Looks like it is set up for 1/4" am I wrong? I hope so
I sent the email hope somebody over there can read English.
Seems the headers are in English but all the rest is in German?
I wonder.... hi or low flow
are the nozzles changeable
etc
Printable View
Looks like it is set up for 1/4" am I wrong? I hope so
I sent the email hope somebody over there can read English.
Seems the headers are in English but all the rest is in German?
I wonder.... hi or low flow
are the nozzles changeable
etc
Looks like a pretty sweet sweet spotQuote:
Originally Posted by Grinch
Any benchies to post??
Quote:
Originally Posted by owcraftsman
yes they use 1/4" threads and there are no fittings included.
I have some 3d mark 03,05,06...super pi....and will work on prime95 tonight..:up:Quote:
Originally Posted by owcraftsman
Quote:
Originally Posted by BAR*B*Q
removing battery = save ? :)
Ive had a bit of correspondenace with MIPS over the past week or so and indeed they speak great english, are very friendly and sell a great product! Their gear is not cheap, but it's handmade and makes every board it's fitted to quite unique :)
Yep they use 1/4" thread... so all the usual DD and EK barbs will fit to the MIPS blocks no problem I believe!
BARBQ: How would you freeze memory? Just unplug your MB, swap the jumper over and then back again, reboot into BIOS and check your VDIMM voltage and you're good to go!!
The DFI board requires very little CMOS resetting :)
dont forget to ask for availability. if they dont have all coolers in stock it could take some time.
you can use any common barbs. for too long threads they deliver distance rings with any cooler.
http://www.mips-computer.de/images/messingringe.jpg
hmm... i dont even want to feed my ram's with 3,04v for seconds :(Quote:
BARBQ: How would you freeze memory? Just unplug your MB, swap the jumper over and then back again, reboot into BIOS and check your VDIMM voltage and you're good to go!!
The DFI board requires very little CMOS resetting
I understand that, but if it makes you feel better I ran my G.Skill GBHZ at 3v for about 10-15mins and there was no meltdown :)Quote:
Originally Posted by BAR*B*Q
[QUOTE=hmm... i dont even want to feed my ram's with 3,04v for seconds :([/QUOTE]
As I understand it, this only happened when loading a saved BIOS profile, not at every BIOS clear.
Hi everyone I just got one dfi 680i, priming my e4300 at 3150mh@1.4vcore(bios)1.36vcore(actual):) . found 2 main problem 1) quite a serious vdrop in vdimmn, nb and core even though I using the updated bios. 2)have to clear cmos whenever oc failed to boot:confused: . Any guide here to get away from the desparated on & off and any way to fix the vdrop?
Wow lots of action in here this weekend... some thoughts:
@Borgod/MikeMK - it looks like the Enermax mystery is solved? Use the 4+4 and not the 8 pin ATX connector?
@MikeMK - If you haven't given up on the DFI... I've had most success with the 2CTEST 521 BIOS on my quad (X3220). You may want to try that. I found with the 4C version, that I couldn't change my multi and that it wasn't stable at lower voltages like the 2C version.
BTW, what OC did you get on the DFI board before you switched to the EVGA?
@Borgod... 80-Deg on the NB?!? :eek:
@Grinch... your image of your new sweet spot isn't showing up for me... where are you at now?
@Jobeo, I'm sorry to hear your supcomm story... what's the latest? Have you found anything in particular that is causing the instability with that program?
There is definitely an undervolt between what you set and what you get in BIOS for CPU VID and NB Voltage... just dial in some extra to compensate. There's more info on my blog (via the link in my sig). VDIMM is similar but not as bad... what exactly is the gap you are observing?Quote:
Originally Posted by denden
The actual droop under load of these values is not bad. CPU droop is much better than most and NB and VDIMM is insignificant.
When you can't boot due to your OC, use the power/reset button trick as described in the manual to quickly clear CMOS. Keep your baseline settings in a "CMOS Reloaded" profile and you can quickly get back to where you were in no time.
Yes 4+4 only with the Galaxy on quadcore. I have been using that since day dot with my DFI since the meltdown of my aforementioned QuadGT.
Yep my NB temps hover around 80degrees. It's not the weight of the 80mm fan pulling it off the core as I have moved the HSF up and down and the temps only get worse - very quickly.
I'm going to watercool it with a mips kit anyway.
Over the weekend I was considering hard how to lower my CPU temps. I came up with an awesome idea of a double loop system with no less than 2x PA120.3 for the CPU and a PA120.3 and PA120.2 for the SLI 880GTX, NB, PWM and SB.
I devised a superb unobtrusive mounting method that would of made it the ultimate water cool system for myself.
But then I did the costs and realised how quickly it added up.
Now I am considering using my existing PA120.2 and PA160 to cool the SLI, NB, PWM * SB and phase change the CPU. kayl lives a 15min drive away from me, it works out cheaper, will be a lot less work and I'll be able to crack 4000Mhz with ease :)
yeah I know to restore the last bootable setting but only feed up with moving the jumper, I will get bk on the manual on the power/reset button, thks a lotQuote:
Originally Posted by virtualrain
That power/reset button trick is a life saver for sure.Quote:
Originally Posted by denden
As for the voltages, I think you mean there is a significant undervolt on the CPU/NB... and while there is a bit of droop (as expected on the CPU) there is no droop on the NB right?
Undervolt = gap between what is set in BIOS and what you get
droop = difference between idle and load voltage
Phase... :D Isn't that going to be noisy though?Quote:
Originally Posted by Borgod
i haven't messed with it today but i'm gonna try some stuff right now. someone recommended 3dmark a few pages ago and i have several versions of that i will try.Quote:
Originally Posted by virtualrain
in supcom the instability seems related to heavy load. somehow the game is pushing the components harder than prime testing... i blame it all on the 400 FSB OC settings i'm using. something is off and i'm just not sure what it is. with those settings i'm barely overclocking the CPU itself by using a lower multi.
from what i've seen the VTT/GTL's are seriously important for 400 FSB. at stock FSB i could raise the multi to 11-12-13 without tweaking the VTT or GTL's...
i'll see what i can do and hopefully report back something good =/
oh...I not reliased the difference of load and idle voltage and will check it tonight, but there is definiately an undervolt, does it matter?Quote:
Originally Posted by virtualrain
well i got somewhere... installed 3dmark03 and loaded my sketchy settings... the reboot crash occured during the benchmark, which actually made me happy cause now i had a better way of testing it.
i tweaked the NB volts higher that didn't help. even tried running the test non-SLI and it seemed to reboot even faster. all of the reboots were occuring at different times during the test there seemed to be no pattern.
so i started fiddling with the GTL's ... i had them set at the "sweet spot" of 1.44v and 170 95 85... tried them a bit higher and the test seemed to run a bit longer before rebooting. finally ended up with GTL's of 180 100 90 and 3dmark03 passed for the first time ! ! i ran the test 3 more times and it passed each time. then ran the 3dmark05 and 3dmark06 tests and both passed. loaded an old supcom save with a lot of units and that seemed to go OK too...
SO.... to make a long story short... i mean, i knew the GTL's were important cause i've mentioned them in almost every post these past couple days.. but MAN they are _really_ important. unless i'm mistaken we're talking about tweaking millivolt settings yet it seems to have a huge impact on stability.
so i think i've found a new sweet spot at 1.44 and 180 100 90. which doesn't really add up cause the Nvidia GTL reference table says the VTT should be 1.55 at those GTL's but whatever works i'll take it... i'll probably need to test further with another prime blending session but for now things are looking good.
Even with a dual core the GTL's make all the difference not for the CPU's sake but for the NB.Quote:
Originally Posted by Jobeo
I am still working on getting 3.7GHz running 8 hours + Orthos stable & have increased my reboot/crash time from 3 to 6 hours tweaking GTL's.
I think GTL's are going to take forever to workout & understand as even the smallest variant in CPU/NB puts you back to the start again with the GTL's.
When they first start up they're a bit noisy - sounds a little like a fridge. However when they're running the noisiest part is usually the fans that cools the condensor, and the bigger compressors may be a little noisy aswell.Quote:
Originally Posted by virtualrain
I'll get mine custom made so it won't be too noisy of course.
It really does make sense as the reference table is just that, a reference. Raju more or less stated this several pages back in this thread.Quote:
Originally Posted by Jobeo
Quote:
Originally Posted by raju
try this:Quote:
Originally Posted by virtualrain
http://s2.supload.com/free/500x7_1.425v.JPG/view/
That's great. Frankly I'm totally amazed you got it prime stable without tweaking GTL.Quote:
Originally Posted by Jobeo
Nice OC!Quote:
Originally Posted by Grinch
thank you...
owcraftsman here are some 3d mark benchs:
3d mark05
http://s2.supload.com/thumbs/default/3d05.JPG
3d mark06
http://s2.supload.com/thumbs/default...0618122723.jpg
not the highest in the world..but decent for 1 8800gtx I guess...
tx... from my experience, and this is all conjecture, the GTL's seem tied to the FSB only. when you fiddled with the FSB the gtl's have to be tweaked to support it. i think thats why my tests with 11-12-13 went OK cause i wasn't changing the FSB at all.Quote:
Originally Posted by virtualrain
on the 400 FSB i've been talking about i was using a low 7x multi... i just bumped it to 8x and it passed all the 3dmark03 tests without any changes to the VTT/GTL's... they are still at 180 100 90
i think thats probably why those of us with quads have so much trouble above 400... the GTL's probably need to be even tighter, say within 1 pt of where they need to be, or the OC won't prove stable.
I got this board last week and so far its reasonably good in most ways and very good in others.
I have found it more stable than the BFG 680i i had before and can get rock solid stability at 450FSB.
Got my chipset water blocks today... Link mofset, Nb and SB so will let you know when all setup and running.
Quote:
Originally Posted by simonmaltby
looking forward to it:up:
For info.
I suffered some really wierd issues over the weekend with the board.
First i flashed to 521 from the official site and all was well with my X6800 (indeed it did 4.5 without issue on freon). The i changed to my E6600 and could not find multi unlocking...
So i flashed to Beta 504 and sure enough the multi unlocking came up, but other parts of the bios were poor. so i flashed back to 521. On doing this the multi unlocked remained... good news i thought... No this is when the really wierd stuff happened.
At this point i could not get stable even at stock. In windows as i moved the mouse, or accessed HD the num lock keyboard loght would flash and ther would be a stuttering. This only happened when using Raid. After a few moments the system would just freeze.
I trued battery out etc - nothing worked. In the end i flashed back to 504. then flashed back to 521 - removing the battery between each flash. I also flashed in Windows XP using Winflash and did the old DFI trick of flash once and then without re booing flash again.
Anyway that got me back to stable, but no multi unlocked.... so for me 521 does not enable downward multi's which is a pain.
hrmm... ive never heard of anything like those issues... when i flash the BIOS i just winflash it once and have never removed the battery afterwards.Quote:
Originally Posted by simonmaltby
as far as switching processors... ya i could see some trouble with that. are you sure you're removing the battery for long enough? the last time i did the BIOS reported a new CPU detected and installed the proper siprom table for it...
Are you using the 4CTST or the 2CTST version of 521. The 2CTST should have unlocked downward multis... at least it does for me.Quote:
Originally Posted by simonmaltby
using 4CTST - ah so ic the 2CTST is 2 core and the 4CTST is quad... That explains why there are so many different versions on the DFI site... wonder why i picked the 4CTST i normally just pick the top one and now the 2CTST is at the top... maybe it was not there when i downloaded originally... oh well.Quote:
Originally Posted by virtualrain
Anyway flashed to 2CTST and multi's are there again. Many thanks for the info :)
4CT bios is for XE edition cpu's including the dual core versions, for all upward locked (non XE's) multi cpu's use 2CT..
regards
Raja
would you share your voltage settings please ? im trying to get 400 x 9 primestable. theres allways a fft length that fails. :shrug:Quote:
Originally Posted by simonmaltby
or you could share yours and we could see if anything is glaringly wrong .. . .Quote:
Originally Posted by BAR*B*Q
couldn't read all the pages, 50+... :)
Where is the strap?
What is the hightest FSB you have got on air (water)?
Ok here we go, a few pics of my water cooled setup and some bios shots.
http://www.motormice.com/images/dfi_water/w2.jpg
http://www.motormice.com/images/dfi_water/w3.jpg
http://www.motormice.com/images/dfi_water/w4.jpg
http://www.motormice.com/images/dfi_water/w5.jpg
http://www.motormice.com/images/dfi_water/w6.jpg
http://www.motormice.com/images/dfi_...os_voltage.jpg
http://www.motormice.com/images/dfi_water/bios_fsb.jpg
http://www.motormice.com/images/dfi_water/bios_ram.jpg
Now, i must say, this board is so typical DFI... Its can be great, but you have to work at it.
Today I spent a great deal of time playing with the settings to get a stable clock... My aim was 3.6Ghz 400x9 prime and Vista stable.
I started by picking what i though would be ok.
That included setting HT to x3 for a 400 fsb. Oddly i found that what everr my ram settings at x3 memtest threw an error on test 3. Changing the HT to x4 gave no errors --- explain that if you can!
Anyway, the above bios setup seems to work. I did 4 cycles through memtest and an hour of Orthos (prime). I get the odd lockup in Vista but that seems to be due to a build up of temps on the northbridge... but then ambiant temps in the UK are very high atm and the issue is only suffered with the case side on, so i'm confident its not a long term issue.
Hope this helps.
Can you post a pic of Smart Guardian to see what NB temps you're getting with the mips gear?
Cheers :)
I'd like to see those temps too. Load + IdleQuote:
Originally Posted by Borgod
Is that two loops or mixing 3/8 & 1/4 inch tubing?
@simon, on dual cores for 400fsb (1600), I find only 1.48-1.51 vnb is usually required, the lock ups will then go away, ldt should be fine too @ 5x...
I will take a screen shot of Smarg Guardian later, but my finding on the temps are as follows: On air at same speeds I was getting both System and Chipset temps in Smart Guardian at about 50c Idle and 60 under load. Those have now droped by about 10c - idle at about 40c and full load i have not seen more than 52c, even with my over inflated NB volts as Raja points out.Quote:
Originally Posted by owcraftsman
My loop.. well the Koolance comes with 3/8" I have taken the 3/8" directly to the CPU then as the 3/8 leaves the CPU i have setup a parallel split using 8mm internal (10mm OT) tube. At the moment one parallel run does nothing and acts as a quick route so i can maximise CPU cooling, the other loop goes to the SB then onto NB then finally Mofset.
Raja - If i set anything less than 1.65v on the NB in the Bios then Orthos (prime) fails with a graphical screen distortion and lock up after time. The time it takes to lock is related to the NB volts i select. i.e. 1.51 gives it about 10 seconds, 1.57 gets to about 30 seconds. I do note that Smart Guardian reports quite low NB volts at 1.65 ut shows the NB as 1.48 / 1.5, so maybe thats why i need more voltage there?
My idea with the unused parallel link is that i can use as a test. Firstly when open i am stopping any major flow restriction because the 3/8 technically goes to 2x8mm or 16mm inner. If i find i needs more flow to the chipset, i can restrict, or cut the flow down that pipe. Lastly its there so i can isolate one of the chipset blocks if needed and run that pipe directly if more cooling is needed... i.e. put the NB block on its own parallel run.
So far tbh, ristricting flow seem to mak no difference at all to any of the temps.
Remember my main aim is to run as quiet as possible at 9x400. So far I can run with the temps described with the fans on the Koolance set at possition 3 out of 10 and thats running prime and in this hot weather. If i wind them up to full speed the temps drop by aprox another 6 - 10c all round.
Quote:
Originally Posted by simonmaltby
i agree with raja ... 1.65v is extreme at 400 FSB ... i'd check your other settings first before cranking it that high... i haven't seen one person need it that high for 400 FSB...
i recall a post from tony a while back where he didn't even need that much NB voltage for 450 FSB...
a lot of the high vnb requirement probably has to do with you running ddr 1100 I guess..
regards
Raja
I initially needed 1.75v on the NB for my quad OC to 400FSB :eek: ... keep in mind that there's a 0.12V undervolt on the NB setting... it's about 1.53 actual reading with a setting of 1.65 which is not that high... this is now what I'm running on the 521 bios.Quote:
Originally Posted by Jobeo
Keep in mind I'm running 4GB of RAM at 1T as well. I think it really depends on your particular proc, NB and RAM as to what voltage you will need. You need what you need... no matter how rediculous it sounds to others!
BTW Simonmaltby, nice water cooling setup! I just got my MIPS mosfet cooler. It looks very good... great attention to detail including slightly different milled planes on the bottom to account for the different height of the PWM components.
I ordered my mips kit lastnight, very nearly ordered the mips RAM freezer 4 aswell for kicks and giggles.
Quote:
Originally Posted by Jobeo
tried so many mate. none of them really prime stable.
all i can say is that im testing with an e6600 + 2x1gb d9gmh memory on 423 bios: 400 x 9 in sync mode 1T.
Spent some time on the reasons why i needed a lot of NB volts.
Firstly With the OCZ 8500 SLI memory I found that I had one dud stick in the 4 1GB modules. But even using the two good ones in Dual Channel i got lockups, These reduced with nore NB volts but were always there - Even if running the ram at stock, or under clocked at 800.
Interestingly running each stick as a single module (no dual channel) windows was perfectly stable. As soon as two sticks went in Dual Channel i got lock ups.
Anyway, i have returned the SLI for a refund and put in my trusty OCZ Rev 2 Plat 6400 at 800 4,4,4,15 and no more lockups - NB volts as sugested at 1.48 and nice and stable.
Conclusion... Either the OCZ 8500 SLI chips I had were all bad or the DFI board just does not get on with these chips.
Out of interest - in these changes I had to reset the CMOS and did it by removing the battery. In all the changes I now have a re occurence of my USB keyboard flashing num lock light and system pauses. I need to iron this out, it could be... reset cmos issue by remove battery, moved GPU from Slot1 (nearest NB) to Slot 2(next one up) or the change of RAM. At the moment my best guess is the movement of the graphics card because it seems to be grahics related.
hi guys i have s strange problem!
i can reach 425mhz stable with 1,51v on the chipset.
but if i increase the vchip the board will not boot up!
i've tried everything and nothing seems to help :(
i tried all vchipset options from 1,51 to the maximum.
everytime the same!
what's wrong with it?
also my cpu-oc seems to be very very poor! 3,2 ghz stable, everything higher than that (regardless how much v-core or which multi) will produce errors in orthos/super-pi and so on!
so dfi owners please help :D
mfg
Your not alone on this one, whenever i setup 1.54v or more on chipset itQuote:
Originally Posted by killingspreez
will not boot. then i figured that in order to use 1.54v or more i need to raise mem to 500mhz or more. I guess if you have low overclock and default mem then a low voltage chipset is enough but when raising o/c it needs more volts.
hm i'll give this mem-oc thing a shot. but i suspect that our 4*1gb ram config causes this problem. did you try 2*1gb ?
i didn't
mfg
turn the power off too your psu when goijng over 1.51 after the board stops booting, power back up and it will run fine..
regards
raja
I have not tried 2gb, im having a hard time trying to set it up in a Coolermaster 830 with a Thermalright 120 Extreme, it just wont fit if i use the 4 fans door. F*ck it, i think im going water. damn you 120 why you have to be so tall.Quote:
Originally Posted by NeoGeo333
Quick followup to my previous post about flashing USB keyboard Num lock light and system pauses. I tracked this down to the Razer Diamondback software that Vista installed via windows update... on removing this driver all is well.
No... if only i could find some Mushkin 4GB DDR2 sticks in the UK... everywhere is sold out and they seem to be unavailable :(
Im about to get a 120 extreme as a backup aircooling solution.Quote:
Originally Posted by NeoGeo333
I better pull out the tape measure :eek:
My OCZ PC2-8500 SLI play nice with my E6600 @ 3.6GHz with the DFI 680i board I think you may have just got a duff set.Quote:
Originally Posted by simonmaltby
http://s106.photobucket.com/albums/m...03600MHz%201T/
Code:Genie BIOS Settings:
Voltage Control:
CPU VID Control................ - 1.50000V
CPU VID Special Add............ - Auto
CPU Drive Strength............. - Weak
DRAM Drive Strength............ - Weak
DRAM Voltage Control........... - 2.30V
NB Voltage Control............. - 1.48V
SB Voltage Control............. - 1.52V
SB Dual Voltage................ - 1.52V
LDT Voltage Control............ - 1.22V
CPU VTT 1.2V Voltage........... - 1.21V
GTL REF Voltage Control........ - Disable
FSB & Frequency Seting:
FSB-Memory Clock Mode.......... - Linked
FSB-Memory Ratio............... - Sync Mode
FSB (QDR), MHz................. - 1600
=====
PCI Bus, Slot 1 MHz............ - 120
PCI Bus, Slot 2 MHz............ - 120
SPP <-> MCP Ref Clock, MHz..... - 300
LDT Frequency.................. - 5x
CPU BSelect.................... - BSel_7
CPU Clock Ratio Unlock......... - Enabled
CPU Clock Ratio................ - 9x
CPU Feature:
PPM Mode....................... - Native Mode
Limit CPUID MaxVal............. - Disabled
C1E Function................... - Disabled
Execute Disable Bit............ - Enabled
Virtualization Technology...... - Enabled
Core Multi-Processing.......... - Enabled
Memory Timing Setting:
Memory Timing Setting.......... - Expert
CAS Latency Control (TCL)...... - 4
RAS# to CAS# Delay (TRCD)...... - 4
Row Precharge Time (TRP)....... - 3
Min RAS# Active Time (TRAS).... - 8
Command Per Clock (CMD)........ - 1T
=====
Bank To Bank CMD (TRRD)........ - 4
Row Cycle Time (TRC)........... - 8
Write Recovery (TWR)........... - 6
Write -Read Commad............. - 9
DRAM Refresh Rate.............. - 7.8uS
I get exactly the same but I can jump over 1.54v i.e miss it out a boot at the next voltage up if I increase the CPU VTT. I just used 1.6v vtt to test this yesterday. One last thing I notice the SB voltage also effects what NB voltage will boot without any CPU VTT.Quote:
Originally Posted by killingspreez
As for the overclock try your 6700 on its native multiplier & see what happens.
with the stacker 832 i have experience mounting it both horizontally and vertically... either way i've gotten it to fit without any permanent case modifications I.E. dremeling. . . . . there are a couple tricks, PM me if ya want help.... unless the 832 side-panel is significantly different than the 830 it should work for you too...Quote:
Originally Posted by NeoGeo333
http://i168.photobucket.com/albums/u161/DaXX001/1-2.jpg
http://i168.photobucket.com/albums/u161/DaXX001/2-2.jpg
any new bios updates coming up?
I'm not sure I want a new BIOS... I'd feel compelled to tweak again and end up wasting hours doing stability testing, likely only to find that I couldn't increase performance measurably at all. ;)
What would you be hoping for in a new bios?
if they do release another update i'll wait until the stable version is released.... i'm kinda hesitant since that "load last working settings" causing the DRAM overvolt . .. .
well i fixed the v-chipset problem.
switching to 2*1gb didn't help.
the old bioses suffer from the same issue.
but if you load defaultsettings and than increase the v-chip it works.
it just doesn't work on a overclocked bios -setting
mfg
Hey guys, have you seen or experienced the 680i lt solid caps version with wifi yet? just wondering if it has any improvements over the original, the one that ate my ram? specifically regaurding the overclock. I've been playing around with the abit ip35-pro and the p5k-d, both are great boards, directly compared to the dfi 680i lt stabillity on my e6400@3.6GHz is greatly improved, memory performance is close, and raid performance is better ten-fold. unfortunately I had to sell my pair of 8600gt cards and purchase an ati 2900hd for the same/better video performance.
still, I enjoyed the DFI bios much better than these other two.
uguru is something else.. I like it.
have not seen....but I would imagine it to be about the same
ya this is something not talked about much in recent times... but i'm really let-down by the raid performance on the DFI board.... its just no good. i guess cause people who really care about raid perfmance just buy an areca card or something... but i'm tired of throwing money at this board to deal with its various problems.Quote:
Originally Posted by xTr33m sammich!
Quote:
Originally Posted by Jobeo
how are you let down? how is it no good? bench pics?
I would say this is about spot on for my 2X150 raptors in raid-0
http://s2.supload.com/thumbs/default/hdtach.JPG
I went from 2x320g seagates in RAID 0 in an Abit QuadGT ICH8R, to the DFI 680i in RAID 0 and although I have no screenshots or facts, the difference in performance was extremely noticeable.
Enough for me to now have 2x 150g Raptors sitting at home ready to go.
That is indeed excellent performance. There were/is some issues with 680i RAID on the EVGA for sure... I haven't seen many results posted with respect to DFI besides this and Raju's... and Raju's results were not looking good.Quote:
Originally Posted by Grinch
Thats really nice for this board with 2x raptors, had mine with 2x 36GB raps in R-0 scoring mid 80's, 3x 500gb raid5 samsungs on the same board were mid 90's. on the ICH9R the same 3x 500GB drives in R-5 mid 160's. but the burst is what I'm talking about, not sure how much difference it makes except I can tell when CS or HL2 or any other app is loading.
http://a293.ac-images.myspacecdn.com...e8ad077fdc.jpg
Wow that is nice whats your setup stripesize/clustersize ?Quote:
Originally Posted by Grinch
I only get about 100mb/s average read 13.? random access & 330mb/s burst from 2 x Hitachi 80gbs sata 2's in raid 0 with either a 32kb or 64kb stripe using XP's default 4kb cluster size & thought it was ok until I saw your figures.
there is a way to get more performance out of you raid arrays....I have been doing this:
http://www.xtremesystems.org/forums/...3&postcount=66
C-N I am using 16K stripe and 4K cluster
Quote:
Originally Posted by Borgod
you will love those drives for sure:up:
Ahhhhh nuts my array has taken a dump.Quote:
Originally Posted by Grinch
I changed the settings rebooted & run HD Tack to find I was down to 58 average read so I changed them back again & its still the same.
I have played with Hitachi's ftool & installed COD2 & office since the last time I run HD Tack so I guess I must have done something in ftool without knowing it & I will just have to have another play later to try to fix it.
I will keep you posted on what happens when I try your method once I work out where my speed has gone.
======
This ones for anyone (help a RAID noob)
I cant get my rig to install XP on my array if I prepare the drive with Partition magic 8 & set the cluster size to anything other than XP's default 4kb.
It will allow me to boot from the XP cd install/copy the RAID driver copy all the files to hard disk after choosing leave file system intact but when it reboots it cant find the array can anyone point me in the right direction to allow me to install XP on my array with a different cluster size.
Right I got it worked out it was the 2 x 250gb storage drives I plugged into the nvidia controllers SATA 3 & 4 sockets that slowed the whole controller down. I moved the 2 x 250's to the SIL controller & its back to normal.
I then tried your earlier suggestion & although my Average read & access stayed the same my burst has gone through the roof.
Every little helps thanks buddy :D :clap: :up:
http://i106.photobucket.com/albums/m...GBHitachis.jpg
http://i106.photobucket.com/albums/m...GBHitachis.jpg
np...ô¿ô
I forgot about atto here is what I am getting:
http://s2.supload.com/thumbs/default/atto.JPG
Glad to see the dialog go this direction. wish i was testing my 680i lt still waiting for FedX to settle the claim on the 8800GTS & Corsair Ram that got lost in shipment.
So I offer this up for comparison. I realize it's 2 different formats but some how I was expecting better performance out of the new board.
http://img300.imageshack.us/img300/6...0001gw7.th.png
There is sth wrong with the RAID controller on that mobo. I have 2x120GB SATA in port 1 and 2 (RAID0, 32kb stripe) and 2x160GB SATA II in port 3 and 4 (RAID0, 32kb stripe). I'm running Windows XP 64. The trick to turn off read caching and command queuing is decreasing my performace :( Moreover my SATA II drives are set by Windows to be 1st generation of SATA :( Please look at performance of my drives, my old 120GB drives are reaching 204.3 MB/s but these new SATA II drives only 196.8 MB/s.
I would have to say you have a stripe size of 32K or bigger in those pics?Quote:
Originally Posted by owcraftsman
Quote:
Originally Posted by AlphaHeX
Have you tried using a 16K stripe?
I will, but wat about Windows not setting these drivers to SATA II?
might just need a windows update? If I were you I would try a 16K stripe and see what you get..:up:Quote:
Originally Posted by AlphaHeX
We haven't given up yet as Grinch shows his drives are doing fine.Quote:
Originally Posted by owcraftsman
I think apart from all the tweaks like stripe & cluster size partition size etc the choice of drives would make a fair difference too. I used to get about the same if not marginally better than you on my NF4 with 2 x 250gb maxline 3 samsung's but as Grinch has shown us 2 x 150gb Raptors put then to shame on the 680i.
I cant see how the 680 performance could be too far down on the older NF4 controller I think we just need to find the groove.
I have a 100mb/s 2 disk array for £40 & if I wanted to get close to Grinch's £300 Raptors I could bung another £20 drive in there & I reckon that would be real nice & quick for the money.
It may be that the older 150mb/s drives are dragging the controller down to there speed & therefor not allowing your 300mb/s drives to run any faster (this is just a guess)Quote:
Originally Posted by AlphaHeX
I can switch from SATA 1/150mb/s & SATA 2/300mb/s using IBM/Hitachis ftool perhaps there is a tool that will allow you to enable SATA 2 on your older drives.
If you de select the "Let BIOS select transfer mode" this will un grey out the transfer mode box & may allow you to change from SATA 150 to 300.
The difference between 196 & 204mb/s on the burst is not worth worrying about imo you can get that much deviation due to the drives being fragged up a bit.
Unsellecting "Let BIOS select transfer mode" is showing only SATA 1-1.5 and PIO Mode :(
agree 100000%:up:Quote:
Originally Posted by C-N
Quote:
Originally Posted by C-N
But 196 I have with SATA II generation drives and 204 with SATA I generation drivers !
Type of drives partition size fragmentation all of these can effect the burst speed like I said imo I don't think its anything to bother about.Quote:
Originally Posted by AlphaHeX
Remember your SATA controller is running at SATA 1 therefore your SATA 2 drives cant shine.
Burst speed don't really mean too much compared to sustained/average speed if you notice the Raptors are SATA 1 & Grinches 150's make about 260mb/s burst & they are probably about the best drives out there atm so I think your 200 ish is ok in that respect.
I am absolutely no hotshot on RAID but read my earlier post about enabling SATA with older drives my Hitachis can do it Raptors can do it maybe yours can too which might allow you to run SATA 2 & get your burst speed up.
burst speed..really dont mean a whole lot...the sustained transfer rate is what is important
here's a pic using hdtune:
http://s2.supload.com/thumbs/default...0629112820.jpg
well... i really started something with my comment ... its like i threw a thread hand-grenade in here... lol. . . ..
anyways, i have no performance charts ... i am just going by the way my rig "feels" . .. . i have 2 raptor 10k's on the first sata ports and then 2 older maxtor drives on ports 3 and 4...
moving the old maxtors to the SI controller seems like an option... maybe it would help things... but i'd have to remove both my 8800 gtx's to get to the connectors... SOo o o . . .. =/
I am using these 2 sata ports for my raptors:
http://s2.supload.com/thumbs/default/big_12.jpg
The sata ports under the battery are not as fast...
hrm ya the ones you have labeled are the ones i'm using...Quote:
Originally Posted by Grinch
anyways i ran this HD tune proggie and my average is 121.7 MB/sec.....
do you think mine is lower cause i'm at a 64k stripe size ?? is there any way to decrease that without losing the drives and having to reinstall windows and such . .. .
They're the same ones I use, and performance is poor.
I'm not going to be using RAID in the future.
hrmmm ya... that may be an idea too .... the whole reason i'm using raid 0 is cause i thought it would increase performance...Quote:
Originally Posted by Borgod
maybe i should just try the drives without it... HRMmm m .. . . .
Quote:
Originally Posted by Jobeo
16K stripe would give you better numbers...you would have to break the raid array in order to create a smaller stripe
Hmm, I've installed Windows Vista to check if it is a problem with Windows XP. Under Vista the performance is even worse :( moreover I've removed SATA I drives and my SATA II performance is still at the level of SATA 150 :(
I take it the SATA 2 drives aren't limited/forced to SATA 1/150 with a jumper like on my Maxtor's check your drives.Quote:
Originally Posted by AlphaHeX