AMD Responds to Ryzen's Lower Than Expected 1080p Performance

[efikkan]

Of course I'm not talking about single cache miss ... rather about max frame times ... with each cache miss at 62.5 ns, any excessive cache misses in one frame compared to previous would show as much bigger variation of the maximum frame time ... you say it adds up to a steady performance drop, but I say it should affect measured frame time variations in a non-steady manner.

You still don't understand the time scale here.
Fluctuations around 1-2 ms is very noticeable, and I would claim anything below ~0.2 ms is hard to notice.
For comparison 0.2 ms = 200 μs = 200,000 ns.

Not necessarily true... There are actually quite a few 8 -12 threaded games on the market.
When there is falloff in framerate on the intel side and the AMD side stays flat at higher res... that shows a cpu bottleneck plain and clear. If the gap does anything other than stay constant... the difference is more than the gpu.

Well there's hope that AMD's push on much more affordable multicore to the masses could result in developers making prettier and higher performance games even though it won't be easy just like AMD's push on lower level API with Mantle and more recently Vulkan/DX12

For both of you;
Multithreading in games mainly comes down to freeing up the rendering thread to work undisturbed building a queue. Granted, Direct3D 12 allows you to use multiple threads to build a single queue, but there's really not any point to it. Having several threads querying the driver this way will create a number a synchronization issues, so the gains will be minimal. So the gains of multiple threads will mostly be limited to having one thread per queue, and since most games use 1-2 queues for most of the load, there will not be a huge potential here. It's not like we can just throw four threads at it and scale nicely.

If a game has a problem with a bottlenecked CPU, it's usually caused by the computations done between each API call. So e.g. precalculating animations in a different thread can help a bit, but of course it mostly comes down to the code structure in the game engine. This is why I started by mentioning "freeing up the rendering thread".

They are also pushing 1000+ dev units out... they are giving away ryzen to game devs...

Too little, too late…
This is all about PR, sending out some dev kits is not going to make developers rewrite their games over night. In ~99% of cases reducing the bloat would require a major rewrite, which is not something that can be done in 10 hours or so.

 

[BiggieShady]

You still don't understand the time scale here.

Are you sure about that ... at 3% cache miss rate there are from 1 to 2 million cache misses inside each frame at 60 fps, you only need extra 1% more misses to happen next frame to have that net variance
... besides we already know the problem is cache latency when L3 gets over 8 MB