Yorkfield Q9450 vs Kentsfield Q6600 at 3.6Ghz

[Orbitech]

The results are disappointing as a matter of fact for me.. The big L2 cache and the less power/heat just don't do the trick for me. I'd have really preferred a bigger multiplier available for Q9450/Q9550, even if that would have meant bigger prices.
I think I'll go for the cheapo Q6600 as I should have done that a long time ago now.. Bad decision this time for me to hold..I should have upgraded to Q6600 for some time now and wait for Nehalem..

Guys I have a question because I haven't kept up much with the Q6600 GOs O/C.. (sry for the OT question I just don't wanna flood the place with new threads if I can avoid it)

I have a E6600 for a year and a half now and I can do 3.6GHz with 1.5v and around 55C idle temp with a TR Ultra xtreme 120.
If I'm to go with a Q6600 GO I have a major question with basically 3 subparts if you look more closely to it..

Is it highly possible to achieve a 24/7 3.6-3.8GHz O/C with this cooler (on high end air that is)?
Any good batch I should look for?
My W/C needs some tweaks and I'll keep it on air for a while so your feedback is really appreciated..
Thnx in advance

[AndrewZorn]

Glad I read this... I was actually just shopping around for Q9450 preorder prices. I have a B3 Q6600 in my machine right now, and a G0 Q6600 sitting here on the desk (that I got for really cheap, and was going to sell...).

I'm going to seriously consider opening this G0. I never really thought about it... to get the 4ghz I'd want (to justify the $100 difference in keeping the G0 and buying the Q9450) my board would have to do 500fsb 24/7. It's an MSI Neo2-FR which is really good for the money, but 500fsb, 24/7? Seems like I'm setting myself up for disaster.

Like dertimaushh said, I think that since I have water and a board that probably won't be doing 500fsb+ all day every day the G0 Q6600 is a better choice. Especially since I already have one AND it's the cheaper option.

[sssf]

JB, are you able to get measurements for the systems at the plug? I'd be interested in total power draw difference for the two setups. Total wattage under full load from a kill-a-watt or similar. Half the amp draw in software is impressive, but what are the savings at the plug?
I'm wondering how much of the juice savings is burned up in the higher fsb etc.
Basically.. how big is the KWH/day difference going to be?

I'm also quite interested in this because I'm going to run it on full load 24/7 (DC you know).
Even a $100 price difference when buying the CPU might not matter if I'm running it for at least 2 years.

Having the numbers enables the math

[Blasar]

For a moderate overclocking with "conventional" cooling (aka air) , let´s say 3600MHz tops, i think that the Q9450 is a better choice,

for higher flights the Q6600 might be the answer, but nothing is guaranteed.

If the prices of the Q9450 stays under 300€ my choice is made, if not Q6600 here i go

[Hymay]

I'm also quite interested in this because I'm going to run it on full load 24/7 (DC you know).
Even a $100 price difference when buying the CPU might not matter if I'm running it for at least 2 years.

Having the numbers enables the math

Exactly why I wanted to know the draw at the plug For me, even a 30% savings at the plug is only about 3-5 bucks a month... depending on your elec cost. So at 3 bucks.. the juice "savings" would take over 33 months to just match the $100 price difference, and even longer to see any benefit.
Meanwhile, your cpu runs cooler.. and may take less wear.. but your MB has to run hotter to maintain a decent OC.

[Cronos]

Intel is working on new 45nm revision for low-middle Yorkfields to fight with high-fsb instability. Also, there is new upcoming P45 chipset with native 1600FSB support.
Provided all goes well, this combination may prove successful for really stable 500FSB OC for 24/7 use.
But i think, we should also wait for the next price drop from Intel and opt for Q9550.

So, to summarize, Q6600 for now, or Q9550(new revision)+ P45 + price drop later.

[lazy]

hm.. still havent found any reason why a Q9450 is worth the change right now.. any more testings?

[Ace-a-Rue]

JB,

any chance of evening up the vcore for the Q6600 and Q9450 to see what the actual amps draw is for both processors in idle and then under a wPrime 1024M bench load?

[WRC]

i'm sure they'll get faster 45nm cpu's in later revision, now it's to soon to see which generation can achieve the highest clocks

[strange|ife]

I think the winner in the locked chips will be the E8600 if you want dualcore.

The new quads are nice, but those multi's just suck. 4.4ghz on E8600 is very possible for 24/7

[Kunaak]

I'd rather wait till they release a Q9650 in like 6 months.
the 8 and 8.5 multiplier are just too low to bother with.

I really liked my Q6600 but not enough to give up the potential behind a good E8400 or E8500.

[Nelly]

Hey Johnny Bravo,

Do you still have your Abit IP35 PRO? would love to see some results with that Q9450 in comparison to the Abit IX38 QuadGT. . . ! i.e especially FSB.

Thanks for sharing your results by the way.

I'll see what I can do mate

Hi Johnny Bravo, did you manage to test the Q9450 on the Abit IP35 PRO?

[adamsleath]

i still get the feeling that q6600's will be more popular when yorkfields turn up; however it will be interesting to see who and how many start spieling about yorkfields when they arrive at retail, and who/how many keep their q6600's - in the rarified atmosphere of the extreme.


new things seem to run like dominoes thru' this forum.

[douirc]

i still get the feeling that q6600's will be more popular when yorkfields turn up; however it will be interesting to see who and how many start spieling about yorkfields when they arrive at retail, and who/how many keep their q6600's - in the rarified atmosphere of the extreme.


new things seem to run like dominoes thru' this forum.

i'd have to agree. who's going to fork up an extra $150 for 1 second faster in super pi. i'd love to get the new cpu, but being fsb limited, what's the point. i'm probably only going to get 450fsb out of my 780i, which means i'll get a better oc out of my 9x q6600.

maybe waiting for the q9650 makes more sense, but then again, maybe the price will be ridiculous. hopefully next gen mobos can do 550-600fsb!

[Pete]

Got mine today, 440fsb max, ppl, vcore, cascade everything tried and tested to make it go some but nowt. Not bad for £200 inc shipping for an ES chip.

Bin fodder now as it's crap!

[Kryckter]

8 Multi is going to make this a hard decision between the Q9450 and the E8400/E8500

[ziddey]

but the lower multi will limit speed since quads still scale less favorably with fsb. after reading on sse4, penryn's have less of an edge over conroe's. while clock for clock, there are still improvements, q6600's are cheaper and are still hitting 3.6-3.8 fairly commonly nowadays.

[Pete]

Okay so it did what i use the pc for faster than anything i had to date so far at the max clock this chip could do. However i got the chip hoping it'd clock more thus making it overall faster...sadly it's not the case.

But if you want lower powerusage, lower clock and less heat but same power for say ya 3.8 q66 then ideal

[jonny_ftm]

Got mine today, 440fsb max, ppl, vcore, cascade everything tried and tested to make it go some but nowt. Not bad for £200 inc shipping for an ES chip.

Bin fodder now as it's crap!

Great feedback. So even cascade won't help it. As I thought, cold usually have few if no effect on FSB wall. Q6600 is the way to go or more expensive 9x multi penryns

[Blacklash]

If the GX2 is solid I may try two of those in a 790i motherboard with a Q9650. Of course I'll have to buy DDR3 too.

[Oblivionlord]

what happened to the numbers on page 1? There isnt any results

[Jor3lBR]

I'm so glad I jumped this upgrade... The Q6600 owns the new Q9450 in my board...

My 24/7 setup runs very cool and with tinny volts with DDR@1200Mhz:
CPU 1.26v Bios 1.28-29v CPU-z
NB 1.36v
VTT 1.255v

And those volts are still workable especially NB and VTT could go lower with a little less Vcore (1.27v real is goal).

Pc is inside a closed case and Fans on Rads working with 7v that's why I get higher temps then normal...

[EJTREM3]

awsome thread, is very helpfoul to decide

[Richard Dower]

Thing is, i held out for over a year waiting for 45nm...so it's worth the upgrade for me, as you can see:

[Jor3lBR]

Thing is, i held out for over a year waiting for 45nm...so it's worth the upgrade for me, as you can see:

My Q6600 without the stress testing I was doing (mix Prime95) run's on similar voltage and temps...