I hear what you are saaying. Skylake-X can get the mem latency down to ~52ns. Sill ~25% higher than CFL, but at that point my 7820x games REALLY well.
I hear what you are saaying. Skylake-X can get the mem latency down to ~52ns. Sill ~25% higher than CFL, but at that point my 7820x games REALLY well.Exactly why I went with an 8700k instead of Ryzen or SkylakeX despite wanting more cores. I'm not even pushing my uncore or memory hard (47uncore, 3600mhz 16-16-16-36) and my mem latency is ~ 41ns. Contrast that with what we see in Ryzen even when tuned to the hilt and you're looking at a 75% increase in latency which can really hurt in certain titles.
Yes and that transform into total higher latency that again impacts drawcall performance on 10 year old api where 20% drawcall difference drags the entire game down. And the entire user experience. Because drawcalls is the weak link.This has been mentioned in passing, but the #1 reason for Ryzen being poor in certain titles is the inter-CCX latency. Certain games really like low latency communication between cores and suffer when dealing with the higher latency introduced by the infinity fabric design.
Remember these benchmarks?
https://www.techspot.com/review/1450-core-i7-vs-ryzen-5-hexa-core/
Take a look especially at Grand Theft Auto V, Far Cry Primal, and Hitman. Those are three of the worst performing games on Ryzen.
Notice that in those same games, the i7-7800X is also much worse than the 7700K. In fact, the 7800X isn't much better than Ryzen 5 1600 when the 7800X is clocked to 4.7Ghz and the R5 1600 is a 4 Ghz. So R5 1600 is nearly as good as the 7800X despite Skylake's IPC advantage AND the 7800X's clock advantage.
So what gives? If it were really just a single-core performance issue, the 7800X should be better.
Well, the 7800X uses the higher-latency mesh interconnect, like all Skylake-X processors, while the 7700K is using the lower-latency ring bus design. The core-to-core latency is the obvious connection between the 7800X and R5 1600. So that has to be the reason.
We found a significant drop off in renderer performance, when the driver thread communicates with processes on another CCX.
How does the game perform when it's assigned to one full 4-core CCX?
Can i point something out?
https://www.youtube.com/watch?v=olI9Mmtw39Y
please check this, check BF1 benchmark. Then show me 1 other tech site which shows that stock R5 1600 with 2933Mhz DDR4 beats i5 8400 with same 2933MHz DDR4. Main thing here is that i5 will be slaughtered by R5 1600 in MP.
There is ONE thing that nobody is showing it. AMD does better in CS:GO.
You also found out that Ryzen does pretty well actually and basically there is so little difference when you draw the line. But everyone is showing ryzen as a non gaming CPU.
Some people talked about how will nvidia put drivers to improve ryzen performance on GTX.
Anyway, we finally can buy decent CPU under 200$.
Almost same 2+2 vs 4+0. CCX shouldn't be a problem. Consoles are same.
Main problem is ryzen is new arch and DF speed.
Why nobody tries to compare i7 8700k 1.6GHz vs R5 1600 1.6GHz with ddr4 3200.
1. If the gap would close we can assume ccx and ram latency.
2. If not, then it is optimization "mostly".
Not sure how to interpret that graph as all the configurations are using different GPUs?!
Can you please not derail this thread and spread FUD? What you posted is completely false, as The ELF said look at the speed the GPUs are running at! 1GHz on the 8400 vs 1.4GHz on the 1600. Those results are completely invalidated.
BF1 MP: https://www.computerbase.de/2017-10/intel-coffee-lake-8700k-8400-8350k-8100-test/5/
Ryzen is behind in avg fps, but actually does pretty well in min FPS, the 1800X is about equal to the 8400 and 1600X about equal to the 7700K, but the R5 1600 certainly doesn't outperform, let alone 'slaughter' the 8400, as you claim.
Now with that out of the way, thank you to all the other posters for keeping this thread clean and on topic!
It does seem like latency is the key here, as some of you mentioned generally in games where Ryzen struggles Skylake-X does too, most likely due to its higher latency mesh bus vs ring bus on the desktop CFL chips.
Does any one think gaming performance can be addressed much in the upcoming Ryzen refresh? Or are we realistically looking at Ryzen 2 before a significant uplift in gaming performance by AMD?
The CCX is indeed a problem. Just the draw call performance impact alone, is at least 20% when the driver thread is on a different CCX. From my draw call thread:
The highest 2x CCX and highest 1x CCX are using 3000MHz DDR4. And there is a 20% performance difference between them.
I really want to see how the performance changes, when the game is constrained to Ryzen 7's 2nd CCX.
That AC Origins benchmark looks all over the place, from what i hear the game is very poorly optimized.
That is old benchmark and this program/tester is now aware of Ryzen.
Since I do game (BF games mostly, some MMOs and new games sometimes), also I do own 144Hz and have few Intel processors. I am not an fanboy, I do prefer AMD against Intel, but I do prefer NVIDIA over AMD as GPU. Since I wanted freesync/gsync I was forced to buy AMD and it turned out to be great.
Anyway, I did notice some weird load and CPU, load jumping from thread to thread (CCX to CCX), after few updates that was gone. I can do "a test" for you (4+0 vs 2+2) in BF to show you guys if there is a difference or not? But, if result is not as you expected please do not blame me, just ignore.
Well, I tried bf1 emty map... 4+0 - 142 vs 2+2 - 138.Sometimes, all the threads of the process may be on one CCX. You'll have to run the program multiple times, to see if it is really constrained to a single CCX.
And yeah, real-world performance comparisons are fabulous. It's not been investigated at all, so if we have some quantifiable data on the CCX penalty in actual gameplay, we get some juicy new data.
Well, I tried bf1 emty map... 4+0 - 142 vs 2+2 - 138.
That AC Origins benchmark looks all over the place, from what i hear the game is very poorly optimized.