hour, and he said the board was new and he couldn't solve the issue. Currently there is no qualified memory for this board ?
SOMEBODY must have this same board and 1333 memory. There has got to be a setting to resolve this.
Did you try rev.1308 bios?
Quote:
BIOS Version 1308
About This Release:
November 12, 2007
BTX3810J.86A.1308.2007.1112.1659
SATA RAID info:
Intel(R) RAID for SATA - v7.6.0.1011
SATA AHCI info:
Version UPSD src 04-20-2007 for ICH9
PXE Nahum info:
Intel(R) Boot Agent GE v1.2.50
New Fixes/Features:
Fixed issues where some BIOS Setup options are hidden.
Added updates to BIOS support for Intel® Desktop Control
Center.
Fixed display of cache for multi-core processors.
Fixed PC28 memory error beeps on warm boot with 1333 CPU / 1333
Memory.
Fixed issue where XMP settings were causing system instability.
Fixed issue where overclocking settings were causing HW
Watchdog Timer to fire on every boot.
Added Hardware-based Watchdog Timer implementation.
Added support for XMP (Extreme Memory Profile).
Added some boot/S3 resume time optimizations.
Fixed a bug where IERR got asserted on certain CPUs.
Well, I'm running at 1333 on the OCZs but like I said, I can't remember what I did. How about going into maintenance mode and doing a bios settings reset?
How's this board's overclocking compared to ASUS or DFI?
Well, at this point what it looks like is I can't overclock the CPU or memory. Tried overclocking the FSB just a little bit and it bounces right back to default. Maybe this board is hosed ? It must have something to do with the watchdog feature. I think I am going to throw in the towel on this one until Intel get's a better bios out. Funny thing is, the latest bios was supposed to fix a ton of related issues, and intel is usually pretty good about that. If that OCZ is 1333 by SPD, I may take a stab at it - why not, I got about $900 of memory sitting in front of me now, what's another $500.
I like where intel is going...Black PCB, hardcore(ish) names for stuff..Remember the SkullTrail monster some time ago? They are showing a little more interest for the xtreme crowd..
But...Yadda yadda yadda...whatever..!
Waiting to see some results!!:up:
does the DX38BT support 1600FSB like asus's x38 lineup? :shrug:
Tried everything at least twice. Can you adjust your mem voltage, or OC your memory at all ?
See, the problem is that the 1333 is not programmed in SPD on my memory, and I think that is what this bios is stuck at. I even try to underclock it to 800mhz, and it will bounce back to the 1066 by SPD settings.
http://i219.photobucket.com/albums/c..._01/patmem.jpg
Downloaded the new CPU-Z and it reports the memory is at 1333 ? Even the bios and Everest report it @ 1066. I wonder if CPUZ and Everest draw the info from different areas.
OK, I did some playing around and I think it's a BIOS bug. I can't boot at all with the memory running above the SPD setting and when I set it below the SPD setting, it still reports the SPD setting. I'll goof more with it later today.
I played with the memory settings yesterday evening after a long nap.
First, don't trust the BIOS main screen memory frequency display. It's only occasionally correct.
With the OCZ Platinum DDR3-14400, here's what combinations worked for me with all other settings at stock...
333/800
333/1066
333/1333
400/800
400/1066
No other reference/memory frequency settings worked (hung POST or 3 beeps) even thought they were well within the capability of the memory.
Since the best setting is 333/1333 at stock FSB, I then started raising the FSB to see what happens. Right now, I'm at 378MHz FSB with a DDR3-1511 and having no problems. So I believe the problem is not that the board can't handle the speed but the BIOS just won't let you manipulate the settings correctly.
Randy: Try my working settings and see what you get. When you try each one, go into the Everest Cache and Memory Benchmark and run a test. Keep track of the results and see what changes. Try raising the FSB as well and see if you're stable. Also create a forum signature that has your system components in it. It'll help.
Ok,
seems to be working here:
333/1333
http://www.acsmicro.com/acsftp/Forumpics/333&1333.JPG
345/1333
http://www.acsmicro.com/acsftp/Forumpics/345&1333.JPG
360/1333
http://www.acsmicro.com/acsftp/Forumpics/360&1333.JPG
Sorry,
I"m kind of a woos, just don't want to hose up this array ! Am I safe or How far should I try to push this ?
Everything @ 1.70V for these benches.
I was able to push to a 378 FSB without any voltage changes but your memory might not take it with the memory freq set to 1333. You have to decide what tradeoffs are acceptable but I'd try going up to a 378 FSB and lowering your memory frequency to 1066. This will put you at 3.4GHz for a processor speed but a slightly underclocked memory speed of 1210MHz.
If you're worried about your array, burn a Memtest86+ CD and use that to do an initial test of new settings.
I also created a version of the XBX2 memory calculator for the X38 so you can use that to see what happens to your memory at different FSB speeds.
Well, here's where I'm at this morning...
Very pleased with the FSB/Processor so far.PHP Code:
vCORE: BIOS: 1.425 SENSOR: 1.37 Measured: 1.397
vMCH: BIOS: 1.375 SENSOR: 1.33 Measured: 1.388
vFSB: BIOS: 1.325
FSB: 434MHz Processor: 3.902GHz
Reference/Memory Freq: 400/800
Timings: 7-7-7-20
vDIMM: BIOS: 1.54 Measured: 1.542
8 Hours Orthos stable
Still some work to go on the memory. At 400/800 it's running at half speed.
If I back the FSB back down to 423 (3.8GHz), I can run the memory at 333/1333 (DDR3-1692) and be perfectly stable.
I just popped the FSB up to 445 for a 4.0GHz processor speed and it started right up and is running Orthos now.
If only it supported SLI...
This board is so far the most horrible mobo that I have ever had the displeasure of working with.
The BIOS is more buggy than a termites nest. It wouldn't even POST my E4300. I put in a Quad and it wouldn't boot that without adjustments. Mobo would show that my CPU had no multiplier, memory display is off....I had problems right from the get-go.
Icannot flash the BIOS to a more recent version becuase there is no f*cking floppy drive AND I cannot get my OS to install (more below). I would have to create a bootable USB drive and that is a huge pain in the ass and that is the only other option that I have.
Now I cannot get my Intel RAID drivers to load in XP using my USB thubbdrive!!
It will sometimes read the thumbdrive to load the RAID drivers, but then again ask me again later to insert a diskette in drive A. If it doesn't at least do this, it will simply tell me that it cannot find a floppy drive....no :banana::banana::banana::banana:, there isn't one...and yes i have USB devices set accordingly in the boot section of the BIOS.
BIOS boot setting keep changing for no damn reason and I have to keep making adjustments.
I've been f*cking with this board for HOURS and HOURS trying to get my damn RAID drivers to work so that I can install XP.
I hate this board so far.
Might I suggest doing a quick XP install with the controller set to IDE mode so you can flash the new BIOS?
I disconnected one of my 80Gb Raptors and I'm trying to reinstall XP on a single drive now...for the second time.
Had major problems trying to get the mobo to boot to my CDROM after the first attempt to install XP failed...
I'm hoping to get to the XP desktop and flash the BIOS ASAP.
EDIT: 2nd BSOD while installing windows...this time it was with a E4300 instead of my Q6600, which was used on the previous attempt.
This mobo has to be defective...no good working board should give someone this many problems....
There IS a bug in the BIOS where the boot order isn't honored in the "normal" mode. If you switch to the "Advanced" view on the boot screen you can set the proper boot order.
Hi.
Yeah I figured that out....
Advanced boot overrides the normal it seems...
I still have so many other problems with the board besides this that it's awful.
I literally spent 8 hours just trying to get RAID drivers to load and XP on RAID, and when that didn't work, trying to XP on a single drive.
I know my hardware is good, except for the mobo. Everything was moved over from another build and put on this board, now I can't use any of the parts.
If I cannot return the board, i'm going to sell it and advertise the board as very unstable at best...though I've had the board since the 19, Sunday was the first day it had a CPU installed.
Looks like i'm going back to Asus...
BL
So.. my board arrived. Just waiting for the DDR3 now... meh...
http://hush.user.fi/pics/this3.JPG
http://hush.user.fi/pics/this2.JPG
Yikes, I'm staying clear of this board for now. Wasn't there supposed to be a V2 of this board coming out next month? Maybe it's because this version is a lemon? My XBX2 has been a stellar performer so it's odd that this board has so many bugs and issues coming from Intel.
Funny, I don't think this board is any less stable than the XBX2 was on its earlier BIOS releases.
Where I'm at now...
I think this is going to be my every-day config.PHP Code:
vCORE: BIOS: 1.375
vMCH: BIOS: 1.350
vFSB: BIOS: 1.300
FSB: 412MHz Processor: 3.7GHz
Reference/Memory Freq: 333/1333
Timings: 8-8-8-24
vDIMM: BIOS: 1.98
24 Hours Orthos stable
I'll stretch it more this week.
I do not remember....I didn't write it down.
I had other problems with the mobo not taking install to XP with two different XP discs; even before I could get to XP setup. One of the discs were sealed and used for the first time on this board.
Again, all hardware was from a previous build that worked fine.
Another note that I forgot to make, the speaker on the mobo does not work at all either. I never ever get a POST beep before the boad allows me into the BIOS....weird.
I just put in an RMA from the place i purchased the mobo, they are giving me a refund minus shipping cost.
I'm happy now.
I'm going witht he P5E3.
BL
hey gtj could you provide us some screenies too.. superPi runs etc.
how many people have this board and what is the consensus compared to the other ddr3 x38 boards out currently. i'm making the move from my formula se to a ddr3 board and am torn between the dx38bt and the maximus extreme.
Well,
I switched to the DX38BT after having the Maximus Formula. There was something wrong with the Asus board. It ate 6GB of Crucial Ballistix, one by one - no overclocking. Then for the grand finale, mysteriously one of my 3 month old 150GB Raptors died. All the memory is out for RMA now.
I got to say that I am damn pleased with the Intel board - but I'm not after any major overclocking, even though this boards bios appears to have the capability.
yes, there will be a BT2 version with an X48 quality bearlake chip it in. Just about everything else about the board will be the same, possibly with a few bios tweaks to take advantage of latency improvements that can be squeezed out of the higher binned northbridge. Depending on the industry acceptance of DFI like tweakability the BT2 might also include a few more of the performance tweaks available on their P35 board.
ok, i've got the board and 4gb of g.skill pc3-1333 on the way. hopefully this thing is as bulletproof as the badaxe2 was, i'm only going to clock the cpu up to about 333fsb and run the ram at default.
A rather uninspiring SuperPI 1M... Should be below 14 at least.
vista...
All, quick hint on easy BIOS upgrades for Intel boards. If you're having issues with not getting an OS install and you don't want to make a bootable USB thumbdrive . . . don't. :)
All you have to do is download the DOS image from their site, uncompress it to somewhere. Next put the .BIO file that gets created and throw it on a thumbdrive. After that remove the BIOS Config jumper on the board, plug in the thumbdrive and wait. Then wait more. Sooner or later (like up to 3 minutes or so) you'll see activity on the thumbdrive. After even more waiting you'll see some feedback on the monitor about the update. Most thumbdrives are supported (though not all). Hope that saves people some time at some point, and removes the need to install an old OS just to update the BIOS. :)
I swapped out the E6850 for a Q6600/G0 to see what the differences were...
E6850
http://www.peakin.com/x38/bert-dx38bt-e6850-g0.jpg
Q6600
http://www.peakin.com/x38/bert-dx38bt-q6600-g0.jpg
Notes:
The SuperPI times are a little better with the Q because of the 2 extra cores. If I set SuperPI so it uses only 2 cores, the times are within a few msec of the E.
The Q needed a whopping 1.375 volts to be stable at 400. I think it's still got some headroom.
Some of the cells are blank because I forgot to run the tests. :)
I'm surprised at how well the OCZ memory is doing. It's rated DDR3-1600 8-8-8-24 at 2.00v. The tests on the E have the vDIMM set to 2.02 which I later found is more than I needed.
The board is very forgiving with it's FSB. Plenty of times I couldn't boot an OS but could still get to the BIOS to change it.
The board is NOT very forgiving with memory timings. One slip and you're back to maintenance mode. In fact, I think there's another BIOS bug... When I was trying to go from 7-7-7-20 to 6-6-6-18 @ 333/1333 on the Q, the board wouldn't boot. So I went into maintenance mode and set the timings back to 7-7-7-20 but the board still wouldn't boot. Finally I had to set 333/1066, 7-7-7-20, let the board boot, then go back to 333/1333 to get back to where I was.
The board, memory and G0 cpus are very stingy with power. Several times I found myself LOWERING the voltage to be stable. Big change from the older conroe's and DDR2 memory.
GTJPI is a PI program I wrote that cross-compiles on both Windows and Linux. I use it because I know EXACTLY what it does and I can compare apples to apples on both operating systems.
Conclusion:
At 400FSB / 3.6GHz, the Q6600 G0 is a better deal than the E6850. 2 extra cores, same core-for-core performance, same price. We'll see how it goes as the FSBs get higher.
thank you gtj for that memory calculator etc, has been a great help.
ive been half the day in maintanance mode. (which is a bit** compared to other bios recovery solutions ive seen)
Very impress. I also got a DX38BT with Super Talent DDR3-1600 memory. I will copy your setting. You save a lot of time for me. Thanks a lot.