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

I rail on Intel for half-assing AVX-512 offset, making overclocking 7900x borderline impossible if you are to run AVX-512 stuff.

As for this, i am okay with it because unlike AMD Intel had some decency to lift NDA together with pre-orders opening.

Just remember no one forced anyone to rush out and order 1 right away, self control is a good thing, but I agree that is the better way to do things. I think more of that had to do with the last minute bios changing, something you Intel guys are going to have to deal with as well. I imagine after a few months most issues will be ironed out tho but they made their life much harder by trying to support more then it was originally intended to.
 
He is only ok with it cause it's Intel, He railed on AMD for having issues with Ryzen tho. But this whole launch feels rushed by Intel.

If I had to care about anyone's opinion or sentiments my life would be terrible, he is entitled to his opinions by free will.

You get tough crowds, such is life.
 
It has everything to do with that. Of course, it would require that you open your eyes, remove your Intel colored fanboy glasses, and set down the drumsticks that you've nearly worn out beating your Intel drum. Objectivity is impossible for you and your trio of loyal compadres, isn't it?

ns in the graph given by lolfail9001 is a SI sorthand for nanosecond, which is a unit of time. The graph is measuring latencies. I can understand your initial confusion (I guess you saw red/yellow colors and believed it was a measurement of tempts), I could even accept you confound temperatures with latencies, but what I cannot accept is that you attack people only because they pointed to your epic fail.
 
It is almost like the criteria for judging CPU's today is overclocking.

Funny when only some few months ago certain camp was pretending that RyZen could hit 5GHz on air and destroy a 10-core Broadwell...

C43uh18WEAEROnq.jpg


Now the same camp has changed the criteria for judging CPUs to moar lanes!
 
Funny when only some few months ago certain camp was pretending that RyZen could hit 5GHz on air and destroy a 10-core Broadwell...

C43uh18WEAEROnq.jpg


Now the same camp has changed the criteria for judging CPUs to moar lanes!

Again they were entitled to their opinions, doesn't make it any different, its just comical techsites on release day. People choose their high ground and stick with it.
 
This has been demonstrated false again and again, with the '95W' AMD chips consuming more power than the 140W chip from Intel and performing worse.

Yes yes we know at 4.1ghz the power spirals beyond what the node can handle without leakage, old news, in normal operating parameters of 14nm LP it has great efficiency. Is this actually relevant now?
 
I have hard time finding these numerous sites which show 1800X consuming more than 6900K at stock frequencies. Can someone Google those for me because most sites I checked have 1800X under the 6900K in the power consumption graphs.

We all already know 1800X consumes more than its rated tdp but I guess it only matters because it's from AMD. But it's just a fluke, measuring error, wrong kind of measuring method etc. When Skylake X also goes over it's rated tdp by hefty amount, according to some reviewers.

Dem double standards :whistle:
 
I have hard time finding these numerous sites which show 1800X consuming more than 6900K at stock frequencies. Can someone Google those for me because most sites I checked have 1800X under the 6900K in the power consumption graphs.

We all already know 1800X consumes more than its rated tdp but I guess it only matters because it's from AMD. But it's just a fluke, measuring error, wrong kind of measuring method etc. When Skylake X also goes over it's rated tdp by hefty amount, according to some reviewers.

Dem double standards :whistle:

don't you know that being under the 6900K/6950X even 6800K is regarded as being less efficient, wait for the spin on it, that is where the funny starts. I don't actually care, I just do it for the lulz.

getgraphimg.php
 
This has been demonstrated false again and again, with the '95W' AMD chips consuming more power than the 140W chip from Intel and performing worse.

Good thing I go by tech sites where at stock it's more efficient. The problem when you lie is you have the chance of actually believing your own lie after awhile.
 
so the 1800X is in between a 6900/6950X and 7700K in terms of efficiency, okay that is tragic. Now tell me why are you even telling me this, if I really wanted to know this I would google information, but more to the point what has this got to do with Skylake X?

The only tragedy is people repeating the same false claims again and again as here, maybe in the hope that repeating the same stuff a hundred of times it will turn to be true.
 
Good thing I go by tech sites where at stock it's more efficient.

Those tech sites that have fallen in the trap of using Prime95, which stresses Intel chips up to the limit, but for some reason doesn't stress RyZen chips? Many RyZen reviews found that RyZen consumes same power under ordinary workloads as Luxrender or Excel than under power virus as Prime95.

Here you have the superefficient '95W' 1800X RyZen consuming either the same power than a 140W Broadwell 6900k (on Prime95) or 20% more power (on Excel)

Power_02.png


Power_01.png


No one complained when 8-core RyZen @4.1GHz consumed 280W. But similar numbers for 10-core SKL are a problem now. Double standard again.
 
Last edited:
It seems we will have to wait to see proper reviews, because the BIOS/firmware issues still remain

http://www.pcgamer.com/the-ongoing-testing-of-intels-x299-and-i9-7900x/

At least reviews aren't falling in the moar lanes trap:

But do you need more than 28 PCIe lanes? For most gamers, the answer is easy: nope. Only dual GPUs really benefit, and even then the benefit is pretty small. And there are plenty of games that don't even properly support multiple GPUs, making that an increasingly niche market.
 
Those tech sites that have fallen in the trap of using Prime95, which stresses Intel chips up to the limit, but for some reason doesn't stress RyZen chips? Many RyZen reviews found that RyZen consumes same power under ordinary workloads as Luxrender or Excel than under power virus as Prime95.

Here you have the superefficient '95W' 1800X RyZen consuming either the same power than a 140W Broadwell 6900k (on Prime95) or 20% more power (on Excel)


No one complained when 8-core RyZen @4.1GHz consumed 280W. But similar numbers for 10-core SKL are a problem now. Double standard again.

I don't care about power consumption at all; just buy more cooling and a bigger power supply is my motto. But if you do care as some inevitably will; here is a test of the new CPUs. First they do power consumption with Excel, Cinebench and then OC it. I think 4.6GHz using 400w on the 7900x is fine as it can do a ton of work. Just need to spend a few bucks on a custom water cooling loop if you want to OC which I think is acceptable. The Corsair 240mm water cooling unit wasn't enough as it was still hitting 90c; going to need some big rads for 4.6GHz OC! :)

Start the video at 13:05.

 
Last edited:
I
I ordered a 7820X, 64GB of ram (4x(2x8GB 3200), and a 1TB 960 evo

I figured my main video card platform (really that is why my box exists) has been doubled in cores, storage, storage speed, and memory.
I mainly game so I'm planning to get a 7820x with 4x4gb DDR4 4000.
 
I don't care about power consumption at all; just buy more cooling and a bigger power supply is my motto. But if you do care as some inevitably will; here is a test of the new CPUs. First they do power consumption with Excel, Cinebench and then OC it. I think 4.6GHz using 400w on the 7900x is fine as it can do a ton of work. Just need to spend a few bucks on a custom water cooling loop if you want to OC which I think is acceptable. The Corsair 240mm water cooling unit wasn't enough as it was still hitting 90c; going to need some big rads for 4.6GHz OC! :)

Start the video at 13:05.



Ouch
 

Solid review. Some things to take away:

I know many don't care about power here, but 400 watts overclocking is full retard mode, especially when over 90* with good cooling.

4.3 GHz seems to be more of the limit for SKL-X as far as a daily overclock. How would this compare to BW-E at the same clock, which some people run? SKL-X seems to be abke to overclock high, but the 'daily overclocks' seem to be far lower than the 'max overclocks'. This doesn't seem to be the case for BW-E and Ryzen.

Why only 2666 mhz ram on BW-E and 3200 on SKL-X? I don't think it matters but I was curious.

Overall the review gave some praise for performance but was still critical of platform value, power, and heat.
 
A fresh review on the i7 7740x:



long story short, want a fast quad core CPU? get a i7 7700k on a good Z 270 board

Also from what i have read/seen so far on on the Skylake-X CPU's is nearly a double in memory bandwidth over Ryzen and (not so much) over Broadwell-E and about the same IPC as Kabylake according to Cinabench R15 results done by Hardware unboxed, the X299 platform will be a major power house for high end workstations that need the massive amounts of memory/memory bandwidth and Kabylake IPC with massive amounts of multi threading, ya man X299 is a serious response to AMD and i love it.

as for the high watts of heat out put when overclocking, I'm going to say that is not as much of a issue as some are making it out, yes these chips are furnaces but they are furnaces that preform, ya you will have to do something extra if you want to have a heavy overclock but I'm sure those of you who want the most of there CPU will make it happen
 
H100 is hardly good cooling for something that can and will consume 300W+ of power, like any Intel's HEDT since Haswell.

Not every HEDT user should have to get a $500 custom water solution. Even with a custom system, temps when o/c still get high as we have seen on other reviews due to the TIM.
 
Maybe this has already been mentioned but how the hell are we going to keep temps and power consumption from being huge issue when overclocking the higher core count cpus if the 10 core is this much of hot power hog? I cant even imagine the 18 core chip being pushed hard after seeing the results of the 10 core being oced.
 
ya it certainly would have been better if they solder the IHS, maybe they will have to for the 12-18 core parts, i think they probably should if not the entire line up, hell if Kabylake-X was solders that might give it a little validation to have basically a soldered 7700K but thats probly pushing it,

Well if anything the extreme temps from these CPU's are going to make some serious cooling solutions that will benefit everyone, or some extreme cooling solutions that only benefit the few...
 
Benches at 4.6 GHz with SKL-X is about as useful as benches with LN2. It is only good for pretty graphs. I doubt anyone will run that as their daily overclock.

I wish there were more overclocking data of the 7820k and 7800. They look to be able to run those clocks on a daily basis.
 
H100 is hardly good cooling for something that can and will consume 300W+ of power, like any Intel's HEDT since Haswell.

A h100iv2 is a 240MM Radiator AIO. I get you can do better, but it I find it very difficult to say its "hardly good cooling". I get 300W is a lot, but if the implication is that you need a custom waterloop with 360mm rads... well, im not sure I find that acceptable for ANY consumer part.
 
A h100iv2 is a 240MM Radiator AIO
And with how powerful of a pump, exactly? The reason custom water beats AIOs has less to do with size of radiator in general, but more with pump used.
but if the implication is that you need a custom waterloop with 360mm rads... well, im not sure I find that acceptable for ANY consumer part.
Enthusiast products are... for enthusiasts?
 
Back
Top