Rejoice! Base Clock Overclocking to Make a Comeback with "Skylake"

[samezra]

Dang it! It requires iGPU be disabled, well this sucks!!

 

[qubit]

links plz...

I haven't seen much real world results from cache in any (intel)platform. Only in a couple of hwbot benchmarks does it matter too...

For bclk only, you won't see any increases really on z170 since it isn't associated with any other bus.

No, I wasn't referring to any specific reviews. I was just saying that full reviews would tell us objectively just how much memory overclocking will get us. I'm sure that TPU is gonna do it. Kinda a given.

 

[newtekie1]

Oh, this makes me happy. I'd love a nicely clocked i3. Though I wish Intel would just release a K series i3 instead of using these janky hacked BIOSes...

 

[EarthDog]

No, I wasn't referring to any specific reviews. I was just saying that full reviews would tell us objectively just how much memory overclocking will get us. I'm sure that TPU is gonna do it. Kinda a given.

there are reviews out which tested this already. Memory is very rarely a bottleneck.. dual or quad channel.

...and I thought you were talking cache anyway?

 

[R-T-B]

I believe without the Bios, the BLCK had a "coded" wall of 102.8Mhz

It didn't. I had it clocked up to 125Mhz no prob at one point. It just hurt everything else because without a special bios (which this news suggests) all other clocks are generated from multiplying the BCLK.

 

[Steven B]

No, I wasn't referring to any specific reviews. I was just saying that full reviews would tell us objectively just how much memory overclocking will get us. I'm sure that TPU is gonna do it. Kinda a given.

Cache increases help a lot with memory latency performance, i already did the proper testing.

 

[Steven B]

Also, the external clock generator shouldn't have much to do with this.

#1 I have a BIOS to unlock non K SKu OC on asrock's mini ITX board which doesnt' have a dedicated clock generator.
#2 You dont' need a dedicated external clock generator to OC BLCK, you only need it if you want to OC BLCK really well (intel says above 250 needs a dedicated external, otherwise the PCH provides both the BLCK and the DMI/PCi-E seperatly).

 

[qubit]

there are reviews out which tested this already. Memory is very rarely a bottleneck.. dual or quad channel.

...and I thought you were talking cache anyway?

Cache increases help a lot with memory latency performance, i already did the proper testing.

Also, the external clock generator shouldn't have much to do with this.

#1 I have a BIOS to unlock non K SKu OC on asrock's mini ITX board which doesnt' have a dedicated clock generator.
#2 You dont' need a dedicated external clock generator to OC BLCK, you only need it if you want to OC BLCK really well (intel says above 250 needs a dedicated external, otherwise the PCH provides both the BLCK and the DMI/PCi-E seperatly).

Guys, I wasn't talking about cache memory or anything fancy. I'm just saying that increasing the bclk speed will also increase the main memory clock which would tend to remove bottlenecks when the CPU core is overclocked, that's all. If it's already not a bottleneck then that's great.

AnandTech is going to review this overclocking functionality with a raised bclk when the new mobo arrives, so it will be interesting to see what performance gains they get, both with the core and the memory speed.

www.anandtech.com/show/9848/bclk-overclocking-intels-non-k-skylake-processors-coming-soon

 

[Steven B]

Guys, I wasn't talking about cache memory or anything fancy. I'm just saying that increasing the bclk speed will also increase the main memory clock which would tend to remove bottlenecks when the CPU core is overclocked, that's all. If it's already not a bottleneck then that's great.

AnandTech is going to review this overclocking functionality with a raised bclk when the new mobo arrives, so it will be interesting to see what performance gains they get, both with the core and the memory speed.

www.anandtech.com/show/9848/bclk-overclocking-intels-non-k-skylake-processors-coming-soon

He just reviewed the OC Formula right? The BIOS for that board is already public. If I had a non-K SKu CPu i would test it, but I don't have one hahahaha, I have more k SKUs than i need, but not one non-K SKU. I am honestly waiting to see what Intel does, they can force manufacturers to reverse course, or they could embrace it, or they could keep quiet. At this point all the manufacturers have an excuse to produce this type of hack, but let's see if they force non reversible BIOS upgrades across the board.

 

[trog100]

market forces control it all.. a chip for every price level.. this situation is artificially contrived.. its not real.. intel need to be able control chip performance and clock speeds..

sales volumes are always more at at the cheaper price levels.. this volume is met by artificially restricting chip performance to match demand.. its all pretty simple..

folks aint gonna buy the more expensive part if the cheaper part with few clicks of a mouse performs just as well.. intel lose nothing with multiplyer unlocked K parts because they are the most expensive parts anyway.. they will lose if folks can make the less expensive parts go faster.. it buggers up the entire pricing structure..

intel try and give the impression they are in favour of "overclocking" and providing its only done with the most expensive parts they are happy bunnies.. they wont be happy with this latest news.. he he

its worth adding that there is no true competition between the two players.. intel set the levels and amd struggle to keep up.. the same applies to graphics cards.. the better player rigs the market..

trog

 

[EarthDog]

If only you knew how few people this really impacted....

Guys, I wasn't talking about cache memory or anything fancy. I'm just saying that increasing the bclk speed will also increase the main memory clock which would tend to remove bottlenecks when the CPU core is overclocked, that's all. If it's already not a bottleneck then that's great.

AnandTech is going to review this overclocking functionality with a raised bclk when the new mobo arrives, so it will be interesting to see what performance gains they get, both with the core and the memory speed.

www.anandtech.com/show/9848/bclk-overclocking-intels-non-k-skylake-processors-coming-soon

You quoted someone talking about cache (and memory), fyi...

It really isn't going to make a difference compared with a memory multiplier.

Why do you believe it will make a difference?

 

[Schmuckley]

Someone wants a decent quite new LGA1150 setup?

EDIT: And for the love of mercy, someone try this on Celerons and Pentiums.

Due to current budget..It will be done
Gonna have to have dat OCF though.
PS: Intel should really only make about 5 chips;They are hurting themselves with production costs for all the silly segmentation.

 

[EarthDog]

They are hurting themselves with production costs for all the silly segmentation.

Actually, they are not. It helps to understand how they make the different bins.

 

[rtwjunkie]

Actually, they are not. It helps to understand how they make the different bins.

Just to add to this @Schmuckley, they are thus able to make use of the entire wafer. Sections are tested. Their quality and response determines what they will be.

 

[Deleted member 24505]

My greatest base overclock

I say welcome back to it, some of my best times fiddling were back when you could oc this way. glad to see it returning