overclocking i3 6100 on a budget, worth it?

[monstreys]

hey everyone,

i'm going to make my first build soon and have sort of settled on pairing the i3 6100 with an r9 280x, i was planning on going with a b150 motherboard which will allow me to install a second 280x in crossfire in a couple of years (when this one becomes too slow)

should i want to overclock the gpu (working with the base clock) i would have to invest in a z170 board and a better cpu cooler, in total this would be an extra 60-70 euros.

would that investment be worth it? considering my build? i'd be working with 1833 ddr3 ram.

already thanks for all your help,
kindest greetings,
monstreys

 

[RedJaron]

When the software suits the FX it is still more than competitive

And that is a HUGE caveat you're conveniently glossing over. "Heavy lifting" tasks like code compiling and video transcoding are not even close to the same kind of load as games, and I'm pretty sure you know that. The LGA115X and AM3+ platforms have far more difference than just core/thread count. Things like instruction set support, IPC, and memory bandwidth also matter a lot, and Intel crushes AMD right now in the latter two.

The 8350 is better at heavy lifting when it can bring all its cores to bear and, more importantly, can keep them saturated at 100% load. When was the last time a game pegged your CPU at 100% load? When was the last time you played an eight-threaded game? The vast majority of games right now don't even use four threads, so please reconsider your argument that an i5 will get overwhelmed in a big multiplayer game. Except for a few outliers, your gaming experience with an FX vs i3 vs i5 will be very similar provided you have a proper GPU for your graphics settings.

You're essentially saying that the 8350 beats the i5 in one eight-thread benchmark, therefore it must win ANY eight-thread benchmark. That's simply not true. Get the right tool for the job. Just because a farming tractor has more raw horsepower and torque and can haul a plow better than a Civic doesn't mean it's better at commuting to work.

 

[Outlander_04]

When the software suits the FX it is still more than competitive

And that is a HUGE caveat you're conveniently glossing over. "Heavy lifting" tasks like code compiling and video transcoding are not even close to the same kind of load as games, and I'm pretty sure you know that. The LGA115X and AM3+ platforms have far more difference than just core/thread count. Things like instruction set support, IPC, and memory bandwidth also matter a lot, and Intel crushes AMD right now in the latter two.

The 8350 is better at heavy lifting when it can bring all its cores to bear and, more importantly, can keep them saturated at 100% load. When was the last time a game pegged your CPU at 100% load? When was the last time you played an eight-threaded game? The vast majority of games right now don't even use four threads, so please reconsider your argument that an i5 will get overwhelmed in a big multiplayer game. Except for a few outliers, your gaming experience with an FX vs i3 vs i5 will be very similar provided you have a proper GPU for your graphics settings.

You're essentially saying that the 8350 beats the i5 in one eight-thread benchmark, therefore it must win ANY eight-thread benchmark. That's simply not true. Get the right tool for the job. Just because a farming tractor has more raw horsepower and torque and can haul a plow better than a Civic doesn't mean it's better at commuting to work.

In just this same way its a huge caveat to say that an i3 or i5 is a better gamer because it can produce higher fps in lighter loaded game situations .
What was your point?

 

[Outlander_04]

...
When the software suits the FX it is still more than competitive

This. Apparently, most benchmarks, for most games (there are exceptions), show that those games are better-suited to higher per-core performance. Rendering and transcoding are not gaming. Rather than try to poke holes in data presented by others, I believe it would be more useful for you to present some data of your own.

If presenting data of your own is important you could try that too . And since its your idea you are welcome to go first .

Most of the intense multiplayer games of the last few years have performed as well on FX cpu's as the intel i5's . There are no way to produce repeatable benchmarks just because of the nature of multiplayer .
But never the less this is the experience I , and many others , who have used these processors have experienced

 

[logainofhades]

Why buy a processor that only does real well, in a few titles? If I am spending hundreds of dollars, on a system, I want something that is going to give more consistent performance, across all titles.

 

[RedJaron]

In just this same way its a huge caveat to say that an i3 or i5 is a better gamer because it can produce higher fps in lighter loaded game situations .
What was your point?

Except no one here is trying to say, "Since the i3 wins at one game, it must win at all games" like you're doing. Also, the specific benches presented here are of Crysis 3, Witcher 3, and Fallout 4. Are you really going to claim those titles are "light gaming"?

 

[Outlander_04]

In just this same way its a huge caveat to say that an i3 or i5 is a better gamer because it can produce higher fps in lighter loaded game situations .
What was your point?

Except no one here is trying to say, "Since the i3 wins at one game, it must win at all games" like you're doing. Also, the specific benches presented here are of Crysis 3, Witcher 3, and Fallout 4. Are you really going to claim those titles are "light gaming"?

I havent said anything of the kind . I have said that single player benchmarks can be misleading .
And all the benchmarks presented here are all single player .

If there is anything else I have written that you cannot understand I suggest you take a deep breath and read it more slowly

 

[Jared2606]

Except no one here is trying to say, "Since the i3 wins at one game, it must win at all games" like you're doing.

I havent said anything of the kind.

Really? Let's try a little fact checking.

The cpu work load in a multiplayer map is much higher . More threads will help,

Not necessarily. A CPU offering more logical cores only helps if the game has been programmed to run on multiple threads. Just because a game is "multi-threaded" doesn't mean it will instantly use every core your CPU has available. A game that has only two threads is still multi-threaded. But that game won't care whether you've got four cores or eight, it still can't use them all. The vast majority of games now use four cores or less. So, considering Intel currently has better per-core performance, and considering that most games can only use up to four cores, why do you think a game will run better on four slower AMD cores than on four faster Intel cores? On games that can use more, yes an FX might have a little benefit, but let's explore that below.

but total usable cpu horsepower is more important. And since an FX 8350 is more powerful than even a skylake i5 the i3's weaknesses appear .
Even an overclocked FX6300 will be better than an i3 in loads like that.

Here we go, what you said you didn't actually say. You've said repeatedly that the 8350 is more powerful than a Skylake i5, though you didn't bother to quantify what that means. The only supporting evidence you gave is a Guru3D link and specifically the Espresso and Handbrake benchmarks. So, you're suggesting that because the 8350 beats an i5 in just one test ( since you haven't offered any other supporting points ), it must be a far more powerful CPU than the i5.

I suggest you look more closely at benchmarks before you provide them as support for your arguments. In the Espresso test, the 6600K loses to the 8350 by two seconds ( 47s to 49s ), but only when AVX acceleration is disabled. Turn on AVX and the i5 can do the task in 12 seconds The i5 actually beats every FX CPU on that list in the Handbrake portion. I'm unsure how you reason these results are evidence of the FX being "a way more powerful CPU" than a Skylake i5. The fact that a CPU with twice the cores as an i5 barely beats it in only one out of three tests suggests to me that the i5 has almost twice the efficiency per core of the 8350. So, even if it can't handle all eight threads of some extremely threaded game, it seems to me it can deal with each of those threads almost twice as fast as the 8350. So by juggling the threads around it could have almost the same overall performance.

Even an overclocked FX6300 will be better than an i3 in loads like that.

Well I would hope so, because an OC'd 6300 platform costs more than a regular i3 platform.

You are right . Multiplayer and multithread is not the same thing . But in those games that support multithreading then multi thread processors perform relatively better under a high multiplayer load.

Whew, I guess it's a good thing that the i3, i5, and most of Intel's CPUs are multi-threaded then, isn't it!

So, all in all, that's another big swing and a miss. Care to try for strike three?


Now then, we have rules here at TH. One of those rules is to not mislead other users asking for advice by giving them false information or presenting opinions as though they were fact. If you insist on making controversial statements, I suggest you back them up with supporting evidence ( the more sources the better, usually ). We enforce the rules here, so breaking them leads to consequences. And if you cannot understand what I have written here, I suggest you take a deep breath and read it more slowly.

And then he went silent xD

 

[Onus]

Closing necrotic thread.

 

[Outlander_04]

FX is simply not a good idea. Fallout 4 is a very demanding title, and the FX 9590 loses to an i3 6100. FX is old, and its age is showing. FX has been stagnant, while Intel keeps making minor improvements. Future titles being more multithreaded, is not going to magically make FX better, than i5. The cost to overclock an FX 8350, to match an FX 9590, would surpass the cost, of just getting the i5. Not to mention the more powerful PSU you would need, to handle it all.

Witcher 3 is pretty demanding. The 9590 does do better, vs i3, but still loses to an older locked ivy bridge i5.

Posts like this are exatly what is wrong with the intel fans perspective .
The OP will be using a 60 Hz monitor . That means that the most the monitor will ever display is 60Hz or 60 fps . If it receives a signal to display more fps it drops the frame or tears the image . Once you know that it becomes obvious that exceeding 60 fps is utterly pointless and will not change the user experience for the better ........BUT COULD MAKE IT WORSE .

With out knowing it you have actually proved that spending more on the intel is pointless .

 

[Outlander_04]

Except you are not paying more for Intel. An i3 can keep up with, or even beat an FX 8350, while being a good deal less. FX 8350, with a proper board, and cooling, would cost as much, if not more than an i5, that doesn't need a more expensive board or cooling. Lower wattage/less expensive psu's can be used with Intel too.

There could be exceptions but the FX is a way more powerful processor and when there is a lot of cpu load as might happen in a 64 player map , and the game can use all cores then the i3 is not competitive .

Unfortunately the only repeatable benchmarks are single player and that can produce a result , but only one that is relevant to single player . Not something that most gamers build pc's for

 

[Outlander_04]

Crysis 3 is a multithreaded title, yet i3 doesn't get blown away by the more expensive FX. Actually, the i3 was better, than FX when looking at frame times.

The FX 8320e is a lower power , low clocked variant of the FX line . Its clock speed is 3200 Mhz , vs the 3500 Mhz of an 8320 .
It also seems to have much more aggressive use of cool'n'quiet compared to a regular FX so it speeds up later and slows down sooner under load .

The only reason you would include it in a comparison is if you wanted to distort the results , or you were too lazy to get a full fat chip

And of course this is another single player benchmark which as I said previously doesnt coincide with the way most people play games on their computer