First Skylake OC report (bring the salt)

lazz

lazz

Limp Gawd
Joined
Apr 15, 2007
Messages
324
So the first report of an OC on a 6700k has popped out from a hardware design company in Hong Kong. They're reporting an OC from the stock turbo speed of 4.2GHz to 5.2GHz, with an air cooler and only 1.35v.

If the actual OCs are even in this ballpark, a lot of enthusiats are going to be very happy with this release. That said, these results are so nuts that big grains of salt are needed.

Linkage: http://wccftech.com/intel-core-i7-6700k-skylake-overclocked-52-ghz-air-cooling-135v/
 
But in retrospect is it really that impressive? Considering we had 5 ghz Sandy overclocks four years ago, and even some on later (Ivy / Haswell) and E variants. Still good stuff to chew on though.
 
ShuttleLuv said:
But in retrospect is it really that impressive? Considering we had 5 ghz Sandy overclocks four years ago, and even some on later (Ivy / Haswell) and E variants. Still good stuff to chew on though.
Click to expand...

yes but add 10% IPC increase to Ivy Bridge and then 10% more to haswell.. Skylake in every leak are reported to add a huge 20-30% IPC increase.. so 40-50% Increased IPC from sandy bridge to skylake should be considered as a major performance increase so at 5GHZ sandy vs skylake should be a huge difference.
 
If these hold up (and even then if they don't as well), it might be time for me to upgrade my lovely i5 2500K Sandybridge machine to one running Skylake!
 
Skylake looks really promising. I was thinking of going the x58 xeon route however at this point the incremental gains are really adding up
 
Devils canyon is ringing in my ears for some reason. Even if this clocks to 5ghz will it worth it for a cpu, motherboard and ram upgrade from a 4.5 2600k? Probably not.
 
My 5GHz 2600K will finally be allowed to retire. Looks impressive. Can't wait!
 
We'll see.

At least they won't be stating from the dark ages of TIM material selection like they were with the early Haswell cores. That was blamed for the ~200 MHz backpedal Haswell had over Sandy. If we can believe they will start with the TIM from Devil's Canyon, then at least this has a possibility of being real :D
 
So my SB has been running (conservatively but cool) at 4.4GHz for four years. If I can get a Skylake running at 4.6GHz, that's a 4 generation jump (SB to IVY to HASWELL to BROADWELL to SKYLAKE). Let's generously assume 10% jump in IPC each, along with a little extra core speed (200MHz or about 5%) to total 45% or so and a slew of new instructions that by now might even be in use (in this case new are F16C, AVX2, FMA3) and it might be a nice improvement.

I'd love to see some SB-Skylake comparisons when the new chips come!
 
I think the first Haswell-e overclocks was 5gh on air too, right? Did anyone ever see those kind of overclocks in real life?

Gimme Skylake E :)
 
Araxie said:
yes but add 10% IPC increase to Ivy Bridge and then 10% more to haswell.. Skylake in every leak are reported to add a huge 20-30% IPC increase.. so 40-50% Increased IPC from sandy bridge to skylake should be considered as a major performance increase so at 5GHZ sandy vs skylake should be a huge difference.
Click to expand...

Please feel free to check my numbers

https://docs.google.com/spreadsheets/d/18i4ux9eY7yc4QfD3bZp1QGwaRfc8R8LU_LZv7vch9oE/edit?usp=sharing

I am using Anandtech's bench tool to compare the 2500k and the 3570k, and then adjusting 3570k results for the 3% difference in clock speed (not accounting for turbo core though).

For Ivy vs Haswell I am using the 3770k and the 4770k as there is more data available - and they are the same base clock speed.

I get an average 6% IPC increase moving from SB to IB, and 8% moving from IB to Haswell. The IB to Haswell includes video benchmarks, which is questionable due to the new instruction set, but also includes a lot of games where there are no or little changes.

There is no sample data in bench for Broadwell. I would say it might be fair to assume Skylake will offer a 25-30% IPC increase over SB.
 
Last edited:
ShuttleLuv said:
But in retrospect is it really that impressive? Considering we had 5 ghz Sandy overclocks four years ago, and even some on later (Ivy / Haswell) and E variants. Still good stuff to chew on though.
Click to expand...

It might be an indication that they are caving, and no longer using paste under the heat spreaders.


This would be fantastic news, for overclock, and people who care about long term endurance of their systems, alike.

I can't help but think that the usage of paste by Intel was a "planned obsolescence" move. They know that the paste is going to degrade over time, and when that happens, the chip will run hotter and hotter until it starts throttling itself, feels "slow" and people will assume their computers have become old and slow, and buy new ones.

A bit sleazy if you ask me. If this is actually the case, it will be good news, but I'm not counting on it from Intel. They are known to try every sleazy (and sometimes illegal) trick in the book.
 
Last edited:
Quix said:
At this rate I'll be waiting for Skylake-E...
Click to expand...

Yeah, performance wise, I still have no reason to upgrade my i7-3930k, and the need to move to DDR4 is a huge cost disincentive.

Whats making me think about it - however - is that the plastic parts on my Asus p9x79 WS motherboard are starting to dry out and become brittle. I have thus far broken off one PCIe clip, and snapped a RAM clip in half. Luckily in both cases everything still works, but I guess I have to be careful not to jiggle the rig too much.

This has never been an issue before I bought the 3930k and motherboard at launch in late 2011. With the longer and longer longevity of these parts (read, no reason to upgrade) motherboard makers are going to have to start considering using different - longer lasting - plastics on their boards, or risk attracting the ire of users.
 
I just don't understand why this is even an issue for them. In a sealed setup why can't they just use some ultra fine ceramic or other TIM. I'm dropping 3-500 for a processor if not 3x that for the X chips when/if they are released for Skylake-E why can't they spend $0.25 cents on each chip for the premium TIM.

Kinda makes me want to risk capping one and seeing what can be done without it like in the old days.
 
Looks like Intel is ratcheting up the hype machine again. Anyone recall back when Devils Canyon was about to be released one of their Exec tweeted that it was going to be the overclocking champ but turned out to be a chump and he had to take it back?
 
Oh I know about the IPC improvements but they were never huge, hence why a 2500k/2600k is still viable. Even so, Skylake sounds pretty good....we shall see.
 
lollerwaffle said:
So my SB has been running (conservatively but cool) at 4.4GHz for four years. If I can get a Skylake running at 4.6GHz, that's a 4 generation jump (SB to IVY to HASWELL to BROADWELL to SKYLAKE). Let's generously assume 10% jump in IPC each, along with a little extra core speed (200MHz or about 5%) to total 45% or so and a slew of new instructions that by now might even be in use (in this case new are F16C, AVX2, FMA3) and it might be a nice improvement.

I'd love to see some SB-Skylake comparisons when the new chips come!
Click to expand...

Not to be a dick but it'd be 1.1^4*1.05 = 54% increase :)
 
Lots of drooling and speculation in this thread. Hopefully I'll see some of those sweet 2600k in the FS forums soon.
 
CompMage said:
I just don't understand why this is even an issue for them. In a sealed setup why can't they just use some ultra fine ceramic or other TIM. I'm dropping 3-500 for a processor if not 3x that for the X chips when/if they are released for Skylake-E why can't they spend $0.25 cents on each chip for the premium TIM.

Kinda makes me want to risk capping one and seeing what can be done without it like in the old days.
Click to expand...

Already answered this above.

Zarathustra[H];1041742904 said:
I can't help but think that the usage of paste by Intel was a "planned obsolescence" move. They know that the paste is going to degrade over time, and when that happens, the chip will run hotter and hotter until it starts throttling itself, feels "slow" and people will assume their computers have become old and slow, and buy new ones.
Click to expand...
 
doh cpus these days only use 0.835 volts? i'm still using an i7 920 (nehalem) lol. i'm waiting for skylake because this rig needs an urgent upgrade. got a current ssd and decent video card at least, though. good to see that there may be potential with this cpu.
 
It's so funny seeing people talk about whether or not this is worth the upgrade from SB/IB systems, and here I am with my Phenom II X2 555, just drooling away.. :D
 
Araxie said:
yes but add 10% IPC increase to Ivy Bridge and then 10% more to haswell.. Skylake in every leak are reported to add a huge 20-30% IPC increase.. so 40-50% Increased IPC from sandy bridge to skylake should be considered as a major performance increase so at 5GHZ sandy vs skylake should be a huge difference.
Click to expand...

I as anyone else share your enthusiasm of wanting higher IPC. Those 20-30% Skylake IPC increases so far that at least I've seen have been on the multimedia benchmarks, which says to me that could likely be due to the improved graphics. All the other benchmark improvements were pretty small in comparison. Though anyone running Ivy Bridge or former has a good incentive to upgrade IMHO since it wouldn't cost much after selling off old gear, sometimes hardly anything at all.
 
Modboxx said:
Looks like Intel is ratcheting up the hype machine again. Anyone recall back when Devils Canyon was about to be released one of their Exec tweeted that it was going to be the overclocking champ but turned out to be a chump and he had to take it back?
Click to expand...

Remember when AnandTech was wall-to-wall "ARM might as well file for bankruptcy now, 'cause once Haswell is released, they're done."?

Haswell was even going to bring us world peace.
 
So many things nowadays' cry wolf.....

Francois.png


Have heard this story before. I will believe it when I see it
 
Gigahertz isn't everything. How will it perform in real-world testing overclocked? Only time will tell...
 
Zarathustra[H];1041742939 said:
Yeah, performance wise, I still have no reason to upgrade my i7-3930k, and the need to move to DDR4 is a huge cost disincentive.

Whats making me think about it - however - is that the plastic parts on my Asus p9x79 WS motherboard are starting to dry out and become brittle. I have thus far broken off one PCIe clip, and snapped a RAM clip in half. Luckily in both cases everything still works, but I guess I have to be careful not to jiggle the rig too much.

This has never been an issue before I bought the 3930k and motherboard at launch in late 2011. With the longer and longer longevity of these parts (read, no reason to upgrade) motherboard makers are going to have to start considering using different - longer lasting - plastics on their boards, or risk attracting the ire of users.
Click to expand...

Plastics like we were used to in the past are gone, my Friend.

There was a global directive to reduce the use of Lead in all things, called RoHS.

Plastics of all kinds used lead compounds to make the plastic more flexible, and more importantly, UV Resistant. :)

Take any plastic bottle you have in your kitchen, set it outside in the sun, and it will be brittle in a week.

Find a 20 year old one in your basement, and try the same test.

That's why your stuff is breaking; if you have nice purple or CCFL lights in your case, it's actually your fault. :)

Sorry. ;)

I had to deal with this from the soldering aspect; life sucked thru the late 00's.
 
DrBorg said:
Plastics like we were used to in the past are gone, my Friend.

There was a global directive to reduce the use of Lead in all things, called RoHS.

Plastics of all kinds used lead compounds to make the plastic more flexible, and more importantly, UV Resistant. :)
Click to expand...

Hey that's some interesting info buddy. I always assumed RoHS was about metallic parts, didn't have a clue plastics also cotained Lead. Thanks for sharing!
 
DrBorg said:
Plastics like we were used to in the past are gone, my Friend.

There was a global directive to reduce the use of Lead in all things, called RoHS.

Plastics of all kinds used lead compounds to make the plastic more flexible, and more importantly, UV Resistant. :)

Take any plastic bottle you have in your kitchen, set it outside in the sun, and it will be brittle in a week.

Find a 20 year old one in your basement, and try the same test.

That's why your stuff is breaking; if you have nice purple or CCFL lights in your case, it's actually your fault. :)

Sorry. ;)

I had to deal with this from the soldering aspect; life sucked thru the late 00's.
Click to expand...

Ahh, I am familiar with RoHS through work, but I didn't know it stretched beyond solder and electrical componentry. I had no idea lead was a common additive in plastics.

No, I have no lights in my case, but it did come with a case window which has probably let some UV light in over the years. It's funny, because personally I would have preferred a windowless case, but none of the high end cases at the time came without Windows. (I guess everyone is a show-off these days)
 
Is still has a 100Mhz bus speed. Welcome to 1999. Meantime my silly 8350 is running 250Mhz bus speed.

Frankly I'm underwhelmed. :rolleyes::eek:
 
You must log in or register to reply here.
Back
Top