Question Will this make a noticeable difference?

[SoCaL081]

I have a intel i5-6600 non-k and was wondering if i should just run it as it came at 3.3 w/3.9 turbo or would it run noticeably faster if i was able to "overclock via BCLK" and have all the cores are running at 3.9 without turbo. Would that .6ghz be noticeably faster?

 

[rgd1101]

got the cpu cooler to handle it?

 

[SoCaL081]

A liquid cooler w/120mm radiator

 

[clutchc]

Yes, all cores at 3.9GHz would be faster than:
3900 MHz (1 core)
3800 MHz (2 cores)
3700 MHz (3 cores)
3600 MHz (4 cores)

But noticeably faster? Only if you are on the margin now. Does your board allow for bclk OC'ing?

 

[Karadjgne]

You'll not notice the difference. Games are not static benchmarks. Games fps bounces all over the place, watch any video, you'll see by just how much. That counter will also average the fps for a single number, but chances are if it's over the refresh of your monitor, you will not see it anyways. If it's under refresh, you can't physically tell the difference between 50 and 55fps. Your brain doesn't work that small. You'd need a considerably larger range to compare to.

It's @ 8% difference in ability, or in a game at stock 100fps, you'd get 108fps, which you wouldn't see, and if the game stock was 50fps, you'd get 54fps, which you can't see.

It's not enough OC to make a difference in 98% of games, many of which weight cores more heavily than fps, and the rest really don't use cpu speeds over @ 3.4GHz to much advantage.

Not to say you shouldn't try it, but just don't be mad if the results aren't what you were hoping for.

 

[SoCaL081]

I have a MSI Z170A KRAIT-GAMING motherboard and it won't allow me to raise the BCLK up to 118.5 i believe which brings me up to the 3.89 speed

 

[Karadjgne]

BCLK is short for Buss Clock. That's a transmission clock multiplier. In pre-haswell cpus that affected everything from ram to gpu to pcie to USB transmission speeds, which beyond @ 105.7 created serious instabilities in many different components. It also artificially jacked up cpu temps. Starting with Haswell, steps were gradually taken to isolate the BCLK in cpu from outside the socket components but there's no guarantee just how successful that's been. 118.5 is a mess. That's a ton of BCLK and usually combined with extreme OC and LN2 to chase world record stable speeds. Without the OC, the bios isn't able to adjust to power levels necessary to pull it off. If BCLK OC worked, and worked well, you could bank on Intel bumping all their cpus to handle it, and every mobo vendor would be using it as default instead of throwing in some freebie OC software which bumps BCLK to 103.x and then stage 2 is 105.x.
Instead, every single vendor actually agrees upon at least 1 thing, if nothing else in the bios is the same, BCLK is always default factory optimized 100.00 for Intel, it has been for years.

 

[damric]

You will need to drop your RAM frequency while keeping timings loose to raise your base clock higher. Tomorrow I plan on overclocking a 2.2GHz i5-6400T to 4.4GHz for 100% OC. Stay tuned.

 

[koson123]

bclk OCing usually has very minimal gains. in games i doubt this will be noticeable but in other workloads and benchmarks possibly. Also bclk is very finicky. you are changing many things when you play with it(i.e ram, cache, pcie) so you could have instability problems not from the cpu. overall would not recommend unless you are very close to a goal

 

[damric]

bclk OCing usually has very minimal gains. in games i doubt this will be noticeable but in other workloads and benchmarks possibly. Also bclk is very finicky. you are changing many things when you play with it(i.e ram, cache, pcie) so you could have instability problems not from the cpu. overall would not recommend unless you are very close to a goal

Base clock overclocking is what separates the men from the boys. I pulled over 50% performance gains when doing my last one. You can find the thread a few pages back. It was last month.