48 replies to this topic

[Smokeyjedi]

Every time I have had an issue from OC was when I was asking for trouble trying to dial in a crazy clock speed or Voltage.......or lazyness causing dust build-up. These days OC is no more difficult than setting up a custom Power profile on a laptop to save battery while not plugged in.

[Lordred]

Pastor, if you have not upgraded yet, contact me, I can give ya the run down on how to squeeze more omph out of that AMD.

[ninjitsu]

I recently upgraded off of an AMD FX 6300. Now I'm running an i5 4670k, got it for $219. Soooo much better than my old FX. FX series chips have a hard time with floating point math if all the cores aren't utilized because they don't have true cores. They use pairs of cores that share the same floating point module.

Personally, I'd avoid AMD until they get their stuff figured out. The way they handle single threaded applications is very poor. I saw a 62% increase in single threaded performance when I went to the 4670k, both chips at 4ghz.

[Pastor Priest]

I'll shoot you a PM, Lordred.

Barring that, what do you all think of the I3 4360? While not a high end I5, it still has gooperformance scores and Micro Center has it for $140.

[ninjitsu]

Pastor Priest, on 10 June 2014 - 10:08 AM, said:

I'll shoot you a PM, Lordred.

Barring that, what do you all think of the I3 4360? While not a high end I5, it still has gooperformance scores and Micro Center has it for $140.

For a dual core, that i3 benches pretty darn high!

[xWiredx]

Considering MWO will use up to 6 cores, 4 real cores is going to perform a pretty decent amount better than 2 real cores and 2 hyperthreading "cores" so an i5 of the same clock speed will net you a better MWO experience. This is the reason why an i7 will perform better in MWO than an i5. In most games, the i5 and i7 perform the same because most games only use 2-4 cores. With MWO that is not the case, though.

[Goose]

Pastor Priest, on 05 June 2014 - 05:34 AM, said:

Well, that deal from Micro Center does seem pretty nice. Will it get any better if I wait until 5 gen. is released at the end of this month?

Sort'a? http://www.tomshardw...clock,3106.html

[Catamount]

Goose, on 10 June 2014 - 06:13 AM, said:

Emphasis mine …

Right, "too many". I'm loving that scientifically defined sample right there. "Too many" might well be 200 anecdotal cases, out of what happened to be a sample of a million readers, the other 1,999,800 of which probably would have mostly said nothing. As far as I know, most people don't go out of their way randomly flood websites with comments about how their hardware behaves perfectly normally. Throw in the fact that GPU failures were included, and the fact that GPUs are absurdly easy to damage with OCing (especially uncooled VRAM that's often a characteristic of non-reference coolers), that might well be 170 or those 200 cases, and then we're left with what percentage of readers? Now, of course, I'm making assumptions, but any estimate based on the words "too many" is going to involve assumptions, and I think you'd have to bend over backwards to come up with a set of assumptions that makes that any significant percentage of OCers.

Also, if a chip was so abused, why would it work for several months after purchase, and then fail, as opposed to being immediately unstable? What are the odds that someone just happens to be selling a chip that acts perfectly fine in every regard, for reasons that had nothing to do with any apparent stability problem, that unbeknownst to anyone just happened to be six months from failure?

Edited by Catamount, 11 June 2014 - 05:29 PM.

[ninjitsu]

Thermal damage can cause an eventual degradation of CPU stability.