First quiet plus overclocked system. Need advice.

[SebRad]

Hi, rumour has it that the 4MB cache Core 2 Duo chips overclock less well than their 2MB counter parts so as budget matters I say go for E6300/6400. Overclocking is luck of the draw, some chips struggle over 2.8GHz while others go to 4GHz, as far as I've seen it doesn't make a lot of difference what it was to start with, in your case E6300 just as good as E6400 for overclocking with a small caveat. The 8x mult on 6400 means for 3GHz need bus speed of 375 so min RAM speed is DDR2 750 (at least on my Asus i965 P5B-E Plus) can change for higher options but not lower. The 7x of E6300 need 429 FSB for 3GHz so then need DDR2 860, which shouldn't be a problem for most vaguely enthusiast DDR2 800 chips, inc your Corsair ones.
Agree with choice of Ninja, it's the best but wouldn't fit in my case, I've got Zalman 9500AT and it wasn't quiet enough at minimum speed of 1300rpm. With some work it's now sorted.
I had Seasonic S12-380 and the fan ran about 800rpm but wasn't really really quiet, had a rumble to it. Maybe just bad sample but the Neo HE 430 I've got now is proper quiet.
Just to whet your appetite for what maybe I've got my E6600 running at 3.325GHz (7x475) on 1.24V (1.375 in BIOS) with Zalman 9500AT on it. According to Speedfan it's 1125rpm with CPU at 55°C, running 2 copies of folding@home (for SPCR naturally).
Good luck, Seb

[ryboto]

Gran....

Use whatever you like. But the Zalman is both quiet and more efficient at cooling capacitors whereas the Ninja is not. Zalman, for some reason, is the choice of professional DAW builders and since I have experience with this particular cooler I can vouch for it's silence. My ASUS board does have fan control though, which may make a difference.

I've dealt with the 9500. Nothing about the stock fan is silent. I can vouch for that. At 1300rpm the thing is still louder than my hard drives. I fan modded the thing, and I can tell you it doesn't do all that awesome with low airflow. I ended up switching to a Thermalright Ultra-120. With the zalman I was using a 92mm fan at 900-1200rpm. Now I use a 120mm fan at 780-1000rpm, and I can't hear it from my chair. I guess the choice of fan matters, but at low rpm it shouldn't matter all that much what I'm using. Idle temps of the passively cooled Northbridge and PWM circuitry have also gone down. Load temps increased slightly on the PWM, but haven't changed for the Northbridge. I used to be very devoted to the 9500, mostly because I spent a lot on it, and figured I had to justify my purchase, but I just can't advise it's use in a quiet or silent system.

[Byakko]

Okay, here's what I've ordered up to this point:

$79.99 Antec Solo
$89.99 Seasonic S12-430
$209.00 Corsair TWIN2X2048-6400
$47.99 Samsung SH-S183L
$39.99 Scythe SCNJ-1100P "Ninja"
$5.99 Arctic Silver 5
$5.99 Arctic Clean ACN
$15.54 Shipping
$5.04 Tax
-$40 Antec Solo MIR
-$40 Corsair MIR

Running Total: $419.52
Left in Budget: $580.48

So far, so good!

I ended up getting the S12 430 instead of the 330 mostly due to the noise concerns around 150W. Better safe than sorry in the long term. I used that same reasoning for getting the Corsair memory rather than some PC5300 memory. And besides, with the $40 rebate, the Corsair was only $40 or so more than the "value" stuff.

Question time: What kind of fans should I be getting? Will the S12 be capable enough to act as a CPU fan in conjunction with the Ninja? Or will I still need a 120mm or 92mm fan (a Nexus one, perhaps). And what about a case fan?

Wow...you have almost the exact build that I just bought last week (waiting for shipments).

Antec p180 = $80 AR
Corsair 520w psu = $80 AR
Corsair TWIN2X2048-6400 = $165 AR
Scythe Ninja = $40
Artic Silver = $5
LG DVD burner = $30

To finish off:
e6400 = $222
Gigabyte 965P-DS3 Mobo= $120 AR
8800 GTS 640mb = $360 AR.
WD Raptor 74gb = $120 AR

pwnage computer. =)

Although, you can save money by buying an e6300 ($180), the Gigabyte 965p-s3 ($110), 7900gs(or 8600 next month for $~150-200) and it'll still be a nice computer.

XD

[Beyonder]

3.5%, 4.1% that only tells you about performance under load - it doesn't take into account how the computer is used.

That's why I said, percentages are useless for without the proper context.

Actually, it does. I think you're misinterpreting the link. It has a wide set of different benchmarks. 3.5% is the total average improvement of all benchmarks. 4.1% is the average of all benchmarks except for gaming. This has absolutely nothing to do with "performance under load"; it has everything to do with "how the computer is used."

If you wanted your context to be video editing, it'd be pretty easy: add up video-related scores, divide by the number of entries. (it ends up being 5.5%, on average, by the way)

I think what you may be trying to suggest is that the improvement isn't related to the amount of time it's going to save the user. (e.g. if a video task takes ten minutes to complete, and a content creation test takes three seconds, then 5% savings is thirty seconds in the former and milliseconds in the latter)

[pyogenes]

I think what you may be trying to suggest is that the improvement isn't related to the amount of time it's going to save the user. (e.g. if a video task takes ten minutes to complete, and a content creation test takes three seconds, then 5% savings is thirty seconds in the former and milliseconds in the latter)

More or less that's what I'm saying. I fully undertand your point, but my point is that those raw numbers are pretty meaningless when deciding whether or not an individual will benefit from the larger cache without taking account usage.

I think we can all agree that a person that is using the computer to surf the web, email, some office apps, listening to music, and watching videos will never see any performance gain from a larger cache. Obviously, someone encoding video 24/7 on the other hand will benefit greatly from the larger cache.

My point is, most people fall in between those types of users. The closer the user is the the former type, the less relevant those benchmarks are. The closer they are to the latter type, the more useful those benemarks are. Recommending a larger cache without knowing where they fall in the spectrum of users is pointless - the entire point of my original and subsequent posts.

Regarding my comment that it only matters under load, that was an incomplete statement. I should have pointed out that most users are idling most of the time, thus most of the time cache is irrelevant. The cache only benefits the users when the computer is actually doing something. The less often applications that benefit from larger cahce is used, the less the user benefits overall.

[gran]

Although, you can save money by buying an e6300 ($180), the Gigabyte 965p-s3 ($110), 7900gs(or 8600 next month for $~150-200) and it'll still be a nice computer.

That reminds me of another question I have. What does the extra $20 or so give you in the 965p-ds3 vs the 965p-s3 in practical terms? Here is what Newegg has to say about that:

...this Durability Enhanced 'DS3' model that is in all essence exactly the same board [as the S3] except that it sports the new All-Solid Capacitor design

Okay... so? I don't know a lot about motherboard design so what exactly does "All-Solid Capacitors" give me? That is, if I had the S3 and DS3 on two computers working on the exact same thing side-by-side, in what way would the DS3 be better?

[pyogenes]

Although, you can save money by buying an e6300 ($180), the Gigabyte 965p-s3 ($110), 7900gs(or 8600 next month for $~150-200) and it'll still be a nice computer.

That reminds me of another question I have. What does the extra $20 or so give you in the 965p-ds3 vs the 965p-s3 in practical terms? Here is what Newegg has to say about that:

...this Durability Enhanced 'DS3' model that is in all essence exactly the same board [as the S3] except that it sports the new All-Solid Capacitor design

Okay... so? I don't know a lot about motherboard design so what exactly does "All-Solid Capacitors" give me? That is, if I had the S3 and DS3 on two computers working on the exact same thing side-by-side, in what way would the DS3 be better?

the capacitors are more reliable. You're less likely for your motherboard to randomly get fried. No guarantee the S3 will die or the DS3 will last longer though.

[Byakko]

Although, you can save money by buying an e6300 ($180), the Gigabyte 965p-s3 ($110), 7900gs(or 8600 next month for $~150-200) and it'll still be a nice computer.

That reminds me of another question I have. What does the extra $20 or so give you in the 965p-ds3 vs the 965p-s3 in practical terms? Here is what Newegg has to say about that:

...this Durability Enhanced 'DS3' model that is in all essence exactly the same board [as the S3] except that it sports the new All-Solid Capacitor design

Okay... so? I don't know a lot about motherboard design so what exactly does "All-Solid Capacitors" give me? That is, if I had the S3 and DS3 on two computers working on the exact same thing side-by-side, in what way would the DS3 be better?

I chose the ds3 over the s3 for two reasons:

1. Solid state capacitors are "supposed" to be more reliable than the old style, which has a chance of expanding and exploding. What are the chances of it happening? I have no idea, but I like the piece of mind.

2. I think I heard the S3's capacitors were taller and some had trouble with installing the Scythe Ninja on the motherboard. The solid capacitors are shorter, so the heatsink should clear it.