CPU Performing Below Average Benchmarks

I'm trying to benchmark my CPU to make sure that everything's performing as it should, but no matter what benchmark I use, I seem to get between 5-15% the scores that other people get for the same CPU.

I'm using a 6600k at stock.

For example, using passmark to bench my CPU, I get a score of around 6900-7200 (7200 was the max I got).

Meanwhile, look at the average score others get:

cpubenchmark.net/cpu.php?cpu=Intel Core i5-6600K @ 3.50GHz

What's going on here? Can CPU performance vary this much between different CPUs of the same model? Did I lose the silicon lottery? Should I ask for a refund?

You'd think that maybe some individuals overclocking would drag the the mean upwards, but on the same website I viewed the page for the 6600, and there was a guy using a H170 chipset motherboard and an i5-6600 who scored 7,837. I don't even fucking know.

I'm lost here, Holla Forums. Help me not regret this purchase.

RAM bottleneck maybe?

I don't understand how that would come into play, but my ram is running at 3000 mhz, would that be a factor?

Also, if anyone is running a 6600k at stock, I'd really appreciate it if you could run passmark and tell me what you get for the CPU section.
I've heard of CPUs varying in how well they can be overclocked, but varying in how well they perform by 10-15%? This seems like too much.

It's shit.
Maybe you're experiencing thermal throttling.

Why do you think passmark is shit?
I also used novabench and cpu-z, but results were roughly the same.
I also don't think I have thermal throttling, none of my cores seem to go above 45 degrees celsius.

Because it's proprietary, which means:
1) It could be partial.
2) It's not compiled to take advantage of your exact CPU, thus isn't fit to measure hardware performances.

I suggest you use something like x264 or x265 to benchmark your hardware.

That might be true, but if I'm comparing my scores against people with the same CPU using the same benchmark, and I'm scoring 10-15% lower then they are.

Might be the OS. Anyway, your fault for using NSAOS and a Jewtel CPU.

windows 7?

also, there isn't really an alternative. The 8320 is shit on single-core performance.

Check voltages and temps of both the CPU and the ram. Drop the ram to 2133 as well.

NSAOS 10 could have SMP and build optimizations improvements over 7.
Works4me (I'm not an AAA gaymes, by the way)

Also, use a less faggy benchmark tool. Run prime95 or something where you know it actually tests the CPU and not random other components.

If I'm using prime95, how can I know that my 6600k is performing as well as other 6600ks?
Also, is this really even a thing (different CPUs of the same model varying wildly in performance)?

bump

Did you even put a heat sink on there ?
I'm hoping you're not a dumbass, but who knows.

Also, post a screenshot of the cpu-z, while under load with something.
And run geekbench.

No. Some are more OCable but if you aren't OCing that's irrelevant. Check voltages, temps, and see what clock rate it's running at. Also that you have the ram in the correct slots and it's using all channels and is running at the right clock (will appear as half what you expect due to dual channel).

Try overclocking, if you overclock badly you lost the silicon lottery, but atleast you might be able to claw back some performance.

Also I think It's unlikely intel will give you a new CPU unless it outright fails.

It's highly unlikely it's the CPU that is the problem

PEBCAK

Overclocking when he doesn't have a grip on things as they are normally would be a big mistake.

I wanted to overclock, since I bought an overclocking motherboard, but my CPU cooler seems a bit insufficient (i'm using the h7). Some cores on my 6600k were getting close to 60 degrees celsius at times while at 4.1 ghz, so overclocking to like 4.3 would be out of the question I guess.
What cpu cooler are you using?
I think I might return my CPU cooler and buy something a bit better, so I'm looking for better options.

On second thought, maybe I applied my thermal paste incorrectly. I turned off the smart fan option, and with my 6600k at 4.1 ghz under load, one of the cores reaches 64 degrees celsius at times, but the coolest core is only reaching 51 degrees celsius. Would this be a sign that I've applied the thermal paste incorrectly?
Also, what temperatures should I aim for? Google says under 60 degrees celsius, but google also said that my current cooler should be fine for overclocking the 6600k to 4.3-4.5 ghz, but that can't be the case if I need to stay under 60 degrees. Something seems off here.

Skylake doesn't throttle until around 90C I think. You'd notice on a tool like CPU-Z that during max load the CPU frequency is dropping (again, check the frequency). The only bad thing about high temps below 90C is that you'll likely have a lot of fan noise.
60-70C is pretty normal load with working thermal paste unless you're testing AVX with a recent prime95. That's some nvidia-tier house fire. A lot of overclockers refuse to test with it because it reveals how shit their "stable" OC is. But those are unprivileged instructions so if your computer melts down then it's broken as far as I'm concerned. You can push most builds into throttling territory so maybe what I consider broken doesn't matter.
Re random core temps, welcome to Skylake. It just does that. It doesn't seem possible to tell if thermal paste is uneven by the temps anymore. Just watch the total package temp.
For temp targets, it's up to you. First realize how fuzzy temperature is. It depends on lots of factors like the PWM curve (different motherboards have different defaults and many are configurable), types of fans, etc.. Think instead of it being a temp vs noise decision up until the CPU fan hits max RPM which is the real 'temp' limit. Faggots will tweak the curve or disable it entirely so they can post that they have 40C at load because of their masterful builder skills but they won't tell you their computer now sounds like a leafblower. Others might gimp the fans with 'low noise adapters' or remove them entirely and be happy at 85C as long as it doesn't throttle. Most gamer setups are designed to hover in the 60-70C range and you should be in good company there.

Thanks a lot for your response.

Wait, shouldn't I be watching the temperatures of the individual cores to watch for excessive temperatures and possible thermal throttling?

A 13 degree difference between max temps on my second and third cores is considered normal? I find this hard to believe. I'd have an easier time thinking I somehow fucked up the thermal paste.

I was also looking at tweaktown's review of the h7, and they say that the stock intel cooler can put the CPU up to 80.25 degrees celsius under load. Would this mean it's probably a safe bet that most skylake CPUs are fine up to the low 80's for temperature? As long as this CPU lasts me a good four years or so, I'd say that's fine.

I'm still very frustrated by how the CPU doesn't seem to be performing as well on benchmarks as I'd like, but if, as you guys say, it isn't the CPU's fault, and it isn't thermal throttling, maybe the scores that I'm referencing my CPU against are at fault.

I really wouldn't mind a noisy cooler if it cooled really well. My H7 is super quiet, is there anything louder with better performance for roughly similar price?

I'm only using a cooler master hyper 212 evo as my cooler. I get around 60-70-ish degrees when ever i put it under heavy load, and I don't really break a sweat about it.

When i get back from work i can put my i5 back to stock so you can compare performance.

No need, I just compared my 6600k's passmark score running at 4.1 ghz to your 6600k's, and they're pretty similar.
Thanks for the help.

no problem