AMD have released their blender render file for the new RYZEN CPU, *For reference, the new 8 core Ryzen and Intel CPUs render the scene in 35s (more or less, they didn't give the exact time). So i tried it on my slightly overclocked i7 and got the following results:
Considering that I kept my old Sandy Bridge for 5 years, I'll probably stick with the 5930K until 2020-2021 so there's no chance that I'll ever buy a Zen CPU (barring any catastrophic failures or natural disasters, of course).
Speaking of my old Sandy Bridge, the old i5-2500 (@ 3.8GHz currently) just did the same render in 2 minutes and 40.70 seconds, or 160.7 seconds. Seing as that PC is still very snappy in general use and the old i5 pushes my 1070 to deliver 60FPS@1080p on ultra settings in GTA5, I wonder how relevant this whole Blender benchmark is to the average user.
@PhaseLockedLoop this is interesting, wenn you have some time, could you maybe crush this on your 5960X? Would be interesting to see, what score you get at 3.4GHz. And your current clock speeds.
Of course AMD claimed that their Ryzen chip was locked at 3.4GHz. But yeah, that is of course something there is no direct proof for. Since they did not show a realy life monitor of the actual core speed during the tests.
Well yeah like i said, they did not show a real time clockspeed monitor during the tests. So they can claim it was running at 3.4GHz with turbo disabled. But there trully is no effective evidence for that.
Hey I remember my own data from my. Downloaded posts from ages ago.. It was 16. 1 on cinebench 11.5 at 3.5 ghz.. Can you extrapolate from that. I'm not home at the moment. Turbo was disabled.. My 4.46 ghz speed was 19.8..
I know ryzen is close to that which is a still a good jump but for all you die hard amd fans.. Guess what Intel probably has something up their sleeves to destroy amd haha
Thats pretty close taking 2 physical cores and 5MB of cache less in concideration. Not sure if you also have overclocked your cache speeds there? Because there might be some improvements to gain there aswell.
No, I merely set the multiplier to x45 and upped the voltage to 1.22V. I could do 4.6 stable at 1.36V, but then my CPU hits 95°C which is okay for the CPU itself but apparently 10°C beyond the socket's specification. I don't want to kill my motherboard, so I backed off as soon as it finished the first Cinebench test at that speed.
Never tried overclocking the cache or RAM, I just haven't done my research into overclocking. In fact I normally run this CPU at stock speeds because it's more than fast enough that way.
