[Zxian]

Intel Core 2 Q6600 G0 Stepping: Cheap Quad Core Just Got Better

The title is a little misleading, but in the end, the biggest advantage to teh G0 stepping is bragging rights. The added performance from the slightly higher overclock is marginal at best.

[puntoMX]

Lower voltage and thus less power eating, smaller cooler and less noise .

However, any one would pick the G0 over a B3, except for the ones who don’t know it or don’t care about a better “model”.

[Zxian]

Sure - lower power consumption for the same performance is always good, but the argument that it would give "substantially" better overclocking performance goes out the window.

Don't get me wrong - I just ordered one of these puppies. Hopefully I'll get the G0, since I'd like to make this machine as quiet as possible. I'll probably end up under-volting the CPU anyways.

[ripken204]

so now you decide to get one
i've been running it for a day and it's nice. hav'nt gamed on it yet, to busy reinstalling everything..
most things seem faster tho, im not sure if it's the ram or the cpu but i like it alot.
i have the feature enabled that automatically adjust the muliplier+voltage and that helps ALOT with temps.
i'm idling at 34C right now with stock cooling 1600MHz.

[Zxian]

@ripken - I decided to go with a quad core, because I got a nice little bonus from work yesterday, and NCIX has them on sale... $289 CAD. Can't really complain. I never said the quad-cores were bad - just the hype about the "better overclocking" from the G0 isn't as much as people said it would be. Sure you sqeeze out a few more Mhz. What does that get you? Some bragging rights, but not much else.

The "feature" you're talking about has been with every post-P4 Intel consumer-line CPU (i.e. not-Xeons). It's called SpeedStep.

I disable that on my desktops, since I'm usually running R@H anyways, and the system never gets a chance to slow down. On my laptop though - boy does it ever work. At minimum speed (800MHz), the CPU runs at 0.7V, and has a temperature of 26C.

Question about your temps - are you using SpeedFan, or TAT to measure those? TAT takes the internal temperature diode on each of the cores, instead of an external temperature reading (inside the CPU vs on the CPU).

[ripken204]

well speedstep is nice. amd sucked with their method.

the temp im getting that from is the same in mbm5 and the asus software. but once i start gaming that temp does soar, especially with my stock cooling..

TAT doesnt work for me.. the second i open it up it just closes.

edit: just tried core temp, 2 of the cores report the same temp and the other 2 are like 8C higher.. do you know which core is on which side of the processor? i might have to remount the heatsink

This post has been edited by ripken204: 23 August 2007 - 02:56 PM

[Zxian]

Right, you're running Vista... from what I understand, TAT is an XP-only program.

With the heatsink - it's pretty hard to mount the stock heatsink incorrectly. You installed it by pressing the diagonal pins first, right?

1  3
4  2

In that order... if you get what I mean. If you went around in a circle, then you did it wrong.

[nmX.Memnoch]

ripken204, on Aug 23 2007, 03:53 PM, said:

edit: just tried core temp, 2 of the cores report the same temp and the other 2 are like 8C higher.. do you know which core is on which side of the processor? i might have to remount the heatsink

I suspect that the two cores with the higher temp are probably doing the bulk of the work right now. And it's probably just one core from that particular die (remember, it's two dual-core dies in a single package). Since those two cores are on the same die, even one core doing some work will heat up both cores.

[ripken204]

Zxian, on Aug 23 2007, 06:23 PM, said:

Right, you're running Vista... from what I understand, TAT is an XP-only program.

With the heatsink - it's pretty hard to mount the stock heatsink incorrectly. You installed it by pressing the diagonal pins first, right?

1  3
4  2

In that order... if you get what I mean. If you went around in a circle, then you did it wrong.

i know how to mount it, i meant more about the thermal paste application method. the arctic silver website just says to put a line of paste across.

nmX.Memnoch, on Aug 23 2007, 06:50 PM, said:

ripken204, on Aug 23 2007, 03:53 PM, said:

edit: just tried core temp, 2 of the cores report the same temp and the other 2 are like 8C higher.. do you know which core is on which side of the processor? i might have to remount the heatsink

I suspect that the two cores with the higher temp are probably doing the bulk of the work right now. And it's probably just one core from that particular die (remember, it's two dual-core dies in a single package). Since those two cores are on the same die, even one core doing some work will heat up both cores.

ya its definatly 2 cores from the same die since they are core 0 and 1. but to see like 8C difference all the time, even at idle..

[Zxian]

If the 8C difference is there regardless of load, then I'd be more willing to place the difference as an offset than an actual temperature difference.

Try running varying numbers of CPUburn, and set the affinity of each burning process to a single core. See if you can find any anomalies with the different combinations.

[nmX.Memnoch]

ripken204, on Aug 23 2007, 08:22 PM, said:

i meant more about the thermal paste application method. the arctic silver website just says to put a line of paste across.

It doesn't take much. But once I put the line across I use a razor blade to spread it evenly across the heatspreader. It should be spread pretty thin...thin enough that you can almost read through it. Again, it doesn't take much at all.

If you just relied on the heatsink to spread it for you then you may not have even coverage.

This post has been edited by nmX.Memnoch: 23 August 2007 - 07:52 PM

[Zxian]

Yup - you want the paste as thin as possible. The thinner the better. If you remove your CPU and you've got a noticable thickness on the IHS, then you've got too much paste.

[ripken204]

nmX.Memnoch, on Aug 23 2007, 09:51 PM, said:

ripken204, on Aug 23 2007, 08:22 PM, said:

i meant more about the thermal paste application method. the arctic silver website just says to put a line of paste across.

It doesn't take much. But once I put the line across I use a razor blade to spread it evenly across the heatspreader. It should be spread pretty thin...thin enough that you can almost read through it. Again, it doesn't take much at all.

If you just relied on the heatsink to spread it for you then you may not have even coverage.

ya that's what i'm thinking. i just figured that arctic silver knew what they were talking about.. i definatly put on enough paste(hopefully not too much) i just hope that it spread correctly. i wont be doing any OCing, some gaming. in a couple weeks i will have better cooling so i will just see what's up then.

[nmX.Memnoch]

Take a really good look at it. If you have any areas that are thicker than others that means the thinner areas aren't getting good contact...and will run warmer. 8C is a pretty good jump from one side of a CPU to the other, especially when the heat spreader is only about 1.25"x1.25" (the CPU dies are much closer together under the heat spreader than that).

The reason for spreading it yourself is so that you make sure you get a nice even, thin spread across the heatspreader.