Official GIGABYTE Forum
Questions about GIGABYTE products => Motherboards with Intel processors => Topic started by: GTZaskarLE on July 31, 2011, 09:44:39 am
-
Hi, I am experiencing an annoying Vertex 3 maxiops 120GB write bottleneck with ATTO disk benchmark on my GA-Z68X-UD3P-B3 mobo having the latest F6 bios. The ssd is connected to the intel Z68 sata3_0 port.
I tried almost everything to locate this problem but it appeared to be the latest motherboard bios F6.
Switched back to F5 and it all just worked as expected.
It didn't make much difference for the performance when I used standard win7pro-64bit MS AHCI or latest Intel RST AHCI drivers.
See also:
http://www.ocztechnologyforum.com/forum/showthread.php?91864-Vertex3-Agility3-Solid3-Revo3-Firmware-ver-2.11-is-live-discuss-it-here.&p=662938&viewfull=1#post662938 (http://www.ocztechnologyforum.com/forum/showthread.php?91864-Vertex3-Agility3-Solid3-Revo3-Firmware-ver-2.11-is-live-discuss-it-here.&p=662938&viewfull=1#post662938)
I am also having this issue with the xmp profiles for the memory switching unexpected cpu frequency's.
(Switching 4x Vengeance CML16GX3M4A1600C9B to 1600 MHz, which is standard)
It looks like Gigabyte has a wide spread issue with their new bios versions because I also read this thread:
http://forum.giga-byte.co.uk/index.php?topic=6194.0 (http://forum.giga-byte.co.uk/index.php?topic=6194.0)
It seems there are more Gigabyte motherboards having similar issues.
Although I am glad with my Gigabyte mobo I hope Gigabyte will solve these bios issues quickly.
I can't wait.
F6 bios (=latest)
(http://www.ocztechnologyforum.com/forum/attachment.php?attachmentid=21307&d=1312028834)
F5 bios
(http://www.ocztechnologyforum.com/forum/attachment.php?attachmentid=21338&d=1312064356)
-
EXACT same issue please see my thread http://forum.giga-byte.co.uk/index.php/topic,6321.0.html (http://forum.giga-byte.co.uk/index.php/topic,6321.0.html)
-
I just read your thread.
A few Gigabyte motherboards seem to have the same issue.
I also have, but didn't place, results from the ATTO results with MS and Intel AHCI driver.
At forums for SandForce SSD's it seems the latest Intel RST driver are hihgly recommended.
However I didn't see almost any difference at all and just stick to MS.
When I switch the memory to XMP profile 1 (1600MHz) the processor switches automatically to 3,8GHz minimum.
This is not supposed to happen, although my cooler can handle that.
It's just weird.
-
There does seem to be a problem with this BIOS version and I asked Gigabyte UK about it today. Their response was to get everyone that has this issue to contact GGTS so that they are aware of the of the problem and hopefully write a fix for it.
Just enter your email address and click on the language of choice.
GGTS http://ggts.gigabyte.com/
Please expect several days for a reply.
-
I already informed Gigabyte about this and I got an answer from the Dutch department.
This person said the issue was clear and reported it into the proper Gigabyte channels.
As soon as there is a bios version available which fixes these issues they would let me know.
And now we wait...
-
UPDATE
Gigabyte just released the F7A (beta) bios for the GA-Z68X-UD3P-B3 motherboard.
Gigabyte description: "Improve OCZ PCIe SSD compatibility"
Well, this probably won't solve the sata 3 SSD write bottleneck issue with F6 because it's meant for PCIe SSD's, but I decided to test it anyway. You never know...
The result:
No, it does not solve any of the issues I described with F6. :(
Let's just wait for Gigabyte to release a new version.
As soon as they do I'll try again.
F7A bios (see attachment)
Latest version of Gigabyte website, 30-aug-2011
Status: issues not yet solved.
-
I had the exact same problem. Try disabling C3/C6 State Support in bios.
-
I see that TAFB got in there first! ;) I was just going to pass on this bit of information in the hope that it might help with your problem GTZaskarLE
-
Hi TAFB and Dark Mantis,
Thank you for this interesting information.
TAFB, I saw your thread about this too.
I'm going to try this right a way and let you know how it went.
The only concern I have, and not have checked with F5, is why F5 did not have any problems with default settings en F6 did.
I don't know if F5 had different C3/C6 States.
To be continued! :D
-
Hi TAFB and Dark Mantis,
Thank you for this interesting information.
TAFB, I saw your thread about this too.
I'm going to try this right a way and let you know how it went.
The only concern I have, and not have checked with F5, is why F5 did not have any problems with default settings en F6 did.
I don't know if F5 had different C3/C6 States.
To be continued! :D
I don't know why but our security system decided not to display your post so I have quoted it here so people can see it. I notice it has relented now and your post is showing!
-
Maybe you will find this relatively useful...and, make sure you read reply (groberts101 03-23-2011 at 07:00:39 PM)
http://www.tomshardware.com/forum/267511-32-experts-wanted-benchmark-scores-vertex-raid0-setup
-
UPDATE
Hi TAFB and Dark Mantis,
Thank you for the "disabling C3/C6 State Support" tip: it worked!!! ;D
I never expected this could make such a difference and it makes me wonder what the setting was for the F5 bios when I had no problems.
I think I'll just go back to F5 to check the setting for [C3/C6 State Support].
Status: one down, one to go.
There is still the weird 'high cpu clock with XMP memory profile 1.
This clocks the CPU to 3.8 GHz minimum when the XMP memory profile 1 is activated.
@Carvetii
Thanks, I'll start reading this thread.
See attachment for the ATTO Disk benchmark with C3/C6 State Support disabled.
-
Good news that! The figures look more like what they should do now. You wouldn't really expect a couple of wait states to make so much difference but then we are talking about computers! ;)
-
Yes, good news.
In mean time I loaded and tested bios F5, F6 and F7A to investigate why F5 had no problems with the write speed and F6, F7A did.
My routine to load a new bios:
- Download the bios from gigabyte and put it on a memory stick
- Boot to bios and load the bios with Q-Flash from the memory stick: keep DMI data [Disable], Load CMOS default [Enable], reset
- Load fail-safe defaults [y], F10 Save & Exit setup [y]
- Load optimized defaults [y], F10 Save & Exit setup [y]
- Manually check and set if needed all but the MIT-settings
- M.I.T. setting: X.M.P. [Profile1] 1600MHz
- M.I.T. setting: C3/C6 State Support [Auto/Enabled/disabled], F10 Save & Exit setup [y]
-> This is the variable to play with
- In Windows 7: run WEI and reboot again
- Done.
For bios F5, F6, F7A the default setting for C3/C6 State Support is [Auto]
TESTS concerning write bottleneck with sata 3 SSD and ATTO Disk benchmark:
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [AUTO]
F5 : OK, no write bottleneck
F6 : Not OK, write bottleneck
F7A : Not OK, write bottleneck
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Enabled]
F5 : Not OK, write bottleneck
F6 : Not OK, write bottleneck
F7A : Not OK, write bottleneck
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Disabled]
F5 : OK, no write bottleneck
F6 : OK, no write bottleneck
F7A : OK, no write bottleneck
Conclusions:
- C3/C6 State Support should always be set to Disabled (thanks TAFB)
- It's not clear why [Auto] works for the F5 bios and don't for F6 and F7A.
Only Gigabyte could probably know the answer.
TESTS concerning CPU frequency 3.8 GHz with X.M.P. memory profile1 1600 MHz:
M.I.T. -> Advanced memory settings -> X.M.P. [Profile1] 1600 Mhz
F5 : M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> CPU Clock Ratio: 34x
F6 : M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> CPU Clock Ratio: 38x
F7A : M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> CPU Clock Ratio: 38x
The processor score in WEI goes from 7.6 to 7.7 with CPU Clock ratio 34x to 38x.
Conclusion:
This is unexpected/weird and I can not explain it.
I hope Gigabyte will do something good with it.
-
No I too am at a loss to explain why it is happening. Doesn't seem to make sense at all. I suggest that you note all that down and send it to GGTS for analysis and rework.
Just enter your email address and click on the language of choice.
GGTS http://ggts.gigabyte.com/
Please expect several days for a reply.
-
Yes, I will try to add this to an open call I have at GGTS about this.
It takes, as you said, some time and I'll first have to wait until they answer the last reply so that I can add a new reply (these results).
Starting a new call for the same issue makes no sense I guess.
The last time they answered they only asked me:
"Did you already update the Intel Management engine Driver and The Rapid Storage Manager?"
I answered I did and do always use the latest drivers/bios/software in common.
(among us: of coarse).
I'm afraid it is going to take some time until this really is going somewhere.
We will have to be patient. They probably do their best.
On the other hand I hope GGTS will read their own forums better.
I sometimes get the idea GGTS only acts if issues are brought to them trough the proper channels.
If issues are only on the Gigabyte forum it seems to not exist to them.
Probably and hopefully this is not true, but it feels like that a bit.
-
Apart from runn3R the official Gigabyte Forum Administrator there are very few occasions when anyone for the company comes onto the forum or reads any posts here. It is all sort of left on an unnofficial footing mostly. I always think it's a shame that more interest isn't taken by the company in one of their biggest assets as far as public liason and marketing goes.
-
Well, I'm very happy to see there is happening a lot on this forum.
Moderators/administrators are really acting here which I appreciate a lot, and many others too I think.
It's a pity there is almost no action on the Dutch Gigabyte forum, really quiet there.
-
It was the same here about 18 months ago, about 3000 members etc but we started putting in some hard work and long hours and built it up to what it is now. Just a shame that Gigabyte as a company doesn't really appreciate it all.
-
Yes that's a pity.
They should notice that reaction time at this forum is aproximately 10 times faster than with GGTS and there are really solved some issues here. It looks like many readers do appreciate the forum since it has a wide variety of Gigabyte related issues, tips, trics and information.
-
Yes I am sure that you are right there, the members do appreciate all the hard work that goes into making it a useful and fast response to their problems. That is what makes it all worthwhile really. Plus it keeps us off the streets!
-
Especially in your case because I just read you are "Deep underground in your EMP proof nuclear bunker" ;)
-
Thanks for this information, I've done my own tests...and they're not as good as most :(
(http://www.wealdcomputing.co.uk/wp-content/uploads/2011/09/110904-ATTO-Transfer-speed-test-agility-3-ocz.png)
(http://www.wealdcomputing.co.uk/wp-content/uploads/2011/09/110904-ATTO-Transfer-speed-test-agility-3-ocz-2.png)
The latter done after a few minor changes. I have disabled the C3/C6 option, as per This post (http://forum.giga-byte.co.uk/index.php/topic,6647.msg52488.html#msg52488) on this forum.
Thanks!
Chris
-
Hi christillis,
It looks like you have connected your SSD to a sata 2 port.
A sata 2 port can not transfer data faster than 3 Gb/s data (=approximately 300 MB/s in practice)
If this is the case then you really should connect it to a sata 3 port.
-
Also make sure that you have AHCI enabled and NCQ in the BIOS.
-
And it doesn't matter that much if you use the standard Microsoft AHCI- or the Intel R(apid)S(torage)T(echnology) AHCI driver.
I tested this with my SSD on the Intel Z67 sata3_0 port and there was almost no difference in performance.
-
Thanks for posting that bit of information GTZaskarLE . Always worth knowing.
-
It's surprising how the charts vary so much compared with earlier charts shown. Using the GA-X58A
same SSD plus SATA 3 PCIe card get consistant just below 300mb/s write & 425mb/s read. As I
plug in various drives have not enabled AHCI. Not as good as test reports shown. Was hoping for
nearer 500GB/s. Also stick to Win drivers for best results in my testing.
Regards tinker.
-
Hi Tinker,
Is there any reason why you haven't enabled AHCI ? I have noticed before that you don't run it.
-
Hi Dark Mantis, several reasons. One is that I progammed it so many times due to Marvell & USB 3 sick of
the sight of it !! Two, I use this PC for plugging in different drives from other PCs for wiping. Three, I didn't like the
added time on boot up. Plus had enough hassle with USB 2 to get it running at full speed. Also this is 2nd rebuild
due to an earlier ver of Gigabyte board & memory problems. Due to a period of ill health & a major rebuild of my
PC/ Hobby room, ( one of 2 desks was buckling under weight of equipment) All monitors, PCs, KVM switches
etc., have been upgraded to HD spec. The PC in question was started OCT 2010 & still not been used other than
programmed. The only thing running on it so far apart from basic software is FTA digital TV or Multi Satellite.
Used to use case for Video editing & drive wiping. Had two different Asus boards in it in past.
Common sense told me I should have returned this MOBO as have wasted hrs of research on net reading about
PCIe cards & faults. If I purchase goods that claim to be USB3 & SATA 3 then that is what I expect. A few yrs ago
any goods that I found were faulty they were sent back either for refund or direct replacement as would never accept
a repaired used item, as was my legal right. Kept board rather than scrap a £200.00 cpu.
Regards tinker.
-
Hey guys,
anyone try yesterday released BIOS F7b (http://www.fileden.com/files/2009/12/20/2696817/z68xud3p.7b.zip)?
-Improve OCZ PCIe SSD compatibility
-Fix Default C-state
-05.Sep 11
Maybe help your issue.
;D
-
If I purchase goods that claim to be USB3 & SATA 3 then that is what I expect. A few yrs ago
any goods that I found were faulty they were sent back either for refund or direct replacement as would never accept
a repaired used item, as was my legal right. Kept board rather than scrap a £200.00 cpu.
Yes fairly much the same as my experience when I first bought the motherboard and several SATA3 components that wouldn't work. As you say in the end decided to keep it as apart from that was a good board.
-
Hi ex58,
Thanks for noting this and no, I had not even seen F7B yet.
As soon as it will become available at the gigabyte website I will test it and present the results here.
(I personally choose to use genuine firmware & software only in general)
-
Hi Tinker,
Do I understand correctly you use a OCZ Vertex 3 120GB SSD connected to a SATA 3 PCIe card on a GA-X58A motherboard?
If yes, which SATA 3 PCIe card do you use?
-
Hi christillis,
It looks like you have connected your SSD to a sata 2 port.
A sata 2 port can not transfer data faster than 3 Gb/s data (=approximately 300 MB/s in practice)
If this is the case then you really should connect it to a sata 3 port.
Hi GTZ, thanks for your observation...could have sworn I plugged it into Sata III! I'll shut down now and check :)
-
Yes thats correct GTZaskarLE . I purchased a Manhatten 6GB/s 150408. It was purchased before I joined Forum & the
only one that came up on google. The cheapest I could find was at North East Peripherals LTD. £46.69 but that included £6.90
P&P + VAT which was a bit over the top. The good thing about it with single drive is fast boot no drivers installed. Fastest slot
was 2nd x16 slot though only PCIex1. 2 outputs 2nd one was fastest. Most recomend the Rocket Raid Highpoint but do some
research as 620 & 640 have slow boot up times. If you want raid you need dual chip & sadly all Marvell 9128. No 9182 about
that I know about. Scan do the Rocket cards & no idea if faster than mine. Startech, Xenta are other names but no reviews.
May be faster if I used AHCI.
The really good cards are around £300.00 plus a lot more. Cheaper to change MOBO !!! Hope that helps.
Regards tinker.
-
Hello again,
Well I checked and it was indeed plugged into SATA3 (after spending all this money on SSD, I thought I would be careful!). However, the SATA cable itself was black and I suspect possibly a SATA2 cable (nothing written on it of course), so I removed it and used a Yellow cable that came with the board and that I suspect is SATA3 capable. Here are the results:
Booted back up after cable swap:
(http://www.wealdcomputing.co.uk/wp-content/uploads/2011/09/110906-ATTO-Transfer-speed-test-agility-3-ocz-cable-changed.png)
Tested again after I noticed the SATA power cable was loose:
(http://www.wealdcomputing.co.uk/wp-content/uploads/2011/09/110906-ATTO-Transfer-speed-test-agility-3-ocz-cable-changed-power-loose-corrected.png)
So an improvement I think...and thanks! ;)
-
Here's a probably stupid question, but it seems that in most transfer instances (using ATTO), the write speed is greater than the read speed - I would have thought this would be the other way round? ???
-
That's quite an improvement christillis. My read is 425MB/s but much poorer write. All
the 6 GB/s cables I have purchased are black.
Regards tinker.
-
All the 6 GB/s cables I have purchased are black.
Ah....bugger :D I do have a stash of Sata Cables, perhaps it was a very old one I picked out the box? I wonder if there's a way of confirming whether a cable is sata 3, other than plugging it in and testing? I have a great Power tester, be nice to have a data cable tester too.
Thanks, Chris
-
Hiya Cris....
Just so you don't get the wrong idea.....This is an extract from the original ........ sent to JEDEC for Registration (In Bold)
In the transition to SATA 6Gb/s, it will be important to use high-quality cabling.
Problems may be related to the use of cables made from marginal materials that
perform at the edges of SATA 3Gb/s tolerances, which could become a failure point at
the faster 6Gb/s signal rates. SATA-IO therefore recommends that only high quality
cables and connectors be utilized for SATA 6Gb/s.
Translation..........there is no physical design difference in Sata cables or the core count with the migration from Sata II to Sata III.... more positive connectors, cable quality, superior shielding Technology and it's implamentation and tighter manufacturing tolerances has and will continue to improve data transfer rates
So what it comes down to is.........."You Get What You Pay For"
I resonantly went over to a newer Lead shielded sata cable with a much more positive clip.......use the best shielded cables you can afford!
Good practices when working with any Data cables......Use the same brand/spec through out.....if you run flat data cables together....even a 1mm gap will make a difference..........
In a given raid set use the shortest cable you can get away with but cables should be the same length on the given raid set regardless of longer tails
Try and route Data cables away from any power cables as much as humanly possible
All these little tricks combined can reap up to 20% better performance........ hope this helps!
Aussie Allan
-
Hi christillis,
Nice to see your SSD performance improvements.
I agree with Aussie Allan concerning cable quality.
Here I have some other questions for you, because I think your SSD performance could still be better.
Which firmware do you use for your SSD: v2.11 or lower?
Especially v2.09 is known to be slower.
OCZ recommends to update to v2.11 in general, for stability and for performance.
I am using v2.11.
If your SSD is your boot drive use the Linux-based firmware update method.
http://www.ocztechnologyforum.com/forum/showthread.php?89670-Bootable-Tools-for-OCZ-Vertex2-3-Agility2-3-Solid3-Revo-and-Ibis-SSD-s (http://www.ocztechnologyforum.com/forum/showthread.php?89670-Bootable-Tools-for-OCZ-Vertex2-3-Agility2-3-Solid3-Revo-and-Ibis-SSD-s)
What sata mode did you set in the bios for your SSD?
It should be AHCI.
Or RAID and then AHCI within the Intel RST driver for a 'RAID ready system.'
The trim function will work in both cases.
On what sata 3 port did you connect your SSD? The intel or the Marvell sata 3 port?
You should connect to the Intel sata3_0 port.
The Marvell sata 3 ports are limited at 5 Gb/s.
The Intel sata 3 ports will do 6 Gb/s and are known to be faster in gerenal for sata 3 SSD's.
-
I know Allan has already got in before me but just to reiterate there are no differences in SATA cables from the very first ones apart from when they ratified the SATA2 standard they incorporated the little metal locks to stop them falling out of the socket. Apart from that all cables are equal but some are more equal than others! You do get what you pay for and even then have to take care in the routing of them.
-
First of all, I have to say how grateful I am to you all for your help with this - Aussie Allan, GTZaskarLE & Dark Mantis. You're all top rate people!
Thanks to Allan for confirming that SATA cable types only vary in quality, not speed. I have looked into some HQ cables such as Overclockers Akasa's (http://www.overclockers.co.uk/showproduct.php?prodid=CB-045-AK&groupid=1929&catid=1928&subcat=), so I may well go down that route a some point...bit of a shame though, as I have dozens of SATA cables of all lengths (from client upgrade/new builds).
I also have to admit, SATA cables in my tower run tight to each other and power cables around them - haven't really sat down and thought "cable management!", yet :) It's a bit of a mess as I have 6 SATA cables in use, but I will have a tidy up when I stop needing the computer for earning money :D
GTZaskarLE - I will definitely find the answer to your questions - especially as they aid me in a speed increase! Thanks for giving me a link to a firmware upgrade where the SSD is boot, I had gone to the OCZ website briefly and only downloaded the upgrade file!
I'm fairly sure from memory the SATA mode is IDE, I did assume that the RAID mode should be used when more than one disk is being used, I also thought that AHCI serves little benefit to 'mono drive' solutions?
I'm unsure of the precise sata port...athough I'm fairly sure it's in the wrong one :)
Hi Dark Mantis, thanks for confirming - always nice to get double confirmation :) Yes I only have the metal locking version these days, although I did prefer the push in ones, handy for tight cases.
As I've not set up a RAID proper, I need not worry about cable lengths...but that is very useful to know - logically similar to tuning the length of motorcycle exhaust pipes! One thing I could do with is an outlet for super short SATA cables (100-200mm), because my motherboard (like most Gigabyte's I think) have side access SATA ports, so they point almost directly at most of my HDD's.
Thanks again everyone ;) - you are the most useful & friendly forum I have ever used thus far (blush).
-
Check Ebay for the very short SATA cables. Obviously you have to be careful what you are buying but I purchased some very nice ones a while ago and picked up a few once I had verified the quality.
Thanks for the kind words by the way. ;)
-
Check Ebay for the very short SATA cables. Obviously you have to be careful what you are buying but I purchased some very nice ones a while ago and picked up a few once I had verified the quality.
Thanks for the kind words by the way. ;)
I had forgotten the link to the SATA cables, my bad. I have updated the post, but if it's too late, here it is: Overclockers UK (http://www.overclockers.co.uk/showproduct.php?prodid=CB-045-AK&groupid=1929&catid=1928&subcat=). :)
Thanks DM, I was considering the eBay way...but we all know what that place is like :o
You're welcomo ;)
-
@Dark Mantis, just a question, you mentioned a while back best to update firmware via IDE
mode for Vertex 3 but the link is for AHCI. Am assuming there are other ways of upgrading as can
always plug my drive into USB. No idea as yet how to find firmware used on SSD unless the update
software tells you.
Regards tinker
-
As with any firmware update it is best and safest to make sure that the drive being flashed is the only one connected as this saves any confusion or errors. ;)
-
I'm back.
With a nice response of GGTS (Gigabyte Global Technical Service)!
They provided me with the beta bios F7B which is not yet available on the Gigabyte website. :)
They asked me: "Please run a test with the attached beta bios and let us know your test results."
So here we go with F7B.
TESTS concerning write bottleneck with sata 3 SSD and ATTO Disk benchmark:
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [AUTO]
F7B : Not OK, write bottleneck
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Enabled]
F7B : Not OK, write bottleneck
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Disabled]
F7B : OK, no write bottleneck
Conclusions:
- C3/C6 State Support should still be set to [Disabled], not to [Auto].
- It's not clear why [Auto] works for the F5 bios and don't for F6, F7A and F7B.
TESTS concerning CPU frequency 3.8 GHz with X.M.P. memory profile1 1600 MHz:
M.I.T. -> Advanced memory settings -> X.M.P. [Profile1] 1600 Mhz
F7B : M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> CPU Clock Ratio: 34x
That's good.
There is still left one inexplicable item.
When I enter the M.I.T. menu (not entering any of the 5 menu's) it shows:
CPU Frequency 3812.32 MHz (??)
However, as showed above, within the Advanced Frequency Settings menu the CPU Clock Ratio is 34 which is expected.
Conclusion:
- Switching to XMP memory profile1 (1600MHz) does not clock the processor to a minimum of 3.8 GHz anymore.
It stays at 3.4 GHz which is what one can expect.
- It is unexpected to see the CPU Frequency status at 3812.32 MHz at the M.I.T. start screen
So the overall conclusion would be there has been made some progress but we are not yet there completely.
It's looks like Gigabyte is taking this serious which I appreciate a lot.
Go for F7C? ;)
(I informed GGTS about this test)
-
That's very useful GTZ, thanks for sharing your response from GGTS :)
-
Once again very impressive GTZaskarLE . Have replied to your last PM re X58A.
Doubt that board will achive those scores with Vertex 3 max iops unless a lot of
money spent.
Regards tinker.
-
Well it does appear the GGTS are doing their best to find a solution and are on the right trail. I think they should be applauded for this as even though it is what they are there for we are quick enough to castigate them for not trying at other times. ;D ;D
-
Hi DM, Christillis & Tinker,
Thanks and yes, it's going into the right direction.
In mean time I got two other reactions from GGTS.
They have sent this thread to the head quarters.
Concerning the wrong CPU frequency status they told me it would be fixed in the next bios version.
There is a so called "running change" on this issue.
The C3/C6 State Support will probably be set to Disabled by default in the next bios version.
We will have to wait for a couple of days for the new bios release.
So I am interested what will happen.
In mean time I will be offline for some time because tomorrow my summer holiday starts.
I will be back in the second half of october.
See you guys later.
-
Not sure if this was already said, but the reason F5 with AUTO doesn't have the issue is because F5 disables C3/C6 if set to auto. I flashed with F5 and the issue was gone, but cpu voltage now stays at 1.2V when idling, instead of dropping to 0.9V, meaning the suspend states are disabled.
If the fix in the new bios is going to be just disabling C3/C6 by default, it's going to be useless. They should make a proper fix.
-
Correction to my previous post - F5 does disable c3/c6 by default, but cpu voltage doesn't stay at 1.2 at idle, but drops like before to about 0.9V. Not sure what I was looking at the first time :).
Idle power consumption increases by just 5W with c3/c6 disabled so it's accceptable.
-
If the new BIOS has C3/C6 disabled by default that is fine as a workaround but it shouldn't be accepted as a permanent fix because some people might want to have the C3/C6 functionality. The actual problem still needs fixing.
-
I use an GA-Z68XP-UD4 and Corsair Force 3 SSD and have this issue with write speed with C3/C6 State Support.
I just updated to the latest F5 Bios (From F4) and the issue is still there, just we now seem to have no C3/C6 State Support available at all. EG no matter what you put C3/C6 too, the VCore voltage doesn't drop on idle.
My Turbo Ratio's are 42-45 hence the higher VCore of 1.36. Some pretty pictures to prove it.
-
After reading this post I remembered that I got the same results with my OCZ Vertex 3 in the beginning. I have the GA-Z68X-UD7-B3 board with F8 bios
I never got around to test it and see what caused it, but the problem is gone now when I'm running 2 OCZ Vertex 3 in RAID0.
This seems to be a problem spread across many different motherboards by the look of this topic!
-
Regarding the vcore not dropping when idle, I've noticed that if you enable/disable C1/C3/C6 on Z68X-UD3P in bios and just restart, for some reason the board always keeps the high vcore. Only after shutting down and powering on again, the board starts dropping clock and voltage properly. C1 enabled alone should be enough for this to work.
This is true if you are using Auto and Normal (Dynamic) vcore settings, and probably not if using Manual, in which case vcore is fixed.
And btw, in the new F7 bios, still have to disable C3/C6 to get rid of the bottleneck issue.
-
That's true, I get the same results with testing F7:
TESTS concerning write bottleneck with sata 3 SSD and ATTO Disk benchmark:
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Auto]
F7 : Not OK, write bottleneck
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Enabled]
F7 : Not OK, write bottleneck
M.I.T. -> Advanced frequency settings -> Advanced CPU core features -> C3/C6 State Support [Disabled]
F7 : OK, no write bottleneck
Conclusions:
- C3/C6 State Support should still be set to [Disabled], not to [Auto] or [Enabled].
- It's not clear why [Auto] works for the F5 bios and don't for F6, F7A, F7B and F7.
I will inform GGTS about this, my ticket about this is still open.
I fully agree with DM who said:
"If the new BIOS has C3/C6 disabled by default that is fine as a workaround but it shouldn't be accepted as a permanent fix because some people might want to have the C3/C6 functionality. The actual problem still needs fixing."
So Gigabyte still has to figure this out.
Let's wait for what they can do.
-
Hello guys
Just wanted inform that 34 vs 3800 mhz are actual for z68xp ud4 as well on f5 bios. Just wanted to ask is therethe same write bottleneck problem on ssd sata 2? Shoul i make corresponding ammendments to bios? And another question is what is NCQ, should i use that with AHCI enabled?
Thx
-
Concerning the first question:
I don't know that because I don't use such a hardware configuration.
Concerning NCQ:
See http://ru.wikipedia.org/wiki/NCQ
(I hope this is in your language :))
And yes, if your drive supports NCQ it's recommended to use it, also in AHCI mode.
For instance, my OCZ Vertex 3 maxiops supports and uses NCQ at sata III AHCI mode.
I hope this helps.
-
UPDATE
It looks like Gigabyte Global Technical Service does not take the problem very serious.
This is their response on my ticket where I check in my F7 test.
Dear Mr.,
Is your windows 7 up to date?
Sp1 installed and all updates?
Please check, if there is an improvement with bios F7
http://www.gigabyte.eu/products/product-page.aspx?pid=3850&dl=1#bios
Did you already update the Vertex 3 SSD to firmware 2.15?
http://www.ocztechnology.com/ssd_tools/OCZ_Vertex_3,_Vertex_3_Max_IOPS,_Agility_3,_Solid_3,_RevoDrive_3_and_RevoDrive_3_X2/
Kind regards
GIGABYTE-Team
Very kind, but yes, yes, yes, I am up to date with everything and so on.
I am getting a little bit tired of this...
This open ticket shows in meantime a very tiny vertical scroll bar.
:(
I of course answered all questions but also mentioned that this probably wouldn't speed up the resolving process.
Let's wait for their next move.
:)
-
My response (which took over a week) from GGTS on this subject was just as disappointing..
Did you download the F5 BIOS from the website?
The released F5 has the same default settings in C3/C6 state as “auto”, the option should be manually changed to disable as the former versions.
::) ::) ::)
Looks like returning the board as "unfit for purpose" is the solution.
-
Responded to GGTS mail, not happy about their lack of effort in resolving this issue (They got informed about it 3 months ago, looking at GTZaskarLE post dates). Giving Gigabyte a week to respond with a satsifactory answer, otherwise returning my board and getting a different manufacturer board.
-
UPDATE:
I got this response from GGTS:
The problem is duplicated now.
The hardwareteam is looking for a solution now.
We will update you once a solution is found.
It finally looks like they have tested it and found the same issue.
Let's wait for what they can do.
I hope it will be a permanent solution, not another workaround.
-
@GTZaskarLE, just for the record, using a small form factor PC first attempt at Raid 0, Read 1000mb/s write, 1020mb/s read. Next step
is to try2 new 240gb OCZ Vertex 3 max iops. Initial test was with Corsair 3 120gb SSDs.
Regards tinker.
-
UPDATE
I tested bios version F8
Results: no difference
It still is not possible to have C3/C6 states enabled without a serious write speed bottleneck with an OCZ Vertex 3 120GB maxiops SSD.
The GGTS still has this under investigation... for more than half a year now...
To be honest, I lost my confidence in them.
This simply takes too much time.
I also noticed bios version F8 was/is removed from the official Gigabyte website.
This is the second day it is not present anymore.
Why?
Was F8 still a beta version?
It was not stated that way.
-
Thanks for the update GTZ, it's certainly a bit of a let down to be honest, but at least 'they' are aware of the problem.
All the best,
Chris
-
Hi Chris,
Thanks.
Yes, they are aware of the issue.
They even have duplicated it.
I'll just have to wait for them to come up with a solution.
There must be others having the same issue.
Of coarse this is not a very severe issue. The SSD is working fine, also with half the write speed.
But it's a pity that I can not make use of the maximum write speed and have C3/C6 state enabled at the same time.
-
UPDATE
I got bios version F9A from GGTS (beta, not yet released) and tested it.
Results: SUCCESS!!
Ok, it took GGTS 8 month to solve my ticket...
And I was helped by the Dutch Gigabyte department to speed things up;
But finally it is possible with the GA-Z68X-UD3P-B3 motherboard with bios F9A to have:
- CPU Enhanced Halt (C1E) ENABLED
- C3/C6 State Support ENABLED
- CPU EIST Function ENABLED
- All CPU Turbo Ratio's set to 40
And last but not least...
- Full SSD write speed as it was meant to be.
;D
As you can see I also slightly overclocked my system, all cores to turbo 4 GHz, but they all idle at 1.6 GHz.
As a result the Windows Performance Index slightly improved.
(The primary HDD score went up 0.1 point...)
There are no issue's left at this point.
I'm happy, assuming final F9 will perform the same as beta F9A, but why wouldn't it?
GGTS, thanks for all the effort you put into this.
Bernice, from Gigabyte Netherlands, thanks for speeding up the process.
See attachment.
(open in new tab)
-
GTZaskarLE, that is great news!...just can't believe how long it's taken, you must be commended for your persistence with this and I for one, thank you!
Look forward to the official F9 to come out!
;)
-
Hi Chris,
You are welcome.
Yes, I'm very happy with this result.
For the record;
In the end this GGTS ticket contained 54 posts before it was solved.
Good luck with F9.
-
Great to hear they have solved this for you!
Can you please share the BIOS for others, thanks!
-
Hi Lsdmeasap,
Yes of coarse, assuming that's ok with GGTS and this forum.
Keep in mind that this is a beta bios.
However it appears to be very stable on my system.
The zipped file size is 1622 kB while the maximum attachment size allowed is 1024 kB...
So I split it up into two zipped, renamed fragments.
(renaming was necessary because not all extensions are allowed here)
How to construct one bios file from two zipped, renamed fragments with 7-zip file manager:
http://www.7-zip.org/ (=freeware)
1. Download Z68XUD3P.zip.001.zip (See attachment)
2. Rename it to Z68XUD3P.zip.001
3. Download Z68XUD3P.zip.002.zip (See next post, because I can not attach more than 1024 kB)
http://forum.giga-byte.co.uk/index.php/topic,6292.msg67819.html#msg67819
4. Rename it to Z68XUD3P.zip.002
5. Install 7-zip, or use another zip-program which can add archives together
6. Select Z68XUD3P.zip.001, both file fragments should be in the same folder
7. right click '7-zip, extract here'
Z68XUD3P.F9A should be recreated which is the beta bios F9A file.
The file size should be 4096 kB (4.00 MB or 4,194,304 bytes)
Good luck!
;)
-
Part 2: Z68XUD3P.zip.002.zip (See attachment)
See previous post for instructions.
http://forum.giga-byte.co.uk/index.php/topic,6292.msg67818.html#msg67818
-
Thanks for sharing! You could have just uploaded it to a file host instead :)
I know how to deal with it though, thanks again for sharing!
Also, no worried about this being a beta BIOS, we host and use beta BIOSes all the time, see here :)
http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-beta-bios.html
-
Yes, I knew this could be done simpler... :P
But I'm afraid I don't have any experience with file hosting.
I saw you already uploaded F9A at TweakTown Forums.
Thanks, this makes it easier for people to download.
-
Ahh no worries, but in case you want to know for future use you can google free file host and pick one. Some let you upload without an account, others you have to have an account, and some of the free ones you can make an account on anyway that way you can keep track of your files and links.
Yes I posted as soon as you shared the file, thanks for sharing! We try to help all Gigabyte users, so having random beta's like that are nice to get, because we'd get it eventually, but since you posted it's now sooner rather than later
-
Man, I had to look at your thread didn't I? :D
I had to test, otherwise it would have bothered me in the back of my mind until I did!
(http://i.imgur.com/p9iOy.jpg) (http://i.imgur.com/0YSNr.png)
C3/C6 setting has no effect! I have not tested EIST and C1E, but I'm sure after the fiasco here they know what to look for and fix.
So, needless to say X79 is also affected by this, report going in now!!
-
Yes.
No complains to you, but if you as 'hero member' had looked earlier into this you could also have pushed Gigabyte harder, together with Bernice (from Gigabyte NL) and me, and then this issue maybe got solved earlier.
It took almost for ever, but they did it.
In mean time DM was temporarely lost 'deep underground in his EMP nuclear proof bunker'.
(just joking DM ;))
And if I understand you correctly;
Yes, the C3/C6 state setting does not matter anymore for SSD performance.
My main issue was that all processor energy saving features like C1,C3,C6 states and EIST should not have impact at SSD performance.
I saw bios version F5 working correctly, but all versions after that gave severe write performance hits.
Until now.
And yes, this is probably something to focus on too for other chipsets.
-
What do you mean "Hero Member"?
But yes, I'm sorry I never saw this thread earlier, I'm not always here as much as I would like, so there is times I may miss many threads I would have been able to help with!
I tested C3/C6 because I thought that was the main setting that initially allowed you to have or not have the bottleneck. Should I also test C1E and EIST and see if those enabled or disabled remove the bottleneck? I guess it really doesn't matter, as I'm sure they know the exact cause and which settings can effect it anyway, so I'm sure they'll fix it up soon since they just recently found this fix for you.
I see P67 and Z68 affected now too, so it must have been some recent BIOS change or addition that created this, because when I reviewed P67 UD4 and UD7 (B2) this was not an issue at all, probably just like you mentioned with the older BIOSes not having it
You are right though, none of these should have ANY effect on the SSD speeds, but I do often see speedstep having a little effect on speeds (Same as using a balanced of power save plan in windows), but nothing like this write bottleneck.
Again, I'm sorry for these hassles you had to go through, and the long amount of time you had to wait for a fix!!
-
Well, I saw in your profile: position 'Hero Member' and you are very active here and at TweakTown.
So I probably was wrong about your involvement at this website.
I thought you are some sort of administrator here, connected with Gigabyte.
But no worries, I know you follow lots of threads, are only one person and cannot be everywhere.
I didn't want to make you feel responsible. In addition, internet makes conversations always different from when you speak in person with somebody about something (and in your own language). Things can easily get out of proportions and that wasn't my intention. I only was a little annoyed about GGTS in general, but not anymore.
Concerning testing C1E and EIST: I already saw it has no impact anymore with F9A.
If you want to verify, please feel free to do so, it is always nice to see something verified/confirmed.
If this is really something recent?
All I know is that is affected my Z68 board for more than half a year, let's say after F5.
It probably affects also P67, but I cannot be sure.
On the internet I found not much about this, other than disabling C3/C6 states which was not the solution for me.
Let's hope Gigabyte fixes this for the whole range of existing P67 and Z68 chipsets.
And pay attention for next generation chipsets too.
-
Ahh, sorry, I didn't even notice that was my user group until you mentioned it! I think that may be based on my post count, or some mod or admin may have set it like that, not sure?
Yes, I am very active for Gigabyte at many places, and I try my best to make my rounds to many forums, this one included! I am just a normal user here, but Admin at Tweaktown. I don't work for Gigabyte though, but I am directly in contact with them, so I guess you could say I am connected with them in that aspect, but in the end I'm just a normal end user like most everyone here.
No worries, I wasn't offended by what you said, just a little confused as I wasn't sure what you meant by Hero until now :D
I'll try to test with C1E and EIST later tonight if I get time, but as you know I've already sent in a report on this issue for X79 and they are working on it already. And I know they are also working on other boards BIOSes for this issue as well, so hopefully they'll have it sorted out for all models very soon. It's odd how it affects ATTO only, but not AS SSD or CDM!?
Ya, I'm not sure when it started, but I know it wasn't always like that as you mentioned, because I did reviews on P67-UD4 and UD7 a while back and they were not affected then, but may be now (haven't checked yet, but will)
I'll keep you posted once I hear back on a fix for X79 in case you know other affected. You can also keep an eye out here for new BIOS releases, as the may be released with fixes mentioned in the change notes, in case you or anyone else reading wants to keep an eye on other boards that may also need corrected
http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-beta-bios.html
-
UPDATE
In mean time OCZ has released new firmware for the Vertex 3 maxiops: v2.22
The test results are good, see attachment.
I wonder when Gigabyte releases F9 for the GA-Z68X-UD3P-B3 motherboard.
As soon as they do I'll test if the F9A improvements are copied into it.
That's the 'SSD write speed in combination with C1E, C3/C6 State Support and CPU EIST Function enabled' fix.
-
That is looking much better now. Those are some good figures.
-
hello everyone.
i resurected this old thread today
Slow SSD Write Speeds? Try This (http://forum.giga-byte.co.uk/index.php/topic,6647.msg72949.html#msg72949)
then i found this one. Its not the same motherboard i have, i have the z68x-ud3h b3 rev 1.3 F12 but i still seem to be locked in write speed even with c3/c6 disabled.
anyone found a workaround for my board especially ?
you guys seem very helpful i tough i'D give it a try
(http://img33.imageshack.us/img33/1157/capturebxz.png)
MERCIIIII
******* EDIT FIXED **********
i had to set c3/c6 to disabled AND cpu halt state to disabled for it to fix my problem !
(http://img820.imageshack.us/img820/4589/capture2gw.png)
-
Hi BIGguiX,
As far as I know there is (still) no Bios version (yet) for GA-Z68X-UD3H-B3 which fully fixes this problem.
The main goal within this thread was to get full write speed for the SSD with all CPU energy saving functions enabled.
So setting c3/c6 to disabled AND cpu halt state to disabled will give you maximum write speed but without CPU energy saving functionality.
If this is enough for you than you could decide to accept it.
If not, then you could ask GGTS for a solution.
Use F9A for GA-Z68X-UD3P-B3 as reference.
Good luck!