MSI Z68A-GD65 G3 LGA 1155 Motherboard Review @ [H]

I plugged my keyboard in the other day after moving the computer and suddenly it worked during POST and in the UEFI. The only difference from before is that I have it plugged into one of the top 2 USB ports instead of one of the other USB 2.0 ports. Too bad the mouse still doesn't work unless I unplug/plug.
 
Astral Abyss said:
I plugged my keyboard in the other day after moving the computer and suddenly it worked during POST and in the UEFI. The only difference from before is that I have it plugged into one of the top 2 USB ports instead of one of the other USB 2.0 ports. Too bad the mouse still doesn't work unless I unplug/plug.
Click to expand...

what mouse?

I plug both my mouse and keyboard into the top 2 ports and it always works perfect. I am also on the beta .N49 BIOS.
 
Well - I just built a 2700k system on this board, and I have to say, this is the worst motherboard I've ever owned.

My main complaint is that despite having flashed the bios with practically every single beta and full version bios out there, the problem of cpu throttling persist.

Basically, you can set the multiplier of the board at say, 50x100, running at say, a 1.4vcore - and as soon as you go under load, like when running intel burn test...if you watch cpu-z and realtemp, what you see is that the motherboard will throttle the cpu to around 42-44multi (it varies, and bounces around a lot...)

Some claim that certain older version's of the bios fix this, trust me, while it might seem to work - it isn't a real fix.

If you search around the MSI user forums, you will see that this is a common issue on almost all of MSI's z68 line....I'm far from the only one.

I've had it with the POS board, stay away from it. I'm returning it to newegg asap.
 
return it ... i wish I had... but I was past my 30days before I really could start overclocking and tweaking the board.

You will have to run Beta BIOS to fix the throttling issues... I am running n49 beta bios and it works flawlessly.... but doesn't change the fact that I hate the board and will sell it once the z77 boards launch :D
 
Nickbaldwin86 said:
return it ... i wish I had... but I was past my 30days before I really could start overclocking and tweaking the board.

You will have to run Beta BIOS to fix the throttling issues... I am running n49 beta bios and it works flawlessly.... but doesn't change the fact that I hate the board and will sell it once the z77 boards launch :D
Click to expand...

See that's the thing. I have tried N49, and N34, and N00 - all of which some people claim fixes the throttling issue. I can say with absolute confidence that it does not fix it.

It may fix it for some, but it is not a universal fix because it doesn't fix it for me. From what I can tell, I think this is because I am using a 50 multi - the consensus is the higher the multi, the more the board tries to throttle the cpu.......for example, with N49 if I set the multi to say.....47 on my 2700k, I will see throttling occur, but it generally takes longer before it does (that is to say, when running Intel burn test on very high, it might take 60 seconds before it starts) and generally, the throttling isn't as aggressive with lower multi's..

For example,

When running 50x100 at 1.4Vcore - I'll see aggressive throttling immediately after starting Intel burn test on very high, de-clocking down to about a multi of 40 or 41 in a matter of seconds...

When running 44x100 at 1.4Vcore - I won't see any throttling for a minute or so, but eventually, it still happens.

I think it stands to reason that this may in some way be related to power draw from the cpu - I've played around with lots of settings, but nothing available in the bios on the Z68a-gd65-g3 allows you to control duration and limits of power in relation to clock

Moreover, I would like to affirm that my issue is in no way related to heat or thermals....realtemp reports that even after 50 runs of intel burn test on very high, my maximum temp was 73c, on core 3 of my particular chip. Currently being cooled with a corsair H100 with 4 high performance 120mm fans on the radiator.

For me, this issue is absolutely unacceptable - MSI has lost my business for life, I honestly feel like this so called "bug" seems, in some way, a very shady way of letting people think they have successfully overclocked their sandybridge cpu - and even making them think it is stable at their overclocked speed, because unless you pay very close attention you would never notice that the throttling is occurring......and you would think that you had successfully overclocked you 2700k to 5+Ghz on a reasonable voltage, when, in fact, the board runs the chip at your overclocked speed most of the time..........except when you need it most, under heavy load, it silently de-clocks and throttles the multiplier...
 
Elipsen the MB has a 5 year warranty. I can understand your frustration with the BIOS problems but MSI has been working on it.

I have seen Nick at the MSI user forums. There is a good group over there that are very patient and full of knowledge to help solve problems. Reading your post I think you are getting paranoid about this problem.

The BIOS has to be set just right for everything to work right. I open CPUZ, RealtempGT with IBT when I do stability testing. I have done the standard setting, very high and max. I'm stable at 4.8GHZ with 2500k and I haven't seen any type of throttling. Time and gflop out put seem to scale evenly for each run and cpuz and realtemGT report a constant 4.8GHZ speed.

I have had BIOS problems as well. I'm running the newest BIOS from the MSI WEB site. I used the MSI flash tool from the MSI user forums and my BIOS doesn't suffer any type of throttling issues. Also after you flash its standard procedure to take out the CMOS battery and do a proper clear CMOS. I'm still learning how to OC this mb. Over all I'm very happy with it. Looking forward to using an IB cpu with it when they are available.

I haven't had any throttling issues. I have BSOD's which is because my system isn't stable at high speeds.

I think there is much more OC potential to be had with this MB but it's going to have to happen at the BIOS level. Right now x50 multiplier stable for 2500k is not an easy task to accomplish.
 
Last edited:
Try n48 then... that should work.

I run my board at 52x100 and at 50x100 I have no throttle issues.

otherwise you have a bad board and need to RMA it.

I hate the board... return it to newegg and buy a Asus z68 board... will be much happier, it is what I have in my other rig and it is an amazing board, but I bet anything is compared to this board.
 
Nickbaldwin86 said:
Try n48 then... that should work.

I run my board at 52x100 and at 50x100 I have no throttle issues.

otherwise you have a bad board and need to RMA it.
Click to expand...


50x and 52x multiplier could be dependent upon cpu. Not all 2500ks will do 5GHZ.

Nick are you Intel Burn Test stable at 5ghz and 5.2GHZ? (on standard or very high settings)
 
Last edited:
SonDa5 said:
50x and 51x multiplier could be dependent upon cpu. Not all 2500ks will do 5GHZ.

Nick are you Intel Burn Test stable at 5ghz and 5.2GHZ? (on standard or very high settings)
Click to expand...


I would suggest a lot more would NOT than do. Even Intel said as much to me directly. You are dealing with a lot of heat at that point. But we have seen some of the better boards get our CPUs higher than we had seen those go in the past, but IIRC, only 2 LGA1155 boards did 5GHz with our 2500K.
 
I have this board. I also have a question.

Is it just me, or is the motherboard incapable of regulating non-PWM fans?

Maybe I should have done more looking into this before I bought the board, but I just figured it was something that the motherboard just did, especially since there are 3 pin fan headers scattered all about the board. Going into SpeedFan, non-PWM fans all read 0 RPM while moving at max, and this is with multiple fans.

Is this something that is a feature on some motherboards and not others, am I overlooking a setting, or is this an oversight on MSI's part?
 
SonDa5 said:
50x and 52x multiplier could be dependent upon cpu. Not all 2500ks will do 5GHZ.

Nick are you Intel Burn Test stable at 5ghz and 5.2GHZ? (on standard or very high settings)
Click to expand...

of course and not all 2600ks will either... with water cooling and enough vcore you can get it to go pretty fast :D

Again... I only run this for bench, not for daily usage, sure if I did I would burn the life out of my CPU pretty fast. yes 100% stable... @ 1.55vcore and 1.9pll :D
 
I ended up buying this board's little brother (Z68MA-ED55), and I'm pretty happy with it so far. The UEFI interface is meh. Double-clicks are awkward, but no biggy. My ASUS P8B WS UEFI interface is definitely "feels" far better. It isn't bad by any means... if I hadn't had the ASUS board, I would be happy with the MSI implementation.

I have PLL set to 1.600v, PLL overvoltage disabled, everything at auto except CPU voltage locked and Vdroop set to low.

Seems my 2500k is IBT-stable at:
4.2 @ 1.272v, 66*C
4.3 @ 1.296v, 68*C

4.4 wasn't 100% stable at 1.320v and was at 70*C. I imagine one or two small notches up and it'd have been stable. 4.5-4.6 wasn't happy at 1.36v, and that got to 75*C pretty quick.. probably would've been close to 80*C. This is with a stock H60.

EIST is on, C-states are on auto, my CPU still downclocks to 1.6GHz at light load. Voltage goes down slightly but nowhere near stock idle volts. Under no load though, it doesn't seem to make more than a couple watt difference on CoreTemp's wattage reading.

Overall, quite pleased. Now if AMD would just release their 78xx cards, I'd be even happier.

EDIT: My only complaint/oddity is that it seems to randomly wake itself up during the day/night. Now I do have a Logitech Unifying receiver plugged in and set to wake on USB (so I can wake it on purpose with keyboard/mouse of course), so I suppose that could be picking up some interference. I find that hard to believe though since I never seem to have a problem with interference while using either device.
 
Last edited:
I got this board after I got a DOA Asus Z68 V board, and I can't say I have any complaints. I didn't go for the 5GHZ overclock but it runs just fine at 4.5 and I haven't had a blue screen since building it back in November. I have had a couple of crashes that were SSD related because I was being lazy and neglected to update the firmware, but since doing so that problem no longer persists. I think this is a stable board, and MSI is one of the more reputable companies out there in terms of their warranties and replacements. Anyone who says ASUS is a better company obviously hasn't had an experience dealing with their customer service before. It's f--king awful and is why I will no longer deal with them on motherboards or monitors.
 
Nickbaldwin86 said:
nicely done. 1.48v is avg both my 2600Ks need that in order to do 5Ghz :(

it is ok for benching but I don't see the need for gaming or daily usage @ that vcore/Ghz
Click to expand...



Working on lowering it. At 1.48 my water cooling loop keeps my 2500k in low 60Cs during prime95.

I read some where that only about 2% of all 2500ks will reach 50+ multipliers so I am happy with the performance. Just need to lower voltage as much as possible.

****update. 1.48v stable prime95 was the best i could do with my 2500k. Worked on it for about 8 hours of a few days off and on whenever I had personal time. I feel confident about this MB abilities and look forward to improved BIOS updates. I5 K series IB is going in my MB ASAP.****
 
Last edited:
I'm done with 5GHZ OC with this 2500k. Intel Burn Test didn't cooperate but I was able to get in a solid hour of Prime95 with 1.48v under Prim95 load as reported by CPUZ. In the BIOS setting I set it to 1.510v but it shows an actual voltage of 1.48v.


The voltage check points read at 1.5v on the CPU voltage.

5GHZvoltage.jpg



Having fun with this MB.
 
I'm seeing some strange behavior on this motherboard. If i stress test (with Prime95 at stock speeds/voltages w/ a 2600k), USB port (closest to PS2) will see poor performance to the point that an input device becomes unusable. If I plug in the mouse, the mouse will work intermittently; if I plug in a keyboard the input will delay and sometimes completely omit keystrokes. Is this just a bad USB port? Is this just a fluke or does it happen to any else? It does not happen on any other USB port except this one.
 
Update from my experience going from 2500k to 3570k with this MB.


MSI came through with a IB ready BIOS a week before release of IB and I was able to update to it with the 2500k in place. A few days later after release of IB 3570k all I had to do was drop the cpu in the socket and then boot into Windows 7 and W7 did an update for the chipset and then reboot. All went well. MSI did a great job with making the swap over from 2500k to 3570k. :)


Everything is working great with this MB and I have been able to get some very nice performance with the MSI Z68A GD65 G3 with Intel i5-3570k.


Here is some of the success I have had with this MB and IB so far.

Here is 3dMark06 with 3570k at 4.8GHZ

http://3dmark.com/3dm06/16640459


5GHZ benchmarks
Super Pi 32m and MaxxMem2

5ghzmemtest3570k5ghz.jpg



24/7 usage stability testing for 4.5GHZ over clock.

LinX

LinX results with ambient temps 26C.

http://valid.canardpc.com/show_oc.php?id=2361977
LinX4point53570kmax4gb.jpg



This is with the latest version of LinX and with the newest Intel linpack binary updated.

I have had alot of fun with this MB and my experience with it has been great so far.
 
Thanks for the update Sonda. Do you have any way to verify if the PCIe 3.0 works with the 3570k? MSI was bragging about it when they released this board so I hope that it does indeed work.
 
Astral Abyss said:
Thanks for the update Sonda. Do you have any way to verify if the PCIe 3.0 works with the 3570k? MSI was bragging about it when they released this board so I hope that it does indeed work.
Click to expand...

Just saw this. Sorry for the late response.

PCI 3.0 does work with this MB with an IB CPU.

Here is a GPU-Z verification that I just ran,

http://www.techpowerup.com/gpuz/nv7f3/


Also one of the nice features about this MB and PCI 3.0 is that this MB is hard built to run PCI 3.0 in either PCI 3.x16 slots. All metal connecting pins in the PCI 3.0x16 lanes are present in both slots.

Most other Z77 MB that are out today only have all the connector pins in the first PCI slot. Only a few of the most expensive Z77 MB have the metal pin connectors in the more than 1 PCI 3.0 lane. Though there is no real benefit to this in terms of multi video card peformance due to the limitations of the IB cpu the only benefit is in placement of PCI 3.0x16 with option to place it in in first or second lane and still run full PCI 3.0x16.


Here is a 3dMark11 score with 2xHD7950s in Xfire that I ran in with this MB.
http://wwwww.hwbot.org/submission/2...___performance_2x_radeon_hd_7950_15081_marks/


http://www.3dmark.com/3dm11/3649832

800205.jpg


800398.jpg
 
You must log in or register to reply here.
Back
Top