H67 vs P67 chipset

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Pondlife
Posts: 5
Joined: Wed Mar 10, 2010 5:10 am
Location: UK

H67 vs P67 chipset

Post by Pondlife » Thu Feb 17, 2011 5:10 am

I'm planning to build a new PC using sandy bridge when the motherboards start shipping again.

I won't be overclocking it, and am not a big gamer, so I'm thinking to go for the H67 chipset and use the integrated video, and maybe add a video card in a few years as a mid-life update (I tend to upgrade RAM, hard disk and video card after two to three years).

The question is: if I add a graphics card to a H67 motherboard, would the GPU on the CPU still be active and end up using more power than the same setup with a P67 mobo?

I'll probably use an Asus motherboard if that makes any difference.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Thu Feb 17, 2011 5:22 am

Pondlife wrote:if I add a graphics card to a H67 motherboard, would the GPU on the CPU still be active and end up using more power than the same setup with a P67 mobo?

AFAIK/IIRC (there are five SPCR's reviews about) no.

tim851
Posts: 543
Joined: Wed Aug 13, 2008 11:45 am
Location: 128.0.0.1

Re: H67 vs P67 chipset

Post by tim851 » Thu Feb 17, 2011 5:28 am

Every review that has compared H67/P67 in power consumption comes to the conclusion that with the same discrete graphics card used, the P67 actually draw more power.

ame
Posts: 488
Joined: Tue Jul 10, 2007 10:35 pm
Location: Israel

Re: H67 vs P67 chipset

Post by ame » Thu Feb 17, 2011 9:01 am

Once you plug in a dedicated GPU, it deactivates the integrated GPU. Since SNB is power gated I'm assuming it shuts down the GPU completely.

As the othe guys said H67 seems to draw slightly less power than P67.

Pondlife
Posts: 5
Joined: Wed Mar 10, 2010 5:10 am
Location: UK

Re: H67 vs P67 chipset

Post by Pondlife » Thu Feb 17, 2011 9:12 am

Thanks for the info. Looks like the H67 chipset will be the best one for me then.

wayner
Friend of SPCR
Posts: 190
Joined: Mon Apr 05, 2004 10:46 am
Location: Toronto, Ontario

Re: H67 vs P67 chipset

Post by wayner » Thu Feb 17, 2011 7:06 pm

ame wrote:Once you plug in a dedicated GPU, it deactivates the integrated GPU. Since SNB is power gated I'm assuming it shuts down the GPU completely.
Why does the integrated GPU shut down? I would be interested in using the integrated GPU and a discrete video card to drive 3 monitors but it looks like you can't do that - why is this?

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Thu Feb 17, 2011 7:24 pm

ame wrote:Once you plug in a dedicated GPU, it deactivates the integrated GPU. Since SNB is power gated I'm assuming it shuts down the GPU completely.
You base your statement on what ? Manuals for H67 boards say otherwise. For example ASUS P8H67-M PRO (manual, page 1-19, 31 of 74):
This motherboard also supports two add-on graphics cards and two onboard VGA output at the same time
This is exactly the opposite of your statement. You can use any of 2 of 3 onboard ports + put aditional 2 graphical cards in PCI-E slots and all of them will work.

ame
Posts: 488
Joined: Tue Jul 10, 2007 10:35 pm
Location: Israel

Re: H67 vs P67 chipset

Post by ame » Fri Feb 18, 2011 5:10 pm

faugusztin wrote:
ame wrote:Once you plug in a dedicated GPU, it deactivates the integrated GPU. Since SNB is power gated I'm assuming it shuts down the GPU completely.
You base your statement on what ? Manuals for H67 boards say otherwise. For example ASUS P8H67-M PRO (manual, page 1-19, 31 of 74):
This motherboard also supports two add-on graphics cards and two onboard VGA output at the same time
This is exactly the opposite of your statement. You can use any of 2 of 3 onboard ports + put aditional 2 graphical cards in PCI-E slots and all of them will work.

According to that quote from the Asus manual it does seem to be possible. The Asus H67 EVO has a slightly differnt manulal, but still seems to indicates that it is possible to run them at the same time, its not specifically mentioned as in the pro, but it reads you can choose what display you want to init first, suggesting there is a second one still active.

I based my statement on a gigabyte H67 motherboard I had for a few days until I replaced it with Intel P67. There was no option to have the integrated and PCIe active at the same time AFAIK. I remember it had integrated graphics options to Disable or Auto. Description said when set to auto it is disabled if PCIe is pluged in. I may be wrong on this, I only had it for a few days and I might have missed something.

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Sat Feb 19, 2011 3:09 am

Then Gigabyte screwed up this option (again). Some motherborads with integrated graphics cards have these problems sometimes.

Then the correct answer is : it depends, look it up in manual of your motherboard.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Sat Feb 19, 2011 3:29 am

faugusztin wrote:Then the correct answer is : it depends, look it up in manual of your motherboard.

While some other else seem even able to run even K-series CPU (the ASRock H67M GE/HT).

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Sat Feb 19, 2011 3:52 am

All H67 boards can run K-series CPU. But none of them can overclock it through multiplier change.

ces
Posts: 3395
Joined: Thu Feb 04, 2010 6:06 pm
Location: US

Re: H67 vs P67 chipset

Post by ces » Sat Feb 19, 2011 5:04 am

faugusztin wrote:Then the correct answer is : it depends, look it up in manual of your motherboard.
As a practical matter, I wouldn't trust how this works until you can get some confirmation from 2 or 3 people who have actually done it with the particular board you are interested in. There is so much BS documentation out there, even when it isn't in Chinglish, it just isn't reliable nor complete. It is just a general starting point. Trust no one or no one thing.

Also, if it suits your needs, you might consider waiting for the Z68 board. Which is a micro-ATX that is capable of overclocking.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Sat Feb 19, 2011 6:12 am

faugusztin wrote:But none of them can overclock it through multiplier change.

AFAIK/IIRC ASRock claims its board is able to do so, and this is exactly what I said, and what you have not understood.

http://en.ocworkbench.com/tech/asrock-h ... review/13/

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Sat Feb 19, 2011 6:22 am

quest_for_silence wrote:AFAIK/IIRC ASRock claims its board is able to do so, and this is exactly what I said, and what you have not understood.
Claim where ? There is none of that on product page, none on the manual. If you mean "- Supports K-Series unlocked CPU", then you overestimate the importance of this line. It means just that - you can use them. But can't use the unlocked multiplier.

Edit: If you meant that "Thecpu multiplier can’t go further than 37x (using ASRock oc tool)." line - i'm pretty much sure any H67 board can do the x37 multiplier, either via Turbo or manual settings. But that is not what K-series support would really mean, and neither could be considerd OC (as you can reach those speeds by simply leaving CPU at base clock and letting Turbo work). K-series support and overclocking means 57x multiplier support.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Sat Feb 19, 2011 6:58 am

faugusztin wrote:Claim where ? There is none of that on product page, none on the manual.

As you should know, pages 6 and 46 of the product manual (besides the product webpage).
faugusztin wrote:If you mean "- Supports K-Series unlocked CPU", then you overestimate the importance of this line. It means just that - you can use them. But can't use the unlocked multiplier.

Your one is a groundless state, and being rash even a bit silly according to me.
Even if marketing is often an obscure art, who are you to invert in the opposite direction what is the literal meaning of those words?
Moreover I'm not aware of any Intel technical document which explicitly states that H67 have sort of "locked CPU multiplier".

However, to answer to your hasty "Claim where?", now that you have quoted them, you have also enough official claimings.
faugusztin wrote:Edit: If you meant that "Thecpu multiplier can’t go further than 37x (using ASRock oc tool)." line - i'm pretty much sure any H67 board can do the x37 multiplier, either via Turbo or manual settings.

OCW explicitly state of OVERCLOCKING ON AIR, not Turbo, or whatsoever.
Again, who are you to invert in the opposite direction what is the literal meaning of their words?

As it's fully OT our talk about it ends here and now.

tim851
Posts: 543
Joined: Wed Aug 13, 2008 11:45 am
Location: 128.0.0.1

Re: H67 vs P67 chipset

Post by tim851 » Sat Feb 19, 2011 9:46 am

quest_for_silence wrote:and this is exactly what I said, and what you have not understood.
Dude, no offense, you are from Italy and spelling/grammar mistakes are excused by default, but you said
While some other else seem even able to run even K-series CPU (the ASRock H67M GE/HT).
which is very a) hard to understand and b) does not imply at all that the ASRock board can overclock a K-Series CPU.

NTNgod
Posts: 33
Joined: Sat Feb 23, 2008 11:09 am
Location: WI, USA

Re: H67 vs P67 chipset

Post by NTNgod » Sat Feb 19, 2011 6:17 pm

quest_for_silence wrote:
faugusztin wrote:If you mean "- Supports K-Series unlocked CPU", then you overestimate the importance of this line. It means just that - you can use them. But can't use the unlocked multiplier.

Your one is a groundless state, and being rash even a bit silly according to me.
Even if marketing is often an obscure art, who are you to invert in the opposite direction what is the literal meaning of those words?
No, he's right - it means they'll work and nothing more. Now, "supports" in English could be interpreted several different ways (your interpretation is undoubtably the one the marketing people hope potential buyers use), but the ASRock H67 boards don't do anything differently than anyone else's, whether you use the BIOS or the ASRockAXTU program.

On a H67 board, a 2500K can Turbo up to 3.7 if only stressing one core, and Turbo to 3.4 if all four cores are stressed. The nominal 3.3 speed (as long as cooling is sufficent) is never seen on my system, even with all four cores stressed - it's always Turbo-ed to 3.4 under load.

I've set it to 3.7 using BIOS/AXTU, and it'll display as such - but then it'll immediately drop back down to 3.4 under full load (where temps top out no higher than ~41C), before bouncing back up to 3.7 when done.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Sun Feb 20, 2011 2:06 am

tim851 wrote:
quest_for_silence wrote:and this is exactly what I said, and what you have not understood.
Dude, no offense, you are from Italy and spelling/grammar mistakes are excused by default, but you said
While some other else seem even able to run even K-series CPU (the ASRock H67M GE/HT).
which is very a) hard to understand and b) does not imply at all that the ASRock board can overclock a K-Series CPU.

Well Tim-dude :mrgreen: this isn't the very first time I talk about it here on SPCR, but still I don't understand why you and several other people don't do the simple math "run a K CPU" = "unlock the multiplier" and nothing more (please forget that Intel IGP). Not to mention that at last even some manufacturer seem to actually speculate about this implicit equation.
But as things are so, well, I will take into account that people have to be pedantic on this matter.
Whether or not ASRock actually overclocks has never been and isn't my concern (I've just said that with reference to product manual, "some board seems able/some one claims to be able", I did not say that some H67 can actually do it: I hope to have addressed both your point :wink:).

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Sun Feb 20, 2011 2:14 am

NTNgod wrote:No, he's right - it means they'll work and nothing more. Now, "supports" in English could be interpreted several different ways (your interpretation is undoubtably the one the marketing people hope potential buyers use), but the ASRock H67 boards don't do anything differently than anyone else's, whether you use the BIOS or the ASRockAXTU program.

Thanks for definitely clarifying this point (about ASRock and OCW states), giving faugusztin credits for being right: your hands on experience helps a lot (even in the way ces pointed out some posts ago, talking about documentation reliability).

ame
Posts: 488
Joined: Tue Jul 10, 2007 10:35 pm
Location: Israel

Re: H67 vs P67 chipset

Post by ame » Sun Feb 20, 2011 3:41 am

quest_for_silence wrote:
NTNgod wrote:No, he's right - it means they'll work and nothing more. Now, "supports" in English could be interpreted several different ways (your interpretation is undoubtably the one the marketing people hope potential buyers use), but the ASRock H67 boards don't do anything differently than anyone else's, whether you use the BIOS or the ASRockAXTU program.

Thanks for definitely clarifying this point (about ASRock and OCW states), giving faugusztin credits for being right: your hands on experience helps a lot (even in the way ces pointed out some posts ago, talking about documentation reliability).
So its settled then? H67 can NOT OC regardless of manufacturer or CPU type. It is what NTNgod said after all.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Sun Feb 20, 2011 4:09 am

ame wrote:So its settled then? H67 can NOT OC regardless of manufacturer or CPU type. It is what NTNgod said after all.

I'm not sure to have well understood you, ame: at anyway, NTN said "I have this board and the 2500K, and it does not OC despite ASRock and OCW words" (so about the same said by faugusztin), and this is why I thanks him, to have somehow definitely cleared ASRock manual's opacity (and OCW claims). As usual there's no free lunch.

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Sun Feb 20, 2011 4:17 am

ame wrote:So its settled then? H67 can NOT OC regardless of manufacturer or CPU type. It is what NTNgod said after all.
Exactly. Even if some H67 board offers you the control for multiplier, you are still limited by maximum multipler of a non-K version of the same CPU. It is useless for you that 2500K and 2600K has x57 max unlocked multiplier, when you are limited to x33/x37 with 2500K and x34/x38 with 2600K. With that ASRock board, you can set x37/x38 as fixed multiplier, but it is pretty much a pointless feature as you reach those multipliers with standard CPU function anyway. If it can fix the 3.7/3.8GHz for all 4 cores, it gives you a slight overclock, but nothing extraordinary and you don't need a K-type CPU for that, as you will get same "overclocking posibility" with 2500 and 2600, without K. In H67 boards 2500K/2600K=2500/2600 with only one big difference - the difference in integrated graphics.

ame
Posts: 488
Joined: Tue Jul 10, 2007 10:35 pm
Location: Israel

Re: H67 vs P67 chipset

Post by ame » Sun Feb 20, 2011 4:19 am

Yes :wink:. I said what you said Luca,




Any chance we can dismiss or confirm Asus's claim to have both iGPU and PCIe GPU active at the same time? Anyone with an Asus H67 motherboard out there want to give it a shot?

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Sun Feb 20, 2011 4:55 am

ame wrote:Any chance we can dismiss or confirm Asus's claim to have both iGPU and PCIe GPU active at the same time? Anyone with an Asus H67 motherboard out there want to give it a shot?
I will be able to confirm it for you in 1st week of March, if that is enough for you. I'm waiting for my 2600K to arrive from Germany (they had none in countries closer to me :( ).

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Mon Feb 21, 2011 6:59 am

faugusztin wrote:If it can fix the 3.7/3.8GHz for all 4 cores, it gives you a slight overclock

But it wouldn't be so, according to NTNgod (it seem still bounded to Turbo schemes, if I've well understood him).
faugusztin wrote:you will get same "overclocking posibility" with 2500 and 2600, without K.

It's questionable, as usually the BCLK seems to hit the wall around 103-105MHz (with reference to current web results).
However, as the fixed multiplier doesn't seem able to oc the four fores, it's doesn't matter at all.

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Mon Feb 21, 2011 7:23 am

quest_for_silence wrote:
faugusztin wrote:you will get same "overclocking posibility" with 2500 and 2600, without K.
It's questionable, as usually the BCLK seems to hit the wall around 103-105MHz (with reference to current web results).
However, as the fixed multiplier doesn't seem able to oc the four fores, it's doesn't matter at all.
No, you again misunderstood me :). I meant that for H67 boards with ability to set multiplier, there is no difference between 2500/2500K or 2600/2600K, as the K versions behave as non-K versions. IN case of 2500/2500 it is x33 standard multiplier and x37 maximum turbo multiplier (means you can set mannually maximum x37 multiplier in that Asrock board) or x34/x38 for 2600/2600K. With H67, you won't be able to go over that x37/x38 multiplier, even if you have 2500K/2600K with unlocked x57 multiplier - simply because the board will not recognize the CPU as unlocked and will behave as if you have the non-K version of CPU (except the iGPU).

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Mon Feb 21, 2011 12:27 pm

faugusztin wrote:
quest_for_silence wrote:
faugusztin wrote:you will get same "overclocking posibility" with 2500 and 2600, without K.
It's questionable, as usually the BCLK seems to hit the wall around 103-105MHz (with reference to current web results).
However, as the fixed multiplier doesn't seem able to oc the four fores, it's doesn't matter at all.
No, you again misunderstood me :). I meant that for H67 boards with ability to set multiplier, there is no difference between 2500/2500K or 2600/2600K, as the K versions behave as non-K versions. IN case of 2500/2500 it is x33 standard multiplier and x37 maximum turbo multiplier (means you can set mannually maximum x37 multiplier in that Asrock board) or x34/x38 for 2600/2600K. With H67, you won't be able to go over that x37/x38 multiplier, even if you have 2500K/2600K with unlocked x57 multiplier - simply because the board will not recognize the CPU as unlocked and will behave as if you have the non-K version of CPU (except the iGPU).

I think it's really difficult to understand each other in a third language (as I just think you :wink: misunderstood me): the way you have described this limited overclocking it is (rectius: it seems to me) the way Turbo does function.
With reference to Turbo mode, the limited overclocking is the just same for -K and not-K SKU CPUs.

Image

But according to Intel, the H67 would not seem able to perform the Turbo mode, with or without a -K SKU CPU:

Image

With this respect, setting a higher fixed multiplier should be doable just on -K SKU CPUs (but we know from NTN it isn't so), while, if in case, the not-K SKU CPUs were limited not to have even the "standard" Turbo mode (1/3 of the slight overclock above described).

To be fair, I've just (tried to) clarified my previous thought, but I'm not really deepening into how oc works in the "SB era" (actually I cannot going into this matter as I have no 1155 systems running), which is fairly OT here as I previously said, so my preliminary notions may well be inadequate to thoroughly comprehend how it operates (or may operate).

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Mon Feb 21, 2011 12:41 pm

Nope. H67 does support turbo, but in case of H67, there is only the green part from those pictures, there is no purple "Overclocked" state. You have the turbo, but no OC higher than maximum Turbo Freq. That is all.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: H67 vs P67 chipset

Post by quest_for_silence » Mon Feb 21, 2011 2:04 pm

faugusztin wrote:Nope. H67 does support turbo, but in case of H67, there is only the green part from those pictures, there is no purple "Overclocked" state. You have the turbo, but no OC higher than maximum Turbo Freq. That is all.

So, sort of "two stage turbo" only for P67s?

At anyway (even with the "green" Turbo ability), it remains unclear (to me) why, in a H67 system with selectable multiplier (just from a pure theoretical standpoint, as the actual so-called selectable multiplier doesn't work in that way), a not-K CPU should have gone up as high in frequency as a -K CPU, providing that only -K CPUs (should) have hardwired the ability to run at a fixed multiplier which is higher than the stock one.

faugusztin
Posts: 450
Joined: Mon Mar 29, 2010 2:47 am
Location: Bratislava, Slovak Republic

Re: H67 vs P67 chipset

Post by faugusztin » Mon Feb 21, 2011 4:28 pm

You base your asumption on false statement. You see a 3.3GHz and asume that i5 2500 has standard x33 multiplier. While it is theoretically true, technically it is not. i5 2500 has minimum multipler of 16, maximum of 37 and maximum unlocked of 41 that is it (for 2500K, it is maximum unlocked of 57). In case of H67, your multipliers are managed automatically between 16 and 37 (or 38 in case of 2600K), and if you have selectable multiplier in BIOS, then you can fix it at that multiplier i guess.

With P67 boards, this barrier goes away and for generic 2500 and 2600 CPU's you get +4 multipliers (x41), while with K version of CPUs you get a maximum of x57, which is pretty much impossible to reach.

That's why you don't see any difference in frequency between 2500 and 2500K or 2600 and 2600K - because the K means instead of x41 maximum unlocked multiplier they have x57 maximum unlocked multiplier.

Of course K versions bring better IGP, but that is another story.

Post Reply