https://www.techpowerup.com/reviews/ASRock/X299_OC_Formula/13.html
Is it me or am i missing a complete review of overclocking features? We see a chip idling at 1200MHz. Was that the final OC? Lol.
That is at idle, and for some users, seeing that the chip still scales down the multi and voltage when under OC is important. It could have been forced higher, but I feel it's important to show exactly what you see in that screenshot. All X299 boards are tested with the i9-7900X set to 4600 MHz on OC. If all the screenshots just sshow 4600 MHz, it's kinda of useless, and doesn't provide you with any real info, IMHO.
Here's an example of how many tests I keep results of (I actually run way more than what is shown here, and multiple times, but I only keep certain results for review purposes):
There's benchmarks there I kept screenshots of that aren't even posted in the review. With every board, results are highly dependent on what specific configuration you have. Memory, VGAs, CPU, power supply, drives, cooling are all changed many time over for every reviewed product. You cna see I do have OC results as well as memory OC results, evn though those aren't reported in the review. I don't just slap a rig together, run benchmarks, and then toss out a review; I build the rig, use it for a couple of weeks at least for 16-18 hours a day, and I tend to run benchmarks for reviews at the very end, but through the entire process I am constantly doing tests to get a true feeling of what a product has on offer. For example, if you check the "Sexy Hardware" thread, you'll see I have an ASRock X299E ITX AC board for review, along with a full cover block from Bitspower. It will be several weeks before you see that review posted, but I'm building that rig today.
Very few other sites go as in-depth on testing as I do. I don't do reviews for money or for access to the hardware; I do reviews to help ensure everyone is getting what they need out of these products, as well as making sure you guys have a place to ask questions and get some info that might not otherwise be available to you. This is my hobby, and I want it to grow so that I can continue to enjoy it for many many years to come.
I don't understand how one can look at X299 and conclude, as here, "This platform really does bring enthusiasts and high-performance computing and gaming together in a single platform" when this very review shows how much better Z270/Z370 is for gaming than X299.
X299 hasn't changed anything about how Intel's platforms stack up... its premium X[]99 platform is worse for gaming than its mainstream Z[]70 and gamer without some other use for 8+ cores is much better served by avoiding X299
You can't judge a platform like this by results with a single CPU. The KabyLake-X chips, by default, are the same as Z270, but have a higher spec right out of the box, for example. If you are NOT overclocking, X299 has higher clockspeeds right out of the gate. With past platforms, I would tend to agree with you, but X299 and it's socket that fits multiple CPU designs into it really changes the game this time, so I have written that to draw your attention those that.
I'm sure Dave can clarify but I'm going to assume he ran the benchmarks at default settings on the motherboard. the massive change Intel made on Skylake-X that hurts gaming is the Mesh and it really low default 2400 mhz clock speed, if one simply overclocks that to 3000 that will resolve most of the poor gaming performance seen on skylake-x vs ring bus desktop CPU's. end of the day though Sky/Kaby/Coffee Lake-S will always deliver you the highest of highest frame rates in many games esp when overclocked.
Yes, I always test and show "out-of-the-box" settings as this shows the board maker's tuning to Turbo profiles and other things like power settings (that idle multi, for example) that can affect overall performance and/or power use. It's actually really hard to create a level playing field in which to evaluate motherboard performance, to be honest, yet the field I put them on allows board makers to show you what they can do, tuning-wise, while also pointing out if that tuning is taking place, and I feel that that is very important for the end user in many cases.