Good review. Thanks. I'd also like to thank you for reviewing Intel boards in this part (as well as the previous part). I've never understood why all review sites don't review Intel boards so they can have a baseline to compare other boards to. Anyway, the review was very interesting and highly useful.

One other point, AFAIK, Gigabyte has ALWAYS used Realtek chips for LAN connectivity.

FLUCTUATING VCORE VOLTAGE?

I noticed that the vcore on the Sandy Bridge i5-2500K fluctuated between 0.8 to 1.2 v. See the voltage chart at the bottom of:
http://www.silentpcreview.com/article1148-page4.html

I have a Clarkdale i5-650 running on an Asus P7H55-M Pro motherboard. My vcore voltage is fluctuating between .096v and 1.2 volts (Correction 0.96v and 1.2 volts) (as measured by speedfan and also by CPUID). (12V, 5v, and 3.3v are rock steady and I am using a Nexus Value PSU)

I have been troubled by these vcore fluctuations on my Clarkdale. My e8200 775 chip has a stable vcore.

QUESTIONS

Apparently these voltage fluctuations are acceptable for the Sandy Bridge. Are they also acceptable for the Clarkdale?

Is my concern over this misplaced?

Does anybody have an explaination for these fluctuations? What is happening and why? Won't these fluctuations interfere with overclocking?

It is totally normal for a modern CPU to vary its operating voltage (and clock speed of cores, independently in many cases) dynamically in accordance w/ load demand. This is precisely how the CPUs have such improved energy efficiency these days.

Great review - thanks!

With the proliferation of >2.19TB drives, I'm suprised Gigayte didn't release with UEFI.

Well, they say :


So maybe they hacked some small UEFI part on top of their BIOS.

Thanks for the review.

The timing couldn't be better too.
I'm waiting for parts to be available. Its gonna be i7 2600K and Intel DP67BG. This time I'll post pics in the gallery.

That's interesting. I notice on mine that the speed of the chip is not varying as the vCore jumps back and forth from 0.96v and 1.2 volts.

So it isn't the speed that is causing the change. Is it that portions of the circuits are turned on and shut off. Is that the cause? Or is there some other mechanism that is causing the increased need for voltage followed by the decrease?

Anyone out there that understands this?

And where do you watch your CPU speed ? In CPU-Z ? Then the reason is simple - CPU-Z can show only 1 value, but every core in Lynnfield and Sandy Bridge can have it's own clock. One core can run at 3.2GHz, while the rest idle at minimal clock. What clock should CPU-Z show ?

So yes, it's the CPU clock which requires the voltage change.

For example look at this image from a Linux monitoring utility : http://img.techpowerup.org/100131/52percent.jpg . Does this help ?

I am using CPU-Z and speedfan, which both have the limitation you mention. But I am using them in conjunction with Windows Taskmaster which shows the load to be spread out fairly evenly among the 4 CPUs.

A swing from 0.9v to 1.2v is a pretty big swing. The swings, as measured by speefan are occurring about once every 5 to 15 seconds (speedfan appears to have a 5 second sampling rate), though Windows Taskmaster, which I am also using appears to have a sampling rate of one second.

It just doesn't seem that the clock speed can account for this... at least if the clock speed is related to the load as measured by Windows Taskmaster. Now it could be that this is occuring at a far faster rate than the Speedfan 5 second sampling rate... but it just doesn't seem like that explains the graphs I am seeing with Speedfan and Windows Taskmaster.

Anyone else out there with a recent generation Intel? What are you seeing on your vCore?

I was thinking about getting a 2600K specifically for overclocking. If 1.2v is a normal paek vcore... what is a normal peak vcore if you are overclocking a 2600K?

Any other thoughts?

Boot Times

Maybe I am missing it, but does the review talk about boot times?

You mad me smile.

Core clock change occurs way more often than on a second level. It occurs every few miliseconds, as clock and voltage change depends on the load itself. 0.9V is the idle voltage, 1.2V is most times the load voltage. So yes, it's normal.

If you want to know what frequency your CPU runs at, download and use this : http://www.cpuid.com/softwares/tmonitor.html

Thanks

I notice you have screenshots of Speedfan 4.42 in this review. I also have an i5-2500K (on ASUS p8p67 pro) and Speedfan 4.42 gave incorrect readings of temperatures and voltages. This was pointed out to me by another poster in the forum. The beta version gave correct readings. Just something to consider.

FYI, the link in the first post is incorrect/broken. It should be http://www.silentpcreview.com/intel-gigabyte-p67-motherboards.

Just a heads up: at a glance it looks like you are saying the Intel board costs $10 USD! This is actually the price difference between this and the Gigabyte board, but is bold in standard SPCR this-is-the-price fashion.
Edit: I mean on the first page, in the first paragraph about the Intel board.

Nice review!

Can the Intel board's UEFI be used to change fan speeds according to temperature readings? Thus eliminating the need for SpeedFan?

I missed that too, the one thing that excites me about UEFI


Love the power consumption charts, though

As we mentioned in the article, only the CPU fan changed speed in response to temperature.

As for boot times...

Power button to "Starting Windows" screen:
P67A-UD4: 18.0 secs
DP67BG: 25.6 secs
DP67BG with Hyperboot: 24.2 secs

For some reason the Intel board always restarts itself before the POST screen on a cold boot, eating up 7 seconds - that accounts for most of the difference.

Can anyone confirm that the Gigabyte supports Dolby Digital Live real-time 5.1 encoding of game audio over the SPDIF port?

Also, can anyone confirm that the intel does NOT do this?

(If you're thinking "optical always supports Dolby 5.1, right?" -- not true! Most motherboard sound support can only pass through pre-encoded 5.1 audio from DVDs/blu-rays through the SPDIF digital ports, and only sends 2-channel stereo for games.)

That was unexpected. UEFI is supposed to be faster. Granted boots speed will vary from board to board... but I was under the impression that UEFI was such a game changer that any UEFI board would be faster than any hard drive (like SSDs compared to HDs).

So what is the big deal with UEFI... that you get to use a mouse??????????

My DP67BG boot times are faster. Power button to "Starting Windows" screen is 19 seconds. With Hyperboot enabled, I get 14 seconds.

Also, I am not getting any cold boot "second restart" issue.

EDIT: I have updated my BIOS for the motherboard and upgraded the firmware on my Intel SSD. Hyperboot is now giving me a speed of only 8 seconds.

This was another of the many excellent and helpful reviews that keep me coming back to SPCR again and again. Thank you.

To avoid confusing people who only research hardware every year or two, like yours truly, it would be good to edit the date at the very start of the article to read 2011 -- it currently reads "January 16, 2010 by Lawrence Lee".