Intel Skylake / Kaby Lake

[dahorns]

Conspiracy..??.lol, i guess that you have no argument, hence this ridiculous straw of yours.

If you did look at the other stress tests you ll see that the max power is exactly the same with 3 decimals, that s 30.263W..

Isnt it extraordinary that the system consume exactly the same max power with such a precison in two different tests..?..

A hint, they just recycled the other windows.

Again, what does any of that have to do with your claim that the prime 95 test alone is doctored? Something is pegging the cpu. And you have a system power consumption number max of 28.812. I really can't understand what point you could possibly be making.

As to why the max power consumption numbers would be the same in the other two screenshots, the obvious answer is because it is part of the same test run. The first image is at 4 minutes and the second at 19 minutes.

 

[Abwx]

No, it definitely does.

If you have Windows auto-hide system tray icons, it can be disturbingly easy to accidentally leave P95 stress test running on your computer for a weekend.

I have the auto-hide implemented, once i close the worker the logo get hidden and there s no more computations..

Anyway when a reviewer do a test he should display the testing tools, unless of course that he s assuming that we must take his claims at face value..

Would do you post a test without displaying that the test is effectively done..??

Besides, look better, nothing is hidden at the time he did the screen capture, also look at the "duration" of the data acquisition in HVinfo : 2 minutes that it was opened :

And of course he neglected to rename this recyclage because the pic is named :

http://www.notebookcheck.com/fileadmin/_processed_/csm_SP4_i5Furmark_solo_7192009965.png


LOL of a LOLZ...

 

[MrTeal]

I have the auto-hide implemented, once i close the worker the logo get hidden and there s no more computations..

Anyway when a reviewer do a test he should display the testing tools, unless of course that he s assuming that we must take his claims at face value..

Would do you post a test without displaying that the test is effectively done..??

Besides, look better, nothing is hidden at the time he did the screen capture, also look at the "duration" of the data acquisition in HVinfo : 2 minutes that it was opened :

What version of P95 are you using? I have 28.7b1, and it definitely doesn't stop the workers if you close the Window. If you have No Tray selected in the Options menu I even have to kill the process if I close or minimize the window, since it doesn't show in the task bar or tray in that case. You can't even alt-tab to it.

 

[Enigmoid]

I have the auto-hide implemented, once i close the worker the logo get hidden and there s no more computations..

Anyway when a reviewer do a test he should display the testing tools, unless of course that he s assuming that we must take his claims at face value..

Would do you post a test without displaying that the test is effectively done..??

Besides, look better, nothing is hidden at the time he did the screen capture, also look at the "duration" of the data acquisition in HVinfo : 2 minutes that it was opened :

And of course he neglected to rename this recyclage because the pic is named :

http://www.notebookcheck.com/fileadmin/_processed_/csm_SP4_i5Furmark_solo_7192009965.png


LOL of a LOLZ...

Or two minutes since it was refreshed as the min load values are 100% (HW monitor either started or was refreshed when the CPU was pegged @100%). No clue how long it was running beforehand. Temperature looks fairly steady as well.

 

[Phynaz]

Snip...

Skylake i7 multitasks 237.5% better than any AMD CPU.

AMD certainly makes better space heaters than Intel.

 

[Phynaz]

Honestly, I can't even figure out the point to your line of attack. Let's say prime 95 isn't running, so what? Clearly something is running that has the processor pegged and the igpu idle. And whether that something is prime 95 or the unicorn fart simulator, isn't pegging the cpu the point of the test? What is the point of your ridiculous conspiracy?

Bizarre, isn't it? How does someone become this emotionally invested? I don't know, maybe his meds stopped working again.

I'm wondering if a permaban is the next stop after the month long vacation he received.

 

[Abwx]

Skylake i7 multitasks 237.5% better than any AMD CPU.

AMD certainly makes better space heaters than Intel.

Lol, hollow sentences and now hollow numbers...

Here the reality, CB ST in a 6700K get from 181 when tested in isolation to
86 when added as a 9th thread over an Integer app that use all 4 cores with 8 threads.

FX CB ST in isolation is 93 and get to 61 when added as a 9th thread over an Integer app that use all 8 cores with 8 threads.

No wonder that in an era of muticore CPUs the usual suspects deseperately insist on ST perf in isolation, they know too well that this perf is a myth as proved by thoses numbers, look like they are not interested in how their multicores CPUs performs in real use, they look at the ST perf and think that it will scale linearly with the number of apps and threads.

FTR the 3770K get to 71, and when i think that people thought that it was better than the FX, lol...

 

[dahorns]

Lol, hollow sentences and now hollow numbers...

Here the reality, CB ST in a 6700K get from 181 when tested in isolation to
86 when added as a 9th thread over an Integer app that use all 4 cores with 8 threads.

FX CB ST in isolation is 93 and get to 61 when added as a 9th thread over an Integer app that use all 8 cores with 8 threads.

No wonder that in an era of muticore CPUs the usual suspects deseperately insist on ST perf in isolation, they know too well that this perf is a myth as proved by thoses numbers, look like they are not interested in how their multicores CPUs performs in real use, they look at the ST perf and think that it will scale linearly with the number of apps and threads.

FTR the 3770K get to 71, and when i think that people thought that it was better than the FX, lol...

So the i7 can run CB while also performing unpacking with winpack at the same speed as the FX without winpack? Is this suppose to be bad? And, as far as I can tell, that article doesn't mention anything about how the winpack run was affected. Presumably the i7 is doing more work in a quicker time there as well.

 

[Abwx]

So the i7 can run CB while also performing unpacking with winpack at the same speed as the FX without winpack? Is this suppose to be bad? And, as far as I can tell, that article doesn't mention anything about how the winpack run was affected. Presumably the i7 is doing more work in a quicker time there as well.

In the combined test the CPU are comparable for the Integer throughput but the i7 does better in the CB score, wich is not that telling as if CB is replaced by Povray for instance the difference will be much smaller for the FP load.

HW doesnt have this advantage as its FP perf is not as good as SKL since it s about the level of the FX in Povray, to get at the same level than SKL in FP the FX should had been upgraded with Steamroller cores or better, with EXV cores to get frequency parity.

 

[Phynaz]

So the i7 can run CB while also performing unpacking with winpack at the same speed as the FX without winpack? Is this suppose to be bad? And, as far as I can tell, that article doesn't mention anything about how the winpack run was affected. Presumably the i7 is doing more work in a quicker time there as well.

It's great when the adf gets owned.

 

[Abwx]

Beware of cooler compatibilities :

The Skylake PCB is about 0.78 mm thick and will bend if you only apply pressure on the die (Haswell PCB is around 1.17 mm thick). So you also need the IHS to keep the PCB straight for a perfect connection to the pins in the LGA socket.

http://www.computerbase.de/2015-11/intel-skylake-scythe-warnt-vor-schaeden-und-ruestet-kuehler-um/

 

[DrMrLordX]

Interesting article. Scythe is already offering different mount screws to compensate.

Anyone had problems from the thin Skylake PCB yet? I haven't heard of any.

 

[AtenRa]

Interesting article. Scythe is already offering different mount screws to compensate.

Anyone had problems from the thin Skylake PCB yet? I haven't heard of any.

I have seen photos of Skylake PCBs experiencing a plastic deformation. But those were on delided CPUs.

 

[crashtech]

A square shim of the proper thickness that takes the place of the removed IHS adhesive will fix that for delidded applications.

 

[jhu]

Bizarre, isn't it? How does someone become this emotionally invested? I don't know, maybe his meds stopped working again.

I'm wondering if a permaban is the next stop after the month long vacation he received.

Is that what happened? Was wondering why it was so quiet in this sub.

 

[Sweepr]

For those of you interested in low-end Skylake, some Pentium G4400 tests from German PCGamesHardware.

Very close to quad-core 3.9-4.2GHz FX4300 MT performance on the synthetic tests. Not bad for a power efficient dual-core (no HT) CPU at 3.3GHz. Also faster than FX6300 in 3 out of 4 games tested.

Wish they had tested the Pentium G4500. A couple extra bucks give you +200MHz and HD Graphics 530. Worth it IMHO.

 

[witeken]

Somewhat off-topic.

I though the gate pitch of a node was the distance between de gates of two transistors. However, this image shows something different:

Then why is it called the gate pitch?

 

[NTMBK]

Somewhat off-topic.

I though the gate pitch of a node was the distance between de gates of two transistors. However, this image shows something different:

Then why is it called the gate pitch?

It looks like the distance between the corresponding point on two gates, to me?

 

[coercitiv]

I though the gate pitch of a node was the distance between de gates of two transistors. However, this image shows something different

Draw a line between two gates, compare it with the line in the pic: they'll be the same. Distance between any 2 identical features in an array of identical elements is the same.
eg: [string]-[string] distance between any 2 letters is identical.

 

[zir_blazer]

Does a Base Clock mod for non-K Skylakes sweetens your ears? The guy doesn't seem to have shared how he did it, but that its possible.



Also, CPU World isn't posting News anymore. Instead, we got this:

Regarding news articles: I had to remove content from two articles at Intel's request in the past two months. It's not worth to research and post new information if we have to take it down shortly afterwards, therefore news stories are currently on hold.

Do anyone knows what that could be, or if its related to Skylake?

 

[ShintaiDK]

Most likely posting unreleased news as a first poster.

 

[MiddleOfTheRoad]

Skylake i7 multitasks 237.5% better than any AMD CPU.

AMD certainly makes better space heaters than Intel.

That is total nonsense.

Intel Core i7-6700K:
Passmark score = 11021 / Single Threaded Score = 2321

AMD FX 9590:
Passmark score = 10265 / Single Threaded Score = 1721

What else are you planning to lie about today?

 

[LTC8K6]

That is total nonsense.

Intel Core i7-6700K:
Passmark score = 11021 / Single Threaded Score = 2321

AMD FX 9590:
Passmark score = 10265 / Single Threaded Score = 1721

What else are you planning to lie about today?

Intel Core i7-4790K 11227 / Single Thread Rating: 2530

The 6700K is worse than Haswell...so he ought to be bragging about Haswell...

 

[ShintaiDK]

LOL@Passmark. Why do people use that again?

They dont separate between stock and OC either. So its all one big pile of rubbish.

 

[LTC8K6]

LOL@Passmark. Why do people use that again?

They dont separate between stock and OC either. So its all one big pile of rubbish.

Well, the 237.5% post was a big pile of rubbish...