Very easy IMO, except that pushpins are annoying as hell. Talk to me on AIM or PM and I can go into detail.

Haha, that's not really a problem, but just keep in mind that this forum exists so you can make threads like that and we can properly answer your questions.

Not so much gone bad. They would always boot up and never did go bad. These are problems they had from the start, ones I'd suggest were flaws in the design, rather than those specific boards I had being broken. But I'm picky as far as stability. I want it to be able to run for months without crashing, if possible. (in other words, not crashing at all) Running for just days without crashing isn't enough for me.

Empirical data is useless isn't it?

SLI/Xfire all use two cards. That means more power is used, generally requiring a beast of a PSU, and an extra cost for using two. The benefit you get from dual graphics is pointless, since a single more powerful card will probably be cheaper, and be faster. For those wanting to buy a budget/expensive card, then get another one to dual later, good luck finding said card. If there is no better card (4870x2), and you really want that sp33d, sure, it's an option, but who has $2000 they want to spend on a computer whose equivalent could be bought for $500 in a year? People who know the fuck what they're doing, that's who.

I do agree with your layout etc..., but again, that's something I mentioned already, even though it was inferred.
It's a 3D rendering... it renders just like a game. My statistics project isn't a very good comparison here, but what I found out was that overclocking the CPU had the most effect. I didn't switch out any component, just changed the clock speed. I got virtually no change for the graphics card, but got a significant change in the CPU, and from that I assume that the CPU has the most effect. Now, my CPU might have been bottlenecking my graphics card, but I doubt it, since changes in the GPU were still visible, and predictable.

My point is that if you're short on money, splurge on a CPU, or better yet, balance the two out. It's called the central processing unit for a reason.

On a brand new computer, I consider overclocking to be pointless. Can you tell the difference between a processor at 2.5 GHz and 3.0 GHz? Can you tell the difference between 80 fps and 125 fps? Is that difference noticeable enough so that you have to overclock? For me, it's no. My processor at 1.8 GHz is about the same as 2.6 GHz, unless I play Oblivion. Lucky for me, Doom only listed some source games, which are based on a really easy engine. If you want to play crysis, sure, you'll probably want to overclock.

Anyone can tell the difference between 20 fps and 25 fps, and if your overclocking will really give you a 25% increase in performance, go right ahead, it's worth it then, unless the action is so ridiculously fast that it doesn't matter. Safari is the fastest browser on the web (except when viewing images), but firefox and IE still are used more because the difference in speeds is so small that it doesn't matter, and the convenience of the two make up for the "slowness". When you get down into the 10% increases is when it's pointless. 20 fps vs 22 fps. 60 vs 66 fps. 2 minutes vs 2 minutes 12 seconds. I know my math's backwards, but the point remains.

A final note: I drive the speed limit. Maybe that will help explain some of my thinking.

LOL, it wouldn't be if it actually made sense. Heatsinks don't make noise, fans do. Heatsinks absorb heat. A heatsink with a higher capacity to absorb and dissapate heat doesn't require as fast as a fan. A slower fan = less noise.
Intel's fan is 92mm and runs at a fairly high RPM. So how exactly is this going to run quieter than a larger fan which runs at a slower RPM?
I don't know how FrostyTech's methods and how they got the numbers they did, but they aren't legit, and by no means is Intel's stock cooler the 2nd quietest. There wouldn't be a market for quiet heatsinks if that were the case. For the record, you will never be able to run a passive setup on a Intel heatsink either, while a lot of other heatsinks can.
SLI/XFire used to use two cards. Now, they can use 3, 4, or 1 and a integrated graphics as well. When it comes to the top of the top performance, nothing matches SLI/XFire. Not everybody is looking out to save money down the line either. If you are an enthusiast purchasing enthusiast products for high performance, you should expect to be spending top dollar.
Down the line, sure it isn't ALWAYS beneficial to use XF/SLI, but it's an option nonetheless. Sometimes a second hand video card for 50$ is the difference between you maxing something out and playing it on high.
Also, you'd be suprised what could run on some powersupplies. I have seen a bunch of SLI and XF rigs powered by middle of the pack PSU's. I'm not advising it, and I would definitely recommend being safe than sorry, but the idea that you're going to need a 800W or some extremely high rated PSU is wrong most of the time. Of course, it will produce more heat and cost, but sometimes both of those are negligible.

Also, with hybrid XF/SLI, power can be saved. Use the intergrated graphics for the easy stuff, powerful GPU for the gaming. Leads to a more quieter, cooler, efficient computer.

No, actually, synthetic benchmarks have found to be pretty useless as far as indicators of gaming performance go. Also, what I'm saying is 3DMark 05 is more heavilly CPU reliant than other 3D Mark benchmarks. Your results you found are because you're using 3DMark 05, which is much more CPU bound.
I don't know your methods on how you tested overclocking, but anyway, there is more to just GPU performance than GPU core clock speeds.

Regardless, I am nearly certain that if you used actual games, or a different 3DMark benchmarks that your information would be different. Although, I don't really know the specifics of the testing anyway.

But that's completely not true. 40 dollars can be the difference between a 9800GT and a HD 4850. There is a significant performance difference between them. 40 dollars is the difference between.. say.. an e2200 and an e7200, and that's the biggest gap in performance I could find. You're saying that overclocking is pointless because there's really no performance gain between .5 GHZ, but then you're advocating paying more money on a CPU in order to gain the extra .5GHZ. That makes no sense at all.

Look at benchmarks. Most modern games are HEAVILLY GPU RELIANT. Shit, some games don't even make use of all the cores CPUs have at this moment.
Not all new computers max out everything at 80 - 125 FPS. In fact, a good example would be mine, I bought my computer for 600~ total. Overclocking my CPU from 2.2ghz (A single core Athlon 64 San Diego on 939 costed about 70$ open box) to 2.9ghz (which would have costed $300-$400 at the time.) However, my processor was only single core, and was limited because of this. Long story short, games like Oblivion were much more playable with that 33%~ increase in clockspeed from my CPU. And I saved myself hundreds of dollars because of it.

And by saying that you admit there is a tangible benefit from overclocking.
And why not overclock anyway? Is there anything wrong with having a faster processor even if you're not gaming? Of course not.

Hey, maybe you can't notice it, but why not. It's not going to cost anything but your time. And there are other reasons as to why Safari isn't used as much as the others. I don't think it has to do with speed.

1500 RPM. My motherboard runs at 7200 RPM. It's pretty quiet.
The same way some fans are just louder than others, and the fact that sound increases exponentially past 1 m/s.
http://www.frostytech.com/testmethod_mk2.cfm
Except that they rated 3 "quiet" heatsinks above it. Look at some fans, and see their size, RPMs, and dB rating. It's not surprising in the least.
No duh.

Power Consumption. Stability. Warranty. I actually pay my power bill. If you don't need to overclock (which you don't for source), there's no need to. In a few years, when my computer is really out of date, sure, but now it's just fine.

Safari is faster, that's why people use it. Most other people think it's stupid, not secure, and think Firefox is much more customizable. My point was that the only thing Safari does better than Firefox is load web pages faster, and since it's not that huge a difference, performance doesn't matter, making Firefox the superior browser.

Post has been edited 1 time(s), last time on Aug 23 2008, 8:00 am by rockz.

A friend of mine just informed me that my family desktop can be upgraded. Apparently it has a PCI-E slot, which I did not know.

So I'm happy!!! This is saving me $600! Thanks for the help though guys, Excalibur I'll keep those links just in case in the future I plan on making one (probably in 1-2 years).

All I need to go is get 2 x 1GB of RAM and this card, and I'll be a lucky man.

HOOOOOOOOOOOOOOOOOOLD on buddy. 8600GT for 93$? I say we can do better.
http://www.newegg.com/Product/Product.aspx?Item=N82E16814121251 for 94.99$ Price:Performance, I'd say this onces a better buy. Why?
2$ more gets you:
192bit memory interface vs 128bit
384MB on board GDDR3 RAM vs 512MB
96 stream processors vs 32
1800MHz memory clock vs 1600
And although the other has 80MHz more core, in benches, the 9600GSO still wins. You could OC the 96GSO's core more than 80MHz with no problem at all.

And on the RAM, I'd rec:
http://www.newegg.com/Product/Product.aspx?Item=N82E16820146118 44.99$
Only a dollar more than the G.Skill and they run at 5-5-5-12 rather than the 5-5-5-15 of the G.Skill.

Post has been edited 1 time(s), last time on Aug 26 2008, 10:21 pm by Excalibur.

Can he use DDR2? If he needs DDR, it's going to be expensive.

In my experience most DDR systems use AGP while DDR2 are usually the ones with PCI-Ex16. Not that this is always the case, but I've found it to be more likely.

I actually found some other cards that were the same as the first (besides 256MB and 512MB), here:

XFX GeForce 8600 GT 512MB PCIe

EVGA GeForce 8600 GT

Also the one in Anonymous thread I was looking at. Same GeForce 8600 GT, what's the difference?

Also, I have no clue if I'll be getting 2GB or 1GB RAM now, so I was looking for 1GB and found a Kingston

Scratch that, I think my desktop only supports 184-pin DDR1 SDRAM and Speed of PC 3200...looking for a $20-30 price range for this.

Post has been edited 4 time(s), last time on Aug 27 2008, 7:32 pm by Xx.Doom.xX.

The GFX card I recommended above still beats the both of those. The 9600GSO is similar to a 8600GTS, rather than a GT. For RAM, Kingston is just fine.

Well, unless that $90 performs like 5x better than the 8600GTS, then I'd consider it. I just want to be able to play Valve/source games mostly. I don't know what else I'll be wanting to play. Maybe Far Cry 2 when it's released and Spore. Perhaps Crysis at Medium settings I don't know yet.

This is why I like AMD. Socket 754/939 is DDR. AM2 is DDR2. Both come with PCI-e.

The DDR ram is going to be expensive. Go ahead and go all out, get 2 GB of it. It shouldn't be more than $100, but I haven't looked at all.

A decent 8600 will play oblivion, which I consider to be one of the most graphics intensive games ever. Make sure you DON'T get a ddr2 version. The total VRAM won't matter that much in older games (oblivion uses all of my 256 MB, source doesn't use it all, except in huge CSS maps). In Savage, I only use 16 MB, but that's a pretty old game.

Oh, and in that 1-2 years, those links will not work, and will prolly be replaced with much better stuff. Just ask ex again when you want to make one.

$100? Hard to believe for just DDR RAM.

Alright, is the EVGA good enough to buy? It's $50 and I like the price a lot...I'd like to spend the least amount possible.

Man, I don't know anything about low or mid range cards.

All manufacturers are pretty much the same. Sometimes they adjust minor things from the reference model, like adjustments in clockspeeds. But you are ultimately getting the same card.

DDR is so out of date, I don't think it is even in production anymore. The price is only going to go up as the supply shrinks.

Well, I assume the desktop doesn't support DDR2, it was made in 2004 though...maybe it's just the memory that it has in right now that is DDR. I don't know. Nevermind, only DDR.

Awesome, I'm going to get the EVGA $50 one then. Thansk for the help.

Post has been edited 1 time(s), last time on Aug 28 2008, 9:48 pm by Xx.Doom.xX.

Before you go through with that, are you sure you have PCIe 16x? You might have AGP, which is older and inferior. And you might have DDR2. Don't assume. Better safe than sorry.