Silver Arrow performance with 2500k

Tephnos · 13 Jul 2011 at 20:27

Hey all, just got my Silver Arrow delivered from OCUK and set it up today, and was wanting to know how my temps compare with others.

CPU is at 4.5Ghz @ 1.304v

Idle temps are around 30c-33c

Load temps are between 50c-55c

Room temp is around 24c if not a bit higher.

How am I doing guys?

(Not too sure on how good I applied the TIM, hate doing that. If you think it's worth redoing, say so. I applied a small rice grain sized amount to both the heatsink and the cpu and then just squashed it down)

Edit: Actually, will probably re apply later tonight. Core 0 is on average 3-4c cooler than all the other cores which are consistent. Hmm

Moothead2 · 13 Jul 2011 at 20:29

Looks perfect to me. Getting mine on Friday

.

WingZero30 · 13 Jul 2011 at 20:36

Is your 'load' prime?

Try running IBT and see what kind of load temps you get?

Plec · 13 Jul 2011 at 20:50

Those temps are fine/great @ 1.3V+.

If you're running small FFTs with Prime (even if it's blend) then those are solid load temps at 1.3V+ - as the continuous stress will be greater than any program/app is going to apply.

EDIT: Howevere, if the load temps are via apps/gaming then run either prime/IBT/ORTHOS etc to get an optimal load temp.

Tephnos · 13 Jul 2011 at 20:50

The IBT set to maximum put me at 55-64ish.

Still good? :confused:

And damn, Core 0 had about a 7c difference compared to 1 & 2. Will have to resit for sure I think. Dammit.

And I got my original load temps running blend mode in Prime, yeah.

WingZero30 · 13 Jul 2011 at 20:51

LC155 said:
The IBT set to maximum put me at 55-64ish.

Still good?

And damn, Core 0 had about a 7c difference compared to 1 & 2. Will have to resit for sure I think. Dammit.

And I got my original load temps running blend mode in Prime, yeah.

Place silver arrow on a flat table and give a slight push sideways. Does it wobble?

Tephnos · 13 Jul 2011 at 20:53

WingZero30 said:
Place silver arrow on a flat table and give a slight push sideways. Does it wobble?

I already checked that whole concave thing with a ruler before I set it up, it seemed flat enough. There was a tiny bit of concave showing but nothing major, very small.

munz · 13 Jul 2011 at 20:54

very nice load temps,

dont worry too much about temp differences at load mine have difference of about 5-7C.

Would be very very very strange to see all 4 cores with exactly same temp under load.

Plec · 13 Jul 2011 at 20:59

LC155 said:
The IBT set to maximum put me at 55-64ish.

Personally, IBT is brutal - it's great quick stress test but doesn't resemble everyday load temps.

LC155 said:
And damn, Core 0 had about a 7c difference compared to 1 & 2. Will have to resit for sure I think. Dammit.

Not a problem i've set up 4 SB rigs and two chips have had discrepencies of (+-) 9degrees (it's a sensor problem across the cores) with one or more cores. The first chip i had i reseated the HS 4 times to be sure - the other chip only twice. The load temps should level out the differences between the cores?

LC155 said:
And I got my original load temps running blend mode in Prime, yeah.

Those temps are fine - try small FFTs if you want to give the CPU greater stress but no app/game is going to test it greater than a blend test.

EDIT: How did you apply the TIM?

Tephnos · 13 Jul 2011 at 21:02

Plec said:
Not a problem i've set up 4 SB rigs and two chips have had discrepencies of (+-) 9degrees (it's a sensor problem across the cores) with one or more cores. The first chip i had i reseated the HS 4 times to be sure - the other chip only twice. The load temps should level out the differences between the cores?
EDIT: How did you apply the TIM?

Load temps level out the difference between the cores?

Looking at idle, they're within 2-3c. At load the sensors/temps get to around 7c difference.

Edit: I applied the TIM by placing a small rice grain sized dot on the heatsink, and another on the cpu. Then I squished down and tightened the thing down (This was very, very hard in a FT02 with a 6950 card in the damn way, so I may have bumped it marginally a few times. Nothing to worry about right? Having to screwdriver those side screws down instead of using the tool supplied made it a pain.)

Plec · 13 Jul 2011 at 21:04

LC155 said:
Looking at idle, they're within 2-3c. At load the sensors/temps get to around 7c difference.

Is it core 0 that's the problem?

Tephnos · 13 Jul 2011 at 21:05

Plec said:
Is it core 0 that's the problem?

Core 0 and 3 are the lowest temperatures. 1 and 2 are the highest.

Plec · 13 Jul 2011 at 21:06

LC155 said:
Edit: I applied the TIM by placing a small rice grain sized dot on the heatsink, and another on the cpu. Then I squished down and tightened the thing down (This was very, very hard in a FT02 with a 6950 card in the damn way, so I may have bumped it marginally a few times. Nothing to worry about right?)

You only need to apply the blob to the CPU and the place the HS on top - but, no, it shouldn't be a problem.

Were they defininetly rice size - as two pea size blobs would be to much.

Tephnos · 13 Jul 2011 at 21:08

Plec said:
You only need to apply the blob to the CPU and the place the HS on top - but, no, it shouldn't be a problem.

Were they defininetly rice size - as two pea size blobs would be to much.

I definitely think so, as I thought to myself that I was applying too little, not too much.

But I do have some spare iso alchol and the TIM that came with the SA, I could resit later if you think that's a good idea.

(Hopefully I don't slice my middle finger on the heatsink when doing it again - that was painful hahaha)

Plec · 13 Jul 2011 at 21:09

LC155 said:
Core 0 and 3 are the lowest temperatures. 1 and 2 are the highest.

Same as my present chip.

My present temps running prime 95 small FFTs:

Core 0 - 50
Core 1 - 57
Core 2 - 59
Core 3 - 51

And my HS and CPU fans are only spinning at 50% - but my Vcore is lower than yours.

This is the CPU i reaseated 4 times with another cooler when i first got it - and i had the same anomolies.

Tephnos · 13 Jul 2011 at 21:11

Plec said:
Same as my present chip.

My present temps running prime 95 small FFTs:

Core 0 - 50
Core 1 - 57
Core 2 - 59
Core 3 - 51

And my HS and CPU fans are only spinning at 50% - but my Vcore is lower than yours.

Yeah that sounds like mine then.

I think both my HS and CPU fans are spinning at 100 since it doesn't make too much noise. Big 200 case fans at the bottom are at 50% though.

And damn, that's a nice vcore. I never really tried to bring mine down to see how stable it could be.

Plec · 13 Jul 2011 at 21:12

LC155 said:
I definitely think so, as I thought to myself that I was applying too little, not too much.

Sounds fine.

You've got quote a high vCore for 4.5GHz which is why your temps will be higher than mine.

How have you set your clock?

Tephnos · 13 Jul 2011 at 21:13

Plec said:
Sounds fine.

You've got quote a high vCore for 4.5GHz which is why your temps will be higher than mine.

How have you set your clock?

Well when I was trying to see how stable I could make it prime95 would keep failing after like 6 hours, I got frustrated, shooting it up to 1.3 and called it a day. Will need to try lowering it sometime, or maybe try for a 4.8

Plec · 13 Jul 2011 at 21:13

LC155 said:
And damn, that's a nice vcore. I never really tried to bring mine down to see how stable it could be.

Experiment - perhaps even drop 100MHz - my vCore goes up dramatically (relatively) when i push past 4.5Ghz.

I wanted a compromise between a moderate clock and quiet/cool running.

Tephnos · 13 Jul 2011 at 21:36

Plec said:
Experiment - perhaps even drop 100MHz - my vCore goes up dramatically (relatively) when i push past 4.5Ghz.

I wanted a compromise between a moderate clock and quiet/cool running.

Dropping vcore to see what I can boot in at 4.5 so far set a 1.280 with CPU-Z reporting that it's drawing 1.264 whilst idle. Seems okay so far although temps didn't change much.