Skylake-X (Core i9) - Lineup, Specifications and Reviews!

Take off the stock heatsink and point a 120mm fan at them. Or better yet, put on some real heatsinks (like these ones) and also get some airflow over them.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
n=1 said:
Take off the stock heatsink and point a 120mm fan at them. Or better yet, put on some real heatsinks (like these ones) and also get some airflow over them.
Click to expand...
those are the ones i was looking at. Gonna wait until board is here on monday and assess the situation before ordering anything.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
cageymaru said:
Got one of those in the bone pile somewhere. :)
Click to expand...

I own 3.I still have the boxes.
You can drop a push-pull fan, true120 copper onto your stove, it will dissipate it NOOOOOO problem.
I am still trying to find the upper limit of a TRUE120 copper.I bet it'll take 1000 watts.
On air.
Fatasses.
:ROFLMAO:
 
The reason it's "4750 MHz average OC on air" is because of this one single entry on air. (yes ONE SINGLE ENTRY, n=1 literally)

Come the fuck on. (seriously, as of this post, every single entry in every single benchmark is a >5 GHz LN2 bench)

And then it says "Air (Custom)" to boot, whatever the hell "custom air" is.
 
juanrga said:
I am giving average clocks; ergo some chips will overclock better and others will overclock worse. Some 10-core chip has been pushed up to 5GHz on air (with delidding)
Click to expand...

You are not hitting a 4.7ghz on that chip for anything more then a suicide bench run. I could give you my custom phase change cooling rig and you couldn't keep that chip cool at 5ghz on a bare die for anything more then a bench run or two. I mean you would need 500 watts just for the damn chip to work at 5ghz. This is why no one believes anything you say on here.
 
Nihilus1 said:
As expected. A couple of nerds did suicide benchmarks on air and this is now the "average o/c on air." Your post was blatently misleading. Your hope was to convince a few lurkers that you can slap on a good air cooler and run the 7900x at 4.7 ghz. Total BS
Click to expand...

"i bought muh 7900X coz some random d00d on teh interwebz saiz it would overclock to 4.7 on air!!!111"

Honestly if that's enough to convince some lurker to buy a 7900X, well then said person deserves to be parted with their money.
 
True, I agree forums are not the best place for research. Everyone is bias to some point even cherry picking information. Even then, it is still information. The 4.75 ghz average on air was a complete joke, though.
 
Nihilus1 said:
As expected. A couple of nerds did suicide benchmarks on air and this is now the "average o/c on air." Your post was blatently misleading. Your hope was to convince a few lurkers that you can slap on a good air cooler and run the 7900x at 4.7 ghz. Total BS
Click to expand...

Stop the BS. A pair of days ago I used the same database when referring to RyZen average overclocks, which I posted in another thread. It is a good database to get a rather accurate idea of how chips overclock on general.

n=1 said:
The reason it's "4750 MHz average OC on air" is because of this one single entry on air. (yes ONE SINGLE ENTRY, n=1 literally)

Come the fuck on. (seriously, as of this post, every single entry in every single benchmark is a >5 GHz LN2 bench)

And then it says "Air (Custom)" to boot, whatever the hell "custom air" is.
Click to expand...

A single entry of 4800MHz does not provide an average of 4750MHz. But irrelevant discussion now because the database has been updated. Average overclocks now are 4270MHz on air and 5633MHz on LN2. Data for watter cooling continues unavailable. We will have to wait a bit more until more submissions reduce the statistical fluctuations.
 
Last edited:
If HWBot is anything like it was 3 years ago, sometimes the data for the old entries that get replaced/deleted still lurk around in the database and affect the average. My guess is that one single air entry in Cinebench R15 was not the first and only submission from that user.

The HWBot database can be useful if there's a large enough sample size, but we're not at that point yet for Skylake-X.

In any case, I think we can at least acknowledge that 4.7 OC on air was a premature statement.
 
n=1 said:
If HWBot is anything like it was 3 years ago, sometimes the data for the old entries that get replaced/deleted still lurk around in the database and affect the average. My guess is that one single air entry in Cinebench R15 was not the first and only submission from that user.
Click to expand...

There are entries from other users, such as this submission at 4.2GHz

http://hwbot.org/submission/3588493_massman_hwbot_x265_benchmark___4k_core_i9_7900x_19.13_fps

n=1 said:
The HWBot database can be useful if there's a large enough sample size, but we're not at that point yet for Skylake-X.
Click to expand...

That is what I wrote in the post just above the yours.

n=1 said:
In any case, I think we can at least acknowledge that 4.7 OC on air was a premature statement.
Click to expand...

Some people is getting 4.7GHz on air, maybe 4.7GHz as average was premature, but we cannot acknowledge it is not possible. Not still. Also remind der8auer words:

On average CPU you might be able to get 4.7-4.8 GHz if you delidded it and on a pre tested CPU 5 GHz should be possible.
Click to expand...
 
Last edited:
Yes, and those entries were submitted after I'd already written my post, which is why the average air OC went down significantly.

If we're going to nitpick semantics, then that "some people" getting 4.7 OC on air is more like one single person, who was running custom air (whatever the hell that means). So it is by no means the norm, and 4.7 average OC on air is definitely premature. I mean for example, let's say I strapped a custom 5 pound copper heatsink with some 4000 rpm Delta fans attached to it, sure that counts as "OC on air", but it would be a ridiculous example to use.

As for der8auber's words, do pay attention to the word "delidded". Also in the context of that post, he was most likely referring to an AIO, not air.
 
Still doesn't change anything as far as "average OC on air" goes. But do provide a link.
 
Regarding game performance on Skylake-X, several sites found it play games worse than Broadwell-E. This is a problem with those reviews using buggy BIOS with broken turbo 3 and other aspects affecting performance on latency sensitive workloads. Updating the BIOS improves performance. Sites as Techspot tested with improved BIOS and got that the Skylake-X is so fast or slightly faster than Broadwell-E playing games

Hitman.png

Ashes.png


Sites like TT tested the i9 with both an older BIOS and with an updated BIOS and found the older BIOS was the cause of the weak gaming performance

8225_41_intel-core-i9-7900x-series-skylake-cpu-review.png

8225_38_intel-core-i9-7900x-series-skylake-cpu-review.png


The issue is that the normal 7900K results were done on older BIOS versions, while the new ones are done on the latest BIOS versions that support Turbo 3 without any software requirements. You will see results in line with the better performing 7900X results, but I do know some other media who were getting the same low gaming scores I was, and that was because Turbo 3 wasn't working. Memory increase from 2133Mhz to 2666Mhz with the same timings also makes a difference in the gaming results. Ashes of Singularity is one of our outliers, but that is most likely because they have to optimize the code for the CPU, just like had to be done for Ryzen.
Click to expand...

However, current BIOS are not still working correctly and some sites delayed their game reviews

http://www.pcgamer.com/the-ongoing-testing-of-intels-x299-and-i9-7900x/
http://www.anandtech.com/show/11550...-core-i9-7900x-i7-7820x-and-i7-7800x-tested/7
 
n=1 said:
Return that shit for a refund and wait for companies to get their shit together. If I paid $300+ for a mobo with VRMs that would blow up without user intervention I'm gonna be pissed as hell. I'm not a paid beta tester so don't push your broken shit on me.
Click to expand...

The Gigabyte Gaming 7 and 9 look to have the best VRM heatsink including a heat pipe granted they are also most expensive at $400 and $500. I went with the gaming 7!
 
Nirad9er said:
The Gigabyte Gaming 7 and 9 look to have the best VRM heatsink including a heat pipe granted they are also most expensive at $400 and $500. I went with the gaming 7!
Click to expand...
I have a gaming 7 and am fighting some weird voltage stuff. It stopped letting me set voltage manually. Anything I set lower than 1.273, it just runs 1.273. I can set higher voltage, but that doesn't help. It was working when first setup and I know my 7820x is stable at 1.225v running 4.7GHz.
 
Alright. I can call my 7820x completely stable at 4.7GHz on 1.225v. Cooling is a H115i with push/pull Noiseblocker 140mm fans. RAM is Corsair 3200MHz running stock 16-18-18-36 timing. Mesh was set to 3200MHz.

Here are the real world numbers...


In Cinebench it scored 2035cb in Multi thread and 204cb Single Core. The hottest core was #1 hitting 72c.

AvdyvQu.jpg


In Geekbench it scored 33189 Multicore and 5831 Single-core. The hottest core was #3 and #7 both hitting 67c.

lhM2zTX.jpg


In Realbench for 15 minute test the hottest core was #3. It hit 78c.

yXuu6pA.jpg


In Prime with AVX in a 20 minute test the hottest core was #2. It hit 88c. AVX offset was -500MHz) Personally I consider Prime with AVX a very unrealistic usage scenarios for me, so I am OK with this.

XY6B9fp.jpg


AIDA Cache and Memory looks like this...

li3ffC2.jpg




Just for shits and gigs here are some 3D benches with a 1080ti FTW3...

http://www.3dmark.com/fs/13034422
uaqJJAv.jpg


http://www.3dmark.com/spy/2015990
XTHgrHC.jpg


tBIOcPL.png[IMG]
 
Last edited:
Was able to Max out the Corsair h75 that came with the board at 4.4ghz. checked VRM and 8 pin temps and saw nothing over 120 degrees. The board does have a cooling plate on the bottom in the vrm area.
CPU is 7820x and board is MSI x299 sli plus
 
Nice build, if you have not you should point some active cooling at your VRM just in case, i know the VRM can handle the high temps but that might not be the same for other components close to the VRM.
 
I'm glad we have this space to share info about Skylake-X. You know, in certain forums people are immediately attacked (even by moderators) if they point out that a 4.5 GHz+ 7800X can beat Ryzen in certain productivity scenarios and CPU-limited gaming. Most people here can recognize the advantages/disadvantages of each chip and make their choices upon that, it's definitely not one-sided. (y)
 
Last edited:
To add to above comments, I want also to emphasize this forum is full of very competent users. Most posts are technical and well-informed. For instance, I cannot imagine a group of people in this forum pretending that the motherboard of 10-core SKL-X overclocked at 4.8GHz is pulling ~400W in a power virus test, because (ridiculous claim --->): "Intel lied the TDP by basically ~3x". I cannot imagine here an entire thread full of that nonsense, without no one here noticing the obvious (TDP is not power consumption at 4.8GHz), but you can find this kind of ridiculous claims in SA forums. They are already in page 2 of the thread and no one even bother to mention that 140W is not the TDP at 4.8GHz.

I am happy of being a member of [H].
 
Kato1144 said:
Nice build, if you have not you should point some active cooling at your VRM just in case, i know the VRM can handle the high temps but that might not be the same for other components close to the VRM.
Click to expand...
I checked 8 pin temps and vrm temps @4.4ghz using IntelBIT after 30 minutes. The CPU was at 95c peak on one core but the 8 pin was under 105f and vrm temps were 120f or lower. I didn't stress with avx as that isn't something that will be used. There are 2 fans in the top of the case. One pulls air in and the other is the radiator exhaust and those are right by the vrms. There is no direct cooling for them but based on my results in comparison to der8's video it's either getting cooling off that or this board doesn't have the issue he outlined.



Are tahoedust and I the only ones who got skylake-x? Are some of you guys still waiting on components to arrive?
 
juanrga said:
To add to above comments, I want also to emphasize this forum is full of very competent users. Most posts are technical and well-informed. For instance, I cannot imagine a group of people in this forum pretending that the motherboard of 10-core SKL-X overclocked at 4.8GHz is pulling ~400W in a power virus test, because (ridiculous claim --->): "Intel lied the TDP by basically ~3x". I cannot imagine here an entire thread full of that nonsense, without no one here noticing the obvious (TDP is not power consumption at 4.8GHz), but you can find this kind of ridiculous claims in SA forums. They are already in page 2 of the thread and no one even bother to mention that 140W is not the TDP at 4.8GHz.

I am happy of being a member of [H].
Click to expand...

Yep, this community is pretty great in general.
 
Last edited:
I am just now figuring out how to get the voltage offset to work how I want with my Gigabyte board. Coming from Asus, I have had some learning to do. The good new is, that now with the voltage no longer pinned at 1.235, the 7820x actually looks to be running a good bit cooler. I am 20 minutes into Prime 95 "Standard Blend" with AVX and my hottest core has been 79c. 9c cooler than it was with fixed voltage. It is actually spending much of it's time at 1.133v, which I would assume is because of the AVX offset lowering clocks.



I am going to let it run for a full hour and see how high the temps will climb. Then on to realbench.
 
i think a lot of us are waiting for TR review before dropping $1k + on skylake-x
 
Prime95 with AVX "Standard Blend" 1-hour temp results with voltage offset working correctly. Hottest core hit 81*...not bad for a AIO

eglpWhm.png


Running realbench now
 
New video from der8auer. Really informative!
Skip to 12:45 to see the Prime 95 chart where the system starts pulling over 500w at the wall which is ~400w from the CPU according to der8auer.



Izdie4H.png
 
Last edited:
As long as you OC without AVX clock offset, you are instantly running AVX512 loads in Prime95 way above stock speed by running them at actual core speed. Even if you just got MCE enabled.

At 4.5Ghz in AVX512 loads, the FP output of a 7900X is higher than 32 and maybe up to 48 SB/IB/Zen cores at 4Ghz.

He also record a 40W difference between 4 boards, just on VRM alone.

He also state the CPU itself at 4.6Ghz pulls 270W. Gigabyte being the exception at 290W.

So no, its not 400W.
 
Last edited:
cageymaru said:
New video from der8auer. Really informative!
Skip to 12:45 to see the Prime 95 chart where the system starts pulling over 500w at the wall which is ~400w from the CPU according to der8auer.



Izdie4H.png
Click to expand...


Prime95 uses AVX-512, so performance for that power is going to be quite a lot.

Anyway, I just ran Prime95 small FFT on my i7 7820X @ 4.3GHz (1.17v) with uncore @ 3.2GHz, and package power maxed out at like 210W.
 
You must log in or register to reply here.
Back
Top