Kaby Lake - 5.0 GHz at 1.35 Volts reported.

[OneWhoKnocks53 51]

And yes, Kaby Lake is no different. Report not long ago of a 30C drop in temp after delidding.

 

Figured as much, thanks! Guess it's time to go for the 5ghz club!

[sizziano 172]

 

 

Great result! Lucky boy. And to be honest there's no much that makes use of HT.

 

Then why get an i7? There are tons of applications that make use of HT including XP.

[martin-w 3,278]

Then why get an i7? 

 

 

The 7600K has a lower base clock and a lower turbo frequency than the 7700K. The 7600K seems to be reaching 5 GHz quite nicely too though. Nothing wrong with opting for the 7600K if we think we will never make use of HT.

 

On the other hand, "Laserit, Floyd" may have felt that by opting for the 7700K, that HT is there if he ever needs it.

 

 

There are tons of applications that make use of HT including XP.

 

 

 

If  running one of those "tons of applications that make use of HT" as you put it, then get the 7700K and leave HT enabled.  

 

It's personnel choice. We all have different needs. What I do know, is that when I ran FSX, and disabled HT, I noticed zero difference in performance. Games ran no different either. If it's different for you, go for what you require.

[ComSimPilot 623]

I've now got her @5.15GHz with 1.36v 

 

top core temp = 76deg

top package temp =78deg

 

no HT

 

Cooling with a Corsair 100i v2

 

I'm on the same cooler. What is your temps if you leave HT on on these voltage? I'm getting over 90..

 

Don't want to turn HT-off as I use it to gain smoothness in P3D.

 

I suspect mine has bad TIM issues and will probably return it to try another 7700K, but would like to compare same cooler same voltage.

[martin-w 3,278]

Deleted

[lownslo 592]

And to be honest there's no much that makes use of HT. 

 

Regarding our simulators, this depends.  I use FSX w/HT on and an affinity mask of 212.  On core 1 I run Active Sky, and on core 5 I run the GTN (except w/ the Turbine Duke where I run one GTN on 5 and another on 3).  I also run a bunch of Orbx as well as A2A and RealAir.

 

When I ran an i5 CPU I had to deal with random stutters (I presumed caused by threads colliding) but those stutters are gone now with an i7 that I try to manage well.  YMMV.

 

Greg

[laserit 93]

I'm on the same cooler. What is your temps if you leave HT on on these voltage? I'm getting over 90..

 

Don't want to turn HT-off as I use it to gain smoothness in P3D.

 

I suspect mine has bad TIM issues and will probably return it to try another 7700K, but would like to compare same cooler same voltage.

 

I've got it stable @4.95 1.39v with HT on and a max temp of 83c

 

Beyond that it seems to get finicky. 

[sizziano 172]

The 7600K has a lower base clock and a lower turbo frequency than the 7700K. The 7600K seems to be reaching 5 GHz quite nicely too though. Nothing wrong with opting for the 7600K if we think we will never make use of HT.

 

On the other hand, "Laserit, Floyd" may have felt that by opting for the 7700K, that HT is there if he ever needs it.

 

 

 

 

If  running one of those "tons of applications that make use of HT" as you put it, then get the 7700K and leave HT enabled.  

 

It's personnel choice. We all have different needs. What I do know, is that when I ran FSX, and disabled HT, I noticed zero difference in performance. Games ran no different either. If it's different for you, go for what you require.

 

Sure but FSX is also over a decade old at this point. New AAA games will defin tely make use of the extra threads as evidenced by superior minimum frames and frame times with an i7 over an i5 even if averga e FPS remains the same.

[ComSimPilot 623]

I've got it stable @4.95 1.39v with HT on and a max temp of 83c

 

Beyond that it seems to get finicky.

 

I actually have a relative good chip, but bad insulation it seems betweed the die and cover. It is stable at 4.8 Ghz and 1.3v and for 5 Ghz at 1.38v. These with HT on. Problem is temps are max 81C and 96C accordingly.

 

With HT off I can get it stable at 5.1 Ghz and 1.365v. Max temp 87C

 

So, comparing with yours it seems my temperatures are high as we have the same cooler.

 

I'll return it hoping the new one has not worse performance and better TIM insulation.

Just to warn future byers of the 7700K that this is possibly an issue on some chips, reading also on the net.

[gboz 157]

 

 

The stock TIM between the die and the heat spreader is worse than I ever imagined!

 

Do you mean the way it physically looked when you delided?

 

Thanks,

gb.

 

 

 

After the delid - 4.8GHz @ 1.32vCore - 54C average per core, 56C CPU package.

 

Did you try to overclock any further?

Seems to be plenty of voltage/temp headroom there.

 

gb.

[OneWhoKnocks53 51]

Do you mean the way it physically looked when you delided?

 

No, although it was the consistency of chalk. What I meant was the performance delta.

 

Did you try to overclock any further?

Seems to be plenty of voltage/temp headroom there.

 

Not yet, but I agree. With these numbers 5ghz should be within reach.

[TheFlightSimGuy 514]

I run 5ghz on my I7 6700k with 1.36v to the CPU. I just don't see any need for any current gen folks to upgrade and for those looking to update an older system, you can probably find Skylake mobo's and CPU's for a discount now. This departure from "tick/tock" for Intel is rather uninspiring, hopefully AMD RYZEN will change that. 

[sizziano 172]

I thought this was relevant to the discussion of voltage versus temp:

 

[ComSimPilot 623]

I run 5ghz on my I7 6700k with 1.36v to the CPU. I just don't see any need for any current gen folks to upgrade and for those looking to update an older system, you can probably find Skylake mobo's and CPU's for a discount now. This departure from "tick/tock" for Intel is rather uninspiring, hopefully AMD RYZEN will change that.

 

I'd say it depends on your 6700K. You have a very lucky one. Most people run the 6700K at 4.6, 4.7 and a few at 4.8 as far as I know. My Skylake could not do more than 4.6 Ghz at 1.41v. In that way a 400-500 Mhz upgrade is worth it in my opinion. Going from 27-28 to 30 fps makes a difference to me. Of course whats worth it or not is also a personal preference.

[martin-w 3,278]

 

Regarding our simulators, this depends.  I use FSX w/HT on and an affinity mask of 212.  On core 1 I run Active Sky, and on core 5 I run the GTN (except w/ the Turbine Duke where I run one GTN on 5 and another on 3).  I also run a bunch of Orbx as well as A2A and RealAir.

 

When I ran an i5 CPU I had to deal with random stutters (I presumed caused by threads colliding) but those stutters are gone now with an i7 that I try to manage well.  YMMV.

 

Greg

 

 

Hi Greg,

 

I Shouldn't comment too much on flight sim, as I don't run it these days, so it wouldn't be fair. But what I can say is that there is no definitive metric that enables us to measure smoothness and stutters. So we are left with subjective opinion, stuff like, "do you know I think that looks smoother", and "I'm pretty sure I see less stutters". It's very easy to be fooled by the placebo effect. Having said that, subjective opinion is all we have, so if it works for you and you're convinced, fair enough.

 

What I will say is that in your case, you didn't just switch from HT off to HT on, you also upgraded your PC from an i5 to an i7. You don't say if you overclocked or not, and how much with the i5 and how much with the i7 [that too is a variable.] but the i5 has a lower base frequency and turbo frequency. And of course, you may have switched to a new architecture entirely, new, faster RAM, new chipset etc.  My point is that there's a myriad of variables at play. As I said though, if you're convinced, fair enough.  :smile:

Sure but FSX is also over a decade old at this point. New AAA games will defin tely make use of the extra threads as evidenced by superior minimum frames and frame times with an i7 over an i5 even if averga e FPS remains the same.

 

 

 

This is a nice review from 2016. Gaming benchmarks with HT on and HT off. Worth noting that none demonstrated an increase in performance and in fact some  some demonstrated a significant decrease in performance....

 

https://www.techpowerup.com/forums/threads/gaming-benchmarks-core-i7-6700k-hyperthreading-test.219417/

 

In regard to newer games that you say "will definitely" be assisted by HT on: Then Floyd [laserit] did the right thing by opting for an i7, as he can indeed enable HT if he ever needs to.

 

There is something else we need to consider, re HT off or on, in my opinion. Namely, with HT off the CPU runs cooler, enabling us to overclock higher. So the question to ask is if the higher overclock is more advantageous to us than HT on, or is HT on and a lower overclock more advantageous to us..  

I actually have a relative good chip, but bad insulation it seems betweed the die and cover. 

 

 

Not to be pedantic, but it's not "bad insulation". On the contrary, we "want" bad insulation and good thermal conductivity. The purpose of TIM is to eliminate the air gap between die and IHS, or IHS and cooler, and increase thermal conductivity..

No, although it was the consistency of chalk. What I meant was the performance delta. 

 

 

 

Actually, it may look like chalk, but it's not horrendous TIM. I recall it's Dow Corning TIM. The issue is more to do with the thick layer of adhesive Intel use to glue the IHS on. The thick layer of adhesive then necessitates a thick layer of TIM. And as you know, the thinner the TIM the better.

 

Clearly Liquid Pro or Ultra is superior though. But there's never a huge difference between different brands of TIM.