Ok, I must try to be careful due the NDA, but you really read my statement right? What happened?
I tried three mainboards to mount a so-called AM4 compatible waterblock.
The goal was to build a powerful, real open loop water cooling to get for all our readers (and especially all the fanboys) the best possible results. And yes, it was mountable and at first glance it was looking ok. I screwed all to the end as every time (each screw has a kind of stopper to prevent the mainboard before bending).
But the performance in extreme situations was not optimal. This was not really mentionable in games, but in all my HPC tests to see (and show you!), what Ryzen can really do. I was thinking first, it depends at the high density, but this was completely wrong. I removed the waterblock and mentioned, that the film of the thermal grease was not so thin as usual. The spring had also a tick too much headroom and the pressure was as follow a tick too low. I can't write until launch any details about temps, sorry, but it was simply too high. Not only for my taste.
I tried mainboards from MSI and Asus, got pictures from a Gigabye - the backplates of all three bords were like at the photos in my news. That means, that all coolers, fixed with screws on the original AM4 backplate will lose min. 1 mm for mounting. This is not so much, right, but it can decide about the usability or not.
And yes, it looks like a minor issue, but ist is not so easy to detect and can cause unawaited problems
because it is not visible at first glance. I had longer phone calls with cooler companies and this guys, using screws, will add thick washers in their kits. For me it makes a big difference if I can save up to 8°C CPU temperature with my expensive cooling to use this CPU simply as I need it.
If you read all our former Ryzen information carefully: Ryzen uses XFR. Simply go back to this article and think about it, what a better cooler can do for you (or vise versa). Small cause, great effect. And no guys, I'm not a fanboy. But I was really confused, that dimensions for the holes and their distance were reported in white papers and not the (changed) heights of the holders. This are such unnecessary trifles that annoy me (I came from industry and also worked in the quality management). Nothing else.
I know the cooling headroom of such solution since years and can mention, if something is not going 100% right. But the normal end-user may think, the CPU is crap, or the cooling, or both. As you know, OC is like a sport. Everybody is trying it. And exactly this won't help AMD to make a good launch. That's all.