Yep, no-one was expecting the 2nd generation of Zen to be 7 nm. It'll be a refresh before they do a proper die shrink.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
*** Official Ryzen Owners Thread ***
- Thread starter Frozennova
- Start date
Caporegime
I don't see why it wouldn't be but it will only run in Dual Channel mode on X370 ^^^
Whatever Global Foundries process is its better than the last so its going to make an already better CPU even better. That's all that matters.
Who cares, i mean really? AMD's chips on 14nm or 16nm or whatever it is are far more efficient than Intel already so maybe Intel "real 14nm???" is a bit crap.
Whatever Global Foundries process is its better than the last so its going to make an already better CPU even better. That's all that matters.
That's more to do with architectural decisions rather than process.
But I'm not saying that jumping to a denser process is bad, it's just that the foundry marketing nowadays is really annoying, and the most egregious example is Global Foundries labeling a glorified 12nm as being 7nm.
Additionally I think people shouldn't have unreasonable expectations from the process jumps, I'm already seeing glee about >5Ghz Ryzens after than 12LP announcement on other forums.
As to which architecture is better and for what purpose, that's highly debatable.
But I'm not saying that jumping to a denser process is bad, it's just that the foundry marketing nowadays is really annoying, and the most egregious example is Global Foundries labeling a glorified 12nm as being 7nm.
Additionally I think people shouldn't have unreasonable expectations from the process jumps, I'm already seeing glee about >5Ghz Ryzens after than 12LP announcement on other forums.
As to which architecture is better and for what purpose, that's highly debatable.
Lots of process nodes have a 0.9x optical shrink, sounds like this is all it is, allows a quicker transition to another node and die cost savings, there will be performance from this sure, but wouldn't expect a great deal, also automotive tends to mean much more conservative.
You probably won't get anywhere near its rated speed, you'll be limited to sub 2400~2666Mhz most likely. If you want 32GB at >3000Mhz you better find a 2x 16GB kit with Samsung ICs, but even then getting to ~3000Mhz or higher might be hit or miss.
Soldato
2x16 is better on ryzen. Though as on every CPU, the more memory and modules you add the lower you can overclock.
Caporegime
That's probably because current Ryzen's are built on a Samsung 3Ghz process, its actually a mobile process, its why they don't overclock well, very power efficient (like the 3Ghz 1700 8 core sipping a tiny 60 Watts or so....) its not a good process for clocks.
You can get about 4Ghz or slightly more out of them but after that the volts needed to keep them stable only work under LN2
Ryzen 2 is built on Samsung's 5Ghz process. that's why....
Last edited:
Soldato
Most I have seen is 32gb@3200
Though that was with some gskill bdie. Results vary a lot on ram overclocks, some cpus have better memory controllers etc
Permabanned
I didn't do anything other than run it at it's stock XMP and then downclock it.
Im using Hynix and it didnt make any difference running 2 or 4 sticks overclocked.