There is actually another product brand that goes by the name Anus http://anuslaptops.com/ though you might want to read between the lines before trying to buy anything. ;)
Printable View
There is actually another product brand that goes by the name Anus http://anuslaptops.com/ though you might want to read between the lines before trying to buy anything. ;)
Thats hilarious...........now I can safely say I have seen it all :ROTF:Quote:
Originally Posted by D_A
Good afternoon!
Advise, what now is the freshest version BIOS for this motherboard and where it to take?
I have two beta version BIOS 501 (downloaded from links from this forum)
In one control of speed of operation of fans doesn't work and POST occupies 2 minutes. In another control of fans works, but periodically BIOS is dropped on default. In 401 too most.
Can there is more stable release BIOS?
knock on wood guys, my Z7S is still kicking.
I've had it over a year and half now.
No issues.
Stock clocks, never overclocked. Thermalright cooled on the CPU's and chipset.
Kingston HyperX memory.
So far so good!
what have you guys moved on to?
I think I'm getting out of the xeon stuff, next pc is an X58 with i7 for me.
Although I got it jacked up on 24gb of ram. :D
Went to a Supermicro X7DAL-E.. also have a UD5 running and an SR-2 just waiting for a case (ordered, shipping in a few days).
Almost a year after my last post (which is just on page 68, lol) and this thread is still kicking. My Z7S is still alive, although yesterday it did just about give me a heart attack.
I leave mine running all the time, and when I came back from watching TV to "do some work," the PC was on, but it wasn't alive. The fans were spinning, the power light was on, but the monitor was still in standby and the keyboard and mouse were not lit. This happens quite frequently, actually, so I did not panic. Something with about either how the computer goes to sleep or how it wakes up. Anyway, I usually reset it and all is well. But this time, I reset it, and... it did not POST. It went right back to the state it was in before. Lights on, fans spinning, but no beep, and no peripherals powering up.
Very first thing I did was browse to Newegg on my phone to make sure I was still under warranty, or I would have really shat a brick. Warranty is up in April of '11. Whew. So I get my flashlight and start looking for fried VRMs. Don't smell anything or see any signs of fried components, but that does not convince me. So I remove the TV tuner card blocking my view of the GP diagnostic card and boot it up once more, expecting to see the dreaded FF. Instead, it shows E4. Crappy manual doesn't list the meaning of the error codes, only the codes that it displays when everything is working properly. So today I call ASUS, and the guy tells me it's a bad memory code.
I remove all my sticks, power it up, and this time I get beeps--long, short short--along with the E4 on the card. So I put them in one at a time, leaving the stick that had occupied channel 2 for last. I figured it was the most likely to have failed, being the only stick sandwiched directly between two more sticks. And I was right. It's a stick of Kingston HyperX that I purchased from a member here nearly a year ago. Looks like I'll be testing Kingston's lifetime warranty service.
Anyway, after reflecting on the whole episode, I am actually disappointed that it was not a VRM failure. The closer I get to the warranty expiration date, the more concern I have that it will fail after the warranty expires rather than before. I'm considering calling ASUS and asking if they will just replace my Z7S with a DSEB-DG even though mine hasn't failed yet. Surely by now they must know that these things are ticking time bombs.
evga SR-2, what else :DQuote:
Originally Posted by pbdm4k
My Z7S is still ticking away, albeit as my brother's main rig - pretty reliable, and no issues 2 years down the track.
I have switched to a SuperMicro X7DWA-N though, doing a great job - yeah, it's not the latest tech, but I cannot justify getting new SR2 - and it won't fit in my Blackbird case - so I'll stick with the SM for a while longer.
Good day to all owners of this "miracle" Asus Z7S-WS!
Discovered a problem with this motherboard. Posted in Asus technical support to the following letter:
My question comes the answer:Quote:
The problem with increasing frequency of the second processor.
Intel QX9775 processors have a free multiplier. Default multiplier for them is x8.
Accordingly, if you change the multiplier to x8, 5 both processors are working
at a frequency of 3400Mhz. And even at increasing the multiplier to x9 crashes - the first processor actually changes the frequency at 3600Mhz, and the second is reset to 3200Mhz, ie remains in the native factor x8.
Same thing happens for the factors x9, 5, and x10. Higher factors are not checked.
Ie it looks like this:
Factor, Real Clock CPU1, Real Clock CPU2
Auto 3200Mhz 3200Mhz
x8 3200Mhz 3200Mhz
x8, 5 3400Mhz 3400Mhz
x9 3600Mhz 3200Mhz
x9, 5 3800Mhz 3200Mhz
x10 4000Mhz 3200Mhz
In this case, the motherboard BIOS at boot time reports about the frequency, which should be a test utility CPU-Z, Coretemp and HWinfo - the frequency of the second processor is lower than the first.
And what do I do?Quote:
************************************************** ******************************
Dear user,
Due to company support policy, we don’t support overclocking issue.
I’m sorry I can’t anything for this.
Best regards,
Ryan Hung
Global Technical Support Department
ASUSTeK Computer Inc.
************************************************** ******************************
Technical support does not want to see the problem.
The problem is not related to the consequences of overclocking. The problem with that, to perform overclocking! The motherboard does not provide the documented functionality.
Dear owners!
I have two more requests:
- If you do overclock, check your system (especially for someone installed QX9775).
- Help to steer the conversation with technical support in the right direction. I, unfortunately, not very familiar with the English language. What and how they write that the case has moved?
Asus = Trash.
edit
For those concerned, the Asus rep that was on the forums and was supposedly investigating the sudden death syndrome hasnt said anything for a few months now.. so I guess we've been left in the cold, again.
Hey guys, just lost my Z7S yesterday - bought in June of 2008. Happened in the exact same way as the others, same heatsink area, and yeah, tons of smoke, had to literally evacuate the room.
I have an email into ASUS for the warranty replacement. What are they substituting the Z7S with? I need something hopefully better to throw my 5410's in.
Here she is guys...
Looks like the Magic Smoke (TM) got out. :shakes:
STEvil - 2
angelshonnny - 1
Jaco/taymans - 3
jcool - 6+
spiro - 1
4DoorGTZ - 1
DanTheBanjoman - 1
Serlv - 1
ZIO -2
colombodk - 1
romanzen - 1
lombardi - 1
Apoptosis - 2
neveroffoline - 1
24
Mine is still kicking, with the warranty up in April. Really starting to panic, thinking the warranty will be up before the apparently inevitable will occur. But I'm concerned... How much of a fire hazard are these things when they fail? Will the whole thing catch fire if I'm not around to cut power when it fails?
Also I'd like to know what Asus did for ZIO, if you're still watching the thread.
shouldnt be a bad fire hazard unless you have a lot of wires right next to the spot that burns up.
Oh and hey, anyone have any more Kingston HyperX FB-DIMMs they want to unload? I can use two more. Also if anyone purchased the FB-DIMM water cooling kit and no longer has a need for it... let's talk.
667 or 800? 1gb or 2gb?
HyperX FB-DIMMs were only made available as a 2 x 1GB 800MHz kit.
funny, i've got 4 similar to these, the non-apple version I guess.. http://www.ec.kingston.com/ecom/conf...KTA-XE667K2/2G
Those aren't the Hyper-X though.
hey ya,
just here to confirm another dead Z7S. :)
as you may guess VRM's burnt (again)
case (lian li b-25) mb tray after fireworks
some more funny things.
asus SERVER support only reachable from monday to friday in regular office times. (board went bbq on saturday morning)
phoned them monday morning 09:00. stuff was done quickly since it wasn't the first replacement. ensured that the board will have latest bios update. they failed there at the first replacement. which then needed a 2nd board to be sent with an actual bios.
replacement board arrival ETA by the support was tuesday.
On thursday with no board arrived or any further information from the server support i called them back to get any information about this.
me: hey whats up board should have been arrived but i have none yet.
happy supporter guy: let me check (busy typing in the background) oh yeah, i got it. we have none of these boards in stock. and there is no delivery date.
i should have informed you about that. but i didn't, my apologies for that.
i will get the delivery date for the mb and call you back in 30 minutes.
guess what?
45 minutes later there was no call.
i went to an appointment then, doing dmg control. cause these board in combination with the extra great support is making my customers very mad. cause i can't hold my schedules without my main workstation.
thats now 24h ago.
guess what? no call, no email, no nothing yet.
i got a question now to you guys.
i am trying to find an alternative board.
is there any tyan board not needing ecc rams ? aka is it possible to switch to tyan (or some other mb) without having to switch the fb-dimms which was needed by the z7s.
best regards,
romanzen
ASUS USA does not have any Z7S's at all. They replaced my Z7S the other month with a DSEB-DG and I was able to use the same ram etc. It is a much more stable board and better quality. You will need a SSI EEB 3.61 chassis support. I am using an Aerocool bgs 500.
If they have no Z7S boards in stock, then I would demand that they replace it with a DSEB-DG. Won't fit in your case, but a new case would be cheaper than new RAM.
Here is a link to the case I am using with my DSEB-DG:
http://www.amazon.com/AeroCool-PGS-B...9256947&sr=8-1
There's supposed to be a new revision of the Z7S now.. not that it matters anymore. At least thats according to Gary Key.
STEvil - 2
angelshonnny - 1
Jaco/taymans - 3
jcool - 6+
spiro - 1
4DoorGTZ - 1
DanTheBanjoman - 1
Serlv - 1
ZIO -2
colombodk - 1
romanzen - 1
lombardi - 1
Apoptosis - 2
neveroffline - 2
Romanzen - 1
26+
Stevil, any idea what the new revision number is and if it addresses the VRM issue? Mine is rev. 1.03g.Quote:
Originally Posted by STEvil
ST, put me down for 2, that Z7S was my second.
I dealt with ASUS USA most of Jan into early Feb. Waited like 5 business days and called. Next day I got an email offering me another MB due to the fact it is discontinued and they have NO supply. I negotiated the DSEB-DG over the other model they offered. I was told she could "try and locate one" (Z7S) in another country, but no guarantee and I didn't want to wait.
So there you have it folks, the Z7S is officially out in the ASUS USA warehouse (and probably elsewhere considering we are a huge market). I'm hoping this DSEB-DG at least lasts 2-3 yrs.
hey ya,
thanks for sharing your experience guys!
got a new DSEB-DG now here from the support in NL. called again monday afternoon. it arrived now not even 24h later. as new case, best option (from my point of view) i found was an zalman gs 1000. the case does not support formfactor SSI EEB 3.61 but still fits extended atx.
decision was made cause:
- can re-use all the 120mm pwm fans i used in the old case.
- top side case fan's to get the heat of the cpu coolers out.
- price around 120€, not 250€+ like silverstone stuff for example
- no open side panel (that noise).
do you have some more suggestions on cases ?
The Zalman looks like it should be a good case. It has a very similar layout to what I am using, a Cooler Master Cosmos 1000. The Cosmos has sound dampening foam on the side panels, no forward-facing vents, and has a door to help block more sound. The intake vents are also filtered. It is still not as quiet as I would like, however, because of the FB-DIMMs. The rear 120mm fan is plugged into the Z7S's memory PWM header and constantly runs at full speed. That is the loudest part of the case. I suspect the Zalman would have similar sound characteristics, maybe a tiny bit louder because of the lack of door plus forward facing vents.
I posted a Aerocool PGS 500, and it has the proper standoffs. I paid less than $120 shipped. Link is above as well as a SS. Can you not get the aerocool there?
Fixed.
Not sure what the rev of the new one is.
That's weird, it looks like DDR-2 800 ECC fully buffered RAM is what the MAC pro uses, I think that is the same for the Z7S? It even says for 8 core harpertown servers.... hmm what do you guys think?
http://www.newegg.com/Product/Produc...-019-_-Product
although I should at least get crucial RAM from the reports.
Zio
Mac Pro RAM will work on the Z7S, but the heatsinks will not. The Mac Pro uses riser cards for its RAM with plenty of space between the slots.
Still looking to buy some Kingston HyperX FB-DIMMs, if anyone has any. Part number KHX6400F2LLK2/2G.
The 3-year warranty on my Z7S ended last month, so I'm flying without a parachute now. Scary. I emailed ASUS a while back and asked them to pretty please just exchange my Z7S for a DSEB-DG even though it wasn't technically broken yet, but they wouldn't go for it. Going to be upset if it doesn't last another 3 years.
fun with this build... gtx 460 2 win...EVGA/Nvidia. Gotta say I've not seen Crysis 2 this smooth on my machine for 400? beats out most of the other offerings out there, SLI on a stick. hella smooth with 672 CUDA cores, 512 bit.
uh nm, it failed after about a month, well it got choppy after an hour or so, is there an ASUS replacement for the z7s-ws yet? Maybe with Sli?Quote:
Originally Posted by ZIO
supposedly there's a new revision of the board, but whats the point this late in the game?Quote:
Originally Posted by ZIO
I have a smile.
:)
Now to hope is stays there.
Tagchen Allerseits! Lurking here for some years now, and since this thread has gone off topic and chatty I hope you don't mind if I bring up another off topic issue which is related to the Asus Z7S WS: There seems to be a linux kernel bug almost exclusivly to the Z7S WS on Kernel Version 2.6.38 and newer. Here is the report entry for Ubuntu Bug tracker.
https://bugs.launchpad.net/ubuntu/+s...ux/+bug/801840
Since it is related to the kernel, it seems to affect all kinds of distros. Tested Ubuntu, SuSE and Fedora and all show Kernel Panics on my machine as described in the report or already fail during installation. This Bug is a kind of a showstopper for several months now, i first run into it when ubuntu 11.04 was released. If anyone has knowlodge of a kernel version newer then 2.6.38 that works stable on the Z7S WS please post.
If you bother feel free to post on the bug tracker of the distro of your choice, so far the kernel developers seem to be indifferent to this issue because it looks like there are only very few linux users with a Z7S WS.
DSEB-DG seems to be unaffected by this issue. I own booth, the DSEB-DG and the Z7S-WS, but I can't use the DSEB-DG with linux because I don't know how to boot it from a discrete graphics card, never got a signal although the adapter works fine with all other boards. Tested both PCI-E 16 Slots on the DSEB-DG, but it seems like linux always boots the onboard graphics on the motherboard. Looked in the bios for an option to disable on board graphics but found none so far and there is also no mainboard jumper for configuration as far as i can tell. Can anyone of those who have switched to a DSEB-DG tell me if Windows is able to boot from a discrete graphics card with that board?
The board has many many unresolved bugs and bios issues that never were, and never will be fixed. For me, it's water under the bridge... I guess most people have moved on as well. Although I certainly won't forget the epic fail Asus delivered with this so-called workstation product.
Well put and many thanks for putting this warning out there. I was almost ready to bid on the Asus Z7S system that's up on ebay.de at the moment but thankfully came across this thread.Quote:
Originally Posted by jcool
Would anyone like to offer suggestions for stable Intel Seaburg 5400 based LGA771 workstation board alternatives? I've currently short-listed the Supermicro X7DWA-N and the Tyan i5400XT. Which company/board is better for a stable workstation?
Both are fine boards really. I'd say, whichever you can get cheaper. The Asus DSEB series of boards is also a valid choice (built & serviced by the Asus server divison which is completely different from the "workstation" line of products)
Hi guys. Im looking at options for an upgrade atm i have an old server sitting here doing bugger all and it has two x5355's im thinking of going down the Z7S WS route but i see there are quite a few problems. Im dont think the xeons will perform as well in gaming terms as the i5 750 i have atm but it seems a shame to have those xeons doing nothing. In an ideal world i would like to run them with 2x crossfire 5850's. Is this an option still or am i completly nuts? :D
Nuts..UNLESS your electricity is free walk away from them and realistically, toss them away.Quote:
Originally Posted by Icy
Just too old now to be viable.
I used X5350's in 2006, moved on to harpers in 2007 I think and even sold those off in 2009
To me it's production vs electrical cost and build cost.
Think on this, a 2600K has TWICE the computational power of both of those X5355's and costs 1/2 as much to build and uses less than 1/2 the electric.
I rest my case
lol just a same to not do something with them.
I still run an old dothan M740 as a media pc dont like to bin old tech.
Your avatar is making me hungry.
</end> :off:
At leasy the dothan isn't a power hog,Quote:
Originally Posted by Icy
IIRC my clovertown with dual X5350's drew 420w on Rosetta back then
Made the equiviliant of 21,000PPD
2600k will make 40,000PPD drawing under 200w and your in England where elec is at leastb TWICE what it is here
Yep, your nuts. A nice buts but still nuts!:p:
FYI: You can get pair of L5420's from ebay for 100USD if you wanna go the Z7S route.
http://www.ebay.co.uk/itm/Intel-Xeon...item35b649f448
Uncle Dave here helped me get my hands on X7DWA-N probably early 2009, and it's a dream board. You can have a look at my Blackbird log to see what I've done with it, running 2 Harpers. I simply cannot justify replacing it, what a beautiful system. Yes, I have a new Sandybridge E3-1245 setup too, but I use that for WCG et al, and other tasks, but my trusty 8-cores of MP goodness I won't easily replace.Quote:
Would anyone like to offer suggestions for stable Intel Seaburg 5400 based LGA771 workstation board alternatives? I've currently short-listed the Supermicro X7DWA-N and the Tyan i5400XT. Which company/board is better for a stable workstation?
For me it's about ROI (return on investment), the cash-outlay for a new system will take a significant time to recover, so, I can't justify - may very well be different if you run 24/7, which I don't....
I say X7DWA-N if you can get it. I have my trusty X7DAL-E still running 24/7.. but its being sold to roomie 'cause his system sucks lol.
My 3rd (and final) Z7S-WS finally died, not even a post beep, got some final mix of strange codes, and the MB still throws missing ram codes if you take them out, so it's dead. The build lasted for many years and while it was functional was the fastest computer I've ever worked with (at home), thanks J-cool and Stevil, Movieman and the rest of you maniacs for the help and direction, I'm thinking of a super-fast Ivy Bridge with at least a 7990. ASUS should have issued refunds instead of faulty board replacements. The supermicro is almost in my cart as a replacement for the Z7S....
About the only positive outcome of the whole ordeal is that the guy responsible for the issue doesnt work at Asus anymore..
Hello,
I am still here with my faithful Z7S after four years. Overclocked, underclocked, still going strong.
I've seen comments on here about the 0501 bios but cannot find it anywhere, PLEASE if someone still has it, can they repost a link or send me so I can host the file. I need the Intel VT-x support.
Many thanks
Sent you a PM... sorry it took so long.
I guess the reason why yours is still alive is the usage of L CPUs. They draw less power, thus straining the crappy voltage regulation less, thus not killing the board (as fast).
Thanks for the PM. Asus also sent me the 0501 BIOS and I compared it against yours, it's identical, but unfortunately doesn't fix the VT-x bug. I'm considering buying another motherboard just for the virtualisation support. I've seen several HP and Dell workstation boards on eBay that should be compatible - the downside is I would lose the control of the CPU performance. I had a HP workstation a few years back and their BIOS support for overclocking/underclocking is roughly zero.
The L-series Xeons were for me a natural choice, I didn't want to blow my electricity bill through the roof with the effective consumption of two computers in one. Even four years later, I don't see any Intel solution that gives 8 physical cores for 100 watts of electricity, and even less when underclocked...
Mine's still chugging along 4 years later with X-series CPUs. I do get this annoying thing however where the PC goes to sleep and then for some reason wakes up but doesn't fully wake up. Power light comes on, fans spin, keyboard lights up, but no video, mouse doesn't power on, and completely unresponsive. Have to reset. I've had that problem for a year or two now.
Oh, but there are... Dual L5630 for example. http://ark.intel.com/products/47927Quote:
Originally Posted by drewhemm
4 Cores, 8 Threads, 40W TDP :) I have one of those too (running alone tho). Or even the L5640 which is a 60W Hex :)
Mmm..L5640. Soon...
i have some of them L5630's ;)Quote:
Originally Posted by jcool
Update on my Z7S-WS:
Tried to wake it from sleep today, and panic ensued. Fans spin up, but then I hear a faint click from somewhere that should never click, and the system is unresponsive, black screen. Think to myself, "well, I've been stupidly lucky with this machine, but the inevitable has finally happened." Dollar signs start spinning in my head.
Unwilling to just accept the inevitable, I start removing components one by one in hopes that it's not actually the voltage regulators on the mobo that have failed, but a bad component. Get it down to just the PSU, mobo and CPUs, but the diag card is still showing the dreaded "FF" and no signs of life from the machine.
Remove the 2nd CPU, FF. Remove the 1st CPU, and the machine turns itself off immediately after powering on. Apparently needs a CPU (duh). Test the PSU under load with a multimeter, seems fine.
Remove the heatsink from the voltage regulators to see which one has burnt itself to a crisp, but see no damage nor smell any smoke. Hmm... Actually kind of disappointed. I wanted to confirm with my eyeballs that the mobo had indeed converted itself to a toaster so I could bin it without any doubts. Scan the entire board, no scorching to be found. A bit frustrating.
Go for the old, "well maybe if I put it back together after having taken it all apart, it will just work."
Start putting things back in one by one. Put a CPU back in the 1st socket, but without a heatsink or fan. Turn it on, card still shows FF. Immediate dead end, I think. But I hear a very faint and steady ticking noise coming from the voltage regulators. Could not hear it earlier probably because of the fans. A little bit more satisfied, as I'm pretty sure a properly functioning voltage regulator does not tick. At least now I'm pretty sure it's toast, even if it didn't do any fireworks.
But still, I have not ruled out every possibility. One of the CPUs might have gone bad. Need to swap them, etc.
So I install the CPUs in reverse order from where they were originally. No go. Try each one separately in the 1st socket (socket 0), FF both times. Then, on a whim, because I'm sure it won't work, I put a CPU in only the 2nd socket. Surely you must have a CPU in the first socket in order to boot the PC... Wrong! Diag card goes through all its steps and shows 00, which means it's handed off to the OS. Moreover, I hear no more ticking from the voltage regulators.
Overly optimistic, I put PC entirely back together, this time with only a CPU in the 2nd socket. Cross fingers, press power button, and well, I'm on the thing right now typing this. Was surprised the mobo would boot with only a CPU in the 2nd socket and nothing in the first, but what a pleasant surprise. Back up and running without having to spend any money. Sure, I'm down a processor, but oh well. Still have a quad-core Xeon working for me.
What, me, lucky?
Anyone else try booting their dead Z7S off only the 2nd CPU?
Also, anyone wanting to get rid of their DSEB-DG? I'm sure it's only a matter of time before the regulators for the 2nd CPU eat it.
you had a cpu without a heatsink in the socket and fired up the hole thing with your ear over the vrms (and your head in close proximity to the cpu)? thats brave man! :D
anyways, nice read. thanks. i almost forgot about my z7s... ;)
yeah i would have put the heatsink but left the fan unplugged!
It was fine. Only had it on for a few seconds to see if it would POST.
In other news, I've modded the 0401 BIOS to include Intel's latest version of the ESB2 OROM, not that anyone cares. Pulled it from Intel's Skulltrail BIOS. Trying to get an SSD to work properly but don't have very high hopes. Sometimes it's recognized, most times it isn't. Might move it to the eSATA JMicron controller instead to see if it does any better there.
Hey guys, one more update. I'm building a new SFF PC to replace my handicapped Z7S and will be parting out the working bits of the Z7S machine. I'll be putting most of it on eBay, but there is one thing that I think might be a bit too "extreme" to gain any interest on eBay.
You may or may not recall that back in 2009 I bought some Kingston Hyper-X FB-DIMMs from Movieman for this machine. Well, the FB-DIMMs did not last. Each and every one of them died. Kingston replaced one set under warranty, but when it was time to send back the other set, they had none to offer in replacement so they gave me a refund. When my last set died, rather than sending them back to Kingston I kept them and scavenged the excellent heatsinks on them to use on some other RAM.
Right now the Kingston Hyper-X FB-DIMM heatsinks are on 2x 4GB sticks of Kingston KTA-MP800K2/8G FB-DIMM RAM. They are working very well--I don't think there was ever a better heatsink solution for FB-DIMM RAM available.
Anyway, if anyone is interested in either the heatsinks or the RAM + heatsinks, shoot me a PM with an offer. Unfortunately I only have two of the heatsinks available since Kingston kept the others. I begged them to send the non-working RAM back just so I could have the heatsinks but they would not do it.
BTW in case you are unfamiliar with the Hyper-X FB-DIMM heatsink, this is what it looks like:
http://images10.newegg.com/NeweggIma...104-021-04.jpg
Heatpipes are in contact with the raised section for the AMB chip, which is the part that generates the insane amount of heat.
This is the RAM the heatsinks are currently installed on: http://www.newegg.com/Product/Produc...82E16820134659
Hi all,
Can someone here please confirm whether this board despite the latest BIOS (0401 or 0501) still doesn't allow changing the RAM timings from within the BIOS? I'm just trying to figure out if there's any point using Kingston HyperX's that do 4-4-4-12 at 800MHz Vs the 5-5-5-15 plain vanilla PC2-6400 FB-DIMM RAM out there?
Thanks.
say what? might have to try this. this friggin' board.... nice discovery!
Scott sighting! :eek:
I haven't played with 0501 but I don't believe it adds anything to do with RAM timing.Quote:
Originally Posted by GFORCE100
I can tell you however that if you use the Kingston Hyper-X FB-DIMMs, they will die. At one point I had 4 sticks in my machine--all are now dead. So if you do buy Kingston Hyper-X FB-DIMMs, buy them from some place that sells them as new and gives you a receipt. Kingston will only refund the original purchaser of the RAM. They no longer have spare sticks for replacement.
Right now I am using 2x 4GB sticks of Kingston's KTA-MP800K2/8G, and I have put Hyper-X heatsinks on them for improved cooling. Works great.
Honestly I wouldn't spend too much--scratch that, any money on a Z7S build, because, well, read this thread.
Thanks for replying.Quote:
Originally Posted by Syndil
Yes, I hear these Elpida IC's in the HyperX FB-DIMM's are the culprits. I do have four of them but was thinking I could use their heatsinks on other FB-DIMM's. The only issue being I see that their notches are somewhat different to other FB-DIMM's so while I can take them off the HyperX's my thought process suggests I'll have some difficulties attaching them tightly to other FB-DIMM's. I could use some thermal epoxy but then that might be messy. Still thinking about it.
As for the Z7S VRM. I've read a lot about it and it seems using L series CPU's (i.e. L5410, L5420, L5430) is a way to go. Whether it's a design issue that could still stand but temperature to me seems to be the #1 reason for why it burns up. I could be wrong but there's no fire without first high temperature hence my logic. The thermal putty Asus uses is poor as are the plastic pegs that hold the VRM heatsink in place. If I do install the board I would first redo all that and go for a metal bolt-thru + washer approach, perhaps even think about getting a custom copper shim under there too and artic silver 5 thermal paste.
Any reason you settled on using only two DIMM's? Are you happy with sacrificing the extra memory bandwidth from employing the full quad channel architecture?
I didn't have any problems putting the heatsinks on different RAM, but I did use Kingston RAM, so that might have something to do with it. I didn't try any other RAM to see if it would fit.
The reason I am only running two sticks now is because I only had two Hyper-X heatsinks available. I had tried modding some regular RAM heatsinks to fit FB-DIMMs in the past, and it failed miserably. Temps would regularly top 100 C. Since I refuse to have a machine that sounds like a turbine, that's when I went to the Hyper-X RAM, until it died. Temps were fine with the Hyper-X, and they're fine using the Hyper-X heatsinks on the Kingston RAM I am using now. If I had two more Hyper-X heatsinks I might run two more sticks as long as the temps stayed reasonable. But without them there is little chance of that.
Anyway once I am done building my SFF PC replacement, if you want the 2x 4GB sticks of Kingston KTA-MP800K2/8G FB-DIMM RAM with Hyper-X heatsinks on them, let me know. The SFF build is taking more time than I had planned, as I am having to shorten just about every cable I put in it.
A fan is much more effective than ramsinks.
Yes, but I am a fan of silent cooling. I'll take good enough but quiet over more effective but noisy any day.
Thanks for your offer though I already also have 8GB of Micron PC2-6400 FB-DIMM's for quad channel mode. I have a OCZ XTC memory dual fan cooler that's quiet should I decide to leave the HyperX heatsinks as-is (i.e. not transfer them over to the Micron's).Quote:
Originally Posted by Syndil
If it's not a bother may I ask what CPU's you've been running your Kingston PC2-6400 RAM with? I hear this board is quirky with 400MHz (1600MHz effective) FSB CPU's and was just wondering whether you were running PC2-6400 RAM at 800MHz with 1333MHz or 1600MHz FSB CPU's?
Thanks again.
More effective in this case means it does something other than insulate the ram. Ramsinks are generally worse than nothing..Quote:
Originally Posted by Syndil
X5482, 3.20GHz, 1600MHz FSB. But the Kingston KTA-MP800K2/8G I am running now runs much cooler than the Hyper-X did.Quote:
Originally Posted by GFORCE100
Generally, I would agree. But in this case specifically (referring to FB-DIMMs and Hyper-X ramsinks), I definitely disagree. They work as intended, and I wouldn't run my FB-DIMMs without them unless I did have a RAM cooler--which I don't and don't need thanks to the Hyper-X ramsinks.Quote:
Originally Posted by STEvil
Wow, so you've been running a 150W TDP CPU x2 in a Z7S WS and your VRM is yet to blow. From everything I've read that's a miracle. I guess you must have a fan blowing over the VRM section, are lucky to have a board sample with "good" components or your board revision isn't 1.03g. I doubt the 0501 version BIOS has any effect on the VRM.Quote:
Originally Posted by Syndil
Put heatsink on buffer chip, thats all that needs it. Ram just gets insulated and soaks up more heat from the buffer chip with normal ramsink stylesQuote:
Originally Posted by Syndil
No, read a few posts back. Mine didn't "blow," but the VRMs for CPU0 did indeed fail. Didn't burn up or anything, but one of them started ticking, and the board would not boot at all if CPU0 was present. My board is now running with only CPU1. I figure it's only a matter of time until the VRMs for CPU1 fail as well.Quote:
Originally Posted by GFORCE100
Yes, it does indeed need it, and that's exactly what the Hyper-X ramsinks do. They have a U-shaped heatpipe in contact with the AMB, taking the heat up to radiating fins. The rest of the chips on the FB-DIMM are not connected to the fins.Quote:
Originally Posted by STEvil