Memory Limitations for GIGABYTE GA-P55-UD3R

[Aves]

I wasn't sure if I should post this in Memory or Motherboards so please excuse.

I'm looking to build a new Core i7 860 PC in the next couple of weeks and I've narrowed down all the components other than RAM.

I've settled on the GIGABYTE GA-P55-UD3R motherboard but I'm just a little confused as to what the RAM limitations are. Right now I'm only looking to add 4GB of RAM but I could see myself going to 8GB down the road and I want to ensure I don't buy anything I can't expand on later.

I've been eying the G.SKILL Ripjaws Series 4GB (2 x 2GB) DDR3 1600 (Model F3-12800CL7D-4GBRH) kit which apparently runs at 1.65v according to the specs on Newegg.


According to the manual for the motherboard it's limited to 4 x 1.5v DIMMs and it doesn't say it supports DDR3 1600

4 x 1.5V DDR3 DIMM sockets supporting up to 16 GB of system memory
Dual channel memory architecture
Support for DDR3 2200/1333/1066/800 MHz memory modules
Support for non-ECC memory modules

So my question is will this RAM even work in that board and if so would I be able to run 4 of these DIMMs for a total of 8GB?


Any help is appreciated, the last PC I built myself was in 2002 so I feel like a n00b again.

 

[n7]

I have that exact mobo & 8 GB Ripjaws DDR3-2000.

Max RAM multi is 10x, so @ stock (133 BCLK), you run DDR3-1333.
If you increase BCLK a little, you can do 1600 or much higher.

 

[RussianSensation]

Please see sig below.

I am pretty sure the 4 x 1.5V is just the 'factory standard'. The motherboard allows you to increase the voltage far beyond 1.650V (which is the maximum safe specification for Core i5/i7 processors).

The G.Skills you linked are rated at 1.65V, however their X.M.P. Profile is actually only 1.60V. Just enable X.M.P. Profile1 on the UD3R with F3f bios and you'll automatically get the motherboard to set 1.60V @ CL7 timings.

In fact, on UD3R I am able to run them at CL 7-7-7-24 at only 1.56V (per bios) -> 1.55V in CPU-Z.

This kit will also do DDR3-1333 at 1.50V.

Also, if you plan on overclocking, UD3R will not reduce your voltages in idle at 4.0ghz. Mine reports 1710mhz (190x21) @ 1.280V at idle. Voltage power savings does not work in overclocked mode.

If this matters to you, get an Asus board, or something else.

 

[Aves]

Excellent. Thanks for the help. I'm really looking forward to the new rig; I've been running a 3-year old Pentium-D from Dell.

 

[vshin]

Originally posted by: RussianSensation

Also, if you plan on overclocking, UD3R will not reduce your voltages in idle at 4.0ghz. Mine reports 1710mhz (190x21) @ 1.280V at idle. Voltage power savings does not work in overclocked mode.

If this matters to you, get an Asus board, or something else.

Is this still true if you leave turbo mode on, as well as the the other power-saving settings in the bios that are usually turned off for higher overclocks? I was planning on modest overclocks to around 150 bclk with no power-saving features disabled, with the main goal of extracting performance while maintaining a very silent system at idle.

 

[RussianSensation]

Originally posted by: vshin

Originally posted by: RussianSensation

Also, if you plan on overclocking, UD3R will not reduce your voltages in idle at 4.0ghz. Mine reports 1710mhz (190x21) @ 1.280V at idle. Voltage power savings does not work in overclocked mode.

If this matters to you, get an Asus board, or something else.

Is this still true if you leave turbo mode on, as well as the the other power-saving settings in the bios that are usually turned off for higher overclocks? I was planning on modest overclocks to around 150 bclk with no power-saving features disabled, with the main goal of extracting performance while maintaining a very silent system at idle.

I just tested this for you:

155 Base Clock

Scenario 1: Leave CPU Voltage at Auto, set all power saving technologies (C1, EIST, etc.) to Enabled. Disable XMP Memory profile. Enable Turbo Mode. I get:

155 Base Clock x 9.0x multiplier = 1395.0 mhz in idle, voltage of 1.264V
155 Base Clock x 22.0x multiplier = 3410.0 mhz at load in WinRAR, voltage of 1.264V

Scenario 2: Same as above, but CPU Voltage manually fixed at 1.300V in the BIOS. I get:

155 Base Clock x 9.0x muliplier = 1395.0 mhz in idle, voltage of 1.280V
155 Base Clock x 22.0x multiplier = 3410.0 mhz at load in WinRAR, voltage of 1.280V.

Therefore, if you want voltage power savings in idle, you cannot consider the UD3R Gigabyte board at this time. In addition, here is a 3rd Party Overclocking Review for UD6. I quote:

"The funny thing is that on two different mainboards, with slightly different voltage settings (Asus mainboard required 1.29375 V processor Vcore), the peak voltage under maximum processor load was still 1.312 V. However, the situation changes dramatically, when the CPU is idle. Asus P7P55D Deluxe has all Intel processor power-saving technologies up and running that is why in idle mode not only the processor clock frequency multiplier but also the core voltage get lower. As for Gigabyte GA-P55-UD6, only the multiplier lowers, and the core voltage remains excessively high."

I recommend you get an Asus P7P55D for this feature. :thumbsup:

 

[vshin]

Thank you very much! This is a very important nuance that most casual overclockers (especially the silent pc enthusiasts) will want to know.

 

[RussianSensation]

If you are running a decent cooler, even having it the processor run at 1.4ghz @ 1.30V will not pose any problems. The main issue here is power consumption and pushing too much voltage than required, not silence imo. The processor idles at 30*C at 1.30V!

 

[imported_Lothar]

Originally posted by: RussianSensation
If you are running a decent cooler, even having it the processor run at 1.4ghz @ 1.30V will not pose any problems. The main issue here is power consumption and pushing too much voltage than required, not silence imo. The processor idles at 30*C at 1.30V!

The power consumption difference was only 7 watts.
Not something for someone to get worked up all over IMO.

 

[RussianSensation]

Originally posted by: Lothar

Originally posted by: RussianSensation
If you are running a decent cooler, even having it the processor run at 1.4ghz @ 1.30V will not pose any problems. The main issue here is power consumption and pushing too much voltage than required, not silence imo. The processor idles at 30*C at 1.30V!

The power consumption difference was only 7 watts.
Not something for someone to get worked up all over IMO.

You are saying the power consumption of running 1.4ghz HT @ 1.30V vs. 1.4ghz @ 0.864V would only be 7 watts? How did you arrive at this estimate?

The 7 watts difference you noted from above Review did not have Turbo Mode enabled which is what was asked by vshin:

"When we discussed different overclocking methods on Asus P7P55D Deluxe, we compared not only the performance, but also the power consumption with dynamic and static Turbo Boost technology implementation. It turned out that in idle mode the dynamic version is more energy-efficient offering almost 30 W advantage. The thing is that when we enabled C3-C7 states for dynamic implementation in the BIOS, we allow the CPU to go into deeper power-saving modes in case there is no load and thus disable more units. This way, if we compared the boards during overclocking using dynamic Turbo Boost, the power consumption difference between them in idle mode would be way greater and would exceed 30 W. In this case the CPU on Asus board could go into deeper power-saving states, unlike the CPU on Gigabyte board."

 

[lopri]

That's kind of a bummer. Is it a common issue on Gigabyte's P55 line-ups?

 

[imported_Lothar]

Originally posted by: lopri
That's kind of a bummer. Is it a common issue on Gigabyte's P55 line-ups?

Yes.

 

[Aves]

Just wanted to say thanks again to all for the help. Got my new PC up and running tonight with relative ease.

 

[imported_Lothar]

Originally posted by: RussianSensation

Originally posted by: Lothar

Originally posted by: RussianSensation
If you are running a decent cooler, even having it the processor run at 1.4ghz @ 1.30V will not pose any problems. The main issue here is power consumption and pushing too much voltage than required, not silence imo. The processor idles at 30*C at 1.30V!

The power consumption difference was only 7 watts.
Not something for someone to get worked up all over IMO.

You are saying the power consumption of running 1.4ghz HT @ 1.30V vs. 1.4ghz @ 0.864V would only be 7 watts? How did you arrive at this estimate?

The 7 watts difference you noted from above Review did not have Turbo Mode enabled which is what was asked by vshin:

"When we discussed different overclocking methods on Asus P7P55D Deluxe, we compared not only the performance, but also the power consumption with dynamic and static Turbo Boost technology implementation. It turned out that in idle mode the dynamic version is more energy-efficient offering almost 30 W advantage. The thing is that when we enabled C3-C7 states for dynamic implementation in the BIOS, we allow the CPU to go into deeper power-saving modes in case there is no load and thus disable more units. This way, if we compared the boards during overclocking using dynamic Turbo Boost, the power consumption difference between them in idle mode would be way greater and would exceed 30 W. In this case the CPU on Asus board could go into deeper power-saving states, unlike the CPU on Gigabyte board."

Interesting...
So there's a 30 watt difference with turbo mode enabled vs. only 7 watts with it disabled?
What could possibly be causing such a large difference?

It seems to me that they're making a prediction(meaning they didn't actually test that) from that underlined statement.
They just extrapolated that since the power savings in ASUS between static and dynamic overclocking was 30 watts, they expect that to be more when comparing to Gigabyte.

Who's to say that power saving technologies during static overclocking on the ASUS isn't broken or working as well as it should?
Are there C3-C7 states available on the ASUS during static overclocking?
If yes, why didn't they enable them during their test? If no, then I feel it's a very important bios feature missing on the ASUS(or possibly ALL boards!!!) that should be implemented.

Whether their prediction does indeed turn out to be correct or not, who knows?

 

[vshin]

Hi guys, just an update. Although I would have been happy with a mild overclock of around 150 bclk at stock voltages, my CPU managed to coast all the way to 166 without any problem at all. This is without increasing the voltage and enabling everything including Turbo, EIST, C1E, C3-7, etc.

It will run at 3.486 (x21) with 4 threads; 3.818 (x23) with 2 threads; 3.984 (x24) single thread. Temps idle at 27 C and peaks at 63 C with Prime95.

The vcore idles at 0.864 V and peaks at 1.280 V at full load. One thing I noticed is that you have to set the vcore to "normal" to keep stock values and get all the powersaving features to work. Leaving it on "auto" will increase the voltage on its own. I didn't change any other voltage settings. I did set the QPI ratio lower and reduced the memory ratio to keep the speed within spec. I also had to enable the advanced CPU feature since leaving it on auto would disable them for some reason.

So far I'm very happy.

 

[lothar]

Russian Sensation
Gigabyte claims they've fixed that feature in their latest BIOS. Are you able to test this fo us?

 

[RussianSensation]

Russian Sensation
Gigabyte claims they've fixed that feature in their latest BIOS. Are you able to test this fo us?

Sure, no problem let me test this for you with F3n bios.

 

[RussianSensation]

Major changes between F3f and F3n BIOS (released November 6, 2009) on Gigabyte P55-UD3R Rev 1.0:

Motherboard Intelligent Tweaker (M.I.T.):

1) Advanced Frequency Settings:

* CPU Clock Ratio = added 1 extra multiplier to all Core i5/i7 processors which can be accessed when Turbo Mode is Disabled (i.e. static overclocking mode). New available upper multipliers are as follows:

i5 750 = 21x (up from 20x)
i7 860 = 22x (up from 21x)
i7 860 = 23x (up from 22x)

* CPU Clock Drive increased to 800mV from 700mV (should help to improve stability)
* PCI Express Clock Drive increased to 900mV from 700mV.

2) Advanced Voltage Settings Tab

*Added Dynamic Vcore (DVID) setting = this allows dynamic voltage reduction of VCore between idle and overclocked modes. To activate, please follow these steps:

CPU VCore = set to Normal (from Auto or actual stable overclocked voltage, 1.3000V for me)
Dynamic VCore = select the additional extra voltage which is required to reach stability from stock voltage (1.19375V for me) to required overclocked voltage at specified overclocked frequency. In practice:

F3f BIOS:

CPU VCore = 1.3000V (unable to lower voltages in idle mode, therefore results in 1.280V in idle and 1.3000V at load)

F3n BIOS:

CPU Vcore = Normal (nominal voltage of 1.19375V)
Dynamic VCore (DVID) = +0.10625
Effective CPU Vcore at Load = still 1.3000V (now able to lower voltages in idle mode to 0.9600V)

*** Dynamic VCore does not work if CPU VCore is set to Auto or to an actual voltage number. Therefore, to get Dynamic VCore to work in idle mode when CPU is manually overclocked, the CPU Vcore MUST be set to Normal first *** (thanks vshin) Don't forget to set all Voltage Power Saving technologies (i.e., C3-C7 states, EIST, etc. ) to Enabled.

New Bios in action:
http://i128.photobucket.com/albums/...-UD3R/i7860186x21391ghzDynamicVCoreatidle.jpg

P.S. Mods, what happened to the advanced Reply action buttons (bold, italics, linking an http:// ????)

 

[n7]

Thanx for the update RussianSensation!

I am going to have to flash my BIOS again...good to see Gigabyte is updating their cheaper boards.

To change the message editor interface (which affects what options are there for new posts), go to User CP > Edit Options > Message Editor Interface; you should be able to change it.

 

[RussianSensation]

Thanx for the update RussianSensation!

I am going to have to flash my BIOS again...good to see Gigabyte is updating their cheaper boards.

To change the message editor interface (which affects what options are there for new posts), go to User CP > Edit Options > Message Editor Interface; you should be able to change it.

This is why I am on my 3rd Gigabyte board. The constant bios improvements keep making the board better and better. It's always a pleasant day when you get extra "free" features or additional overclocking headroom from a new BIOS.

The best part for future UD3R owners is that the "A" version introduces even more improvements (hopefully at the same price):

Major changes for GA-P55A-UD3R include:

* 12-phase power VRM design vs. 8-phase currently, for more optimum power efficiency and thermal solution
* NEC SuperSpeed USB 3.0 with transfer rates of up to 5 Gbps (10x speed)
* LOTES processor sockets which should have less burning issues than Foxconn ones.
* 3X USB power delivery for greater compatibility and extra power for USB devices
* Marvell SE9128 high speed SATA3 storage interface with superfast 6Gbps link speed
* Realtek ALC889 codec vs. current 888

 

[Howard]

Man, I just bought one. Bummer.

 

[Madhukar]

small query. If i have this motherboard, can i out in 2 * 8 gb ram dual channel or do i need to limit it to 4* 4 gb ram. (Since the limitation is 16 gb)

 

[Valantar]

Wow, that is one epic necro. Congratulations. When the thread is 8 years old, I'd say it's safe to create your own, brand-new thread. There's not that much chance that the original posters are still around to follow up.

As 8GB DIMMs are standard DDR3 fare, and not part of some extension or something like that, I'd say 2x8 should work. Then again, that board is a first-gen DDR3 product. Might not work. Have you tried googling it?