Caporegime
mine only runs at 3.6-3.7ghz in cinebench r23, what gives? Score is 24k
That's well under, you should be scoring about 27.5K, maybe something in your BIOS? have you tried resetting it?
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.
mine only runs at 3.6-3.7ghz in cinebench r23, what gives? Score is 24k
How do you get 4.2ghz? When I run CBR23 its 3.6-3.7ghz and pulling 120-130w. Only when I enable PBO does it go up to 4.45ghz
Yeah, I couldn't find anything. I was thinking more along lines of an "at least" max all-core clock seeing as the 5950X is listed as having a base clock of 3.4 and a max boost clock of up to 4.9. It means above the base clock it's incredibly vague.AMD don't publish fixed boost clocks because there are none, they are not fixed clock CPU's.
Looking at the R23 thread there are scores in the 24 to 25k range so it's not incredibly low.mine only runs at 3.6-3.7ghz in cinebench r23, what gives? Score is 24k
mine only runs at 3.6-3.7ghz in cinebench r23, what gives? Score is 24k
CB23 is super temp sensitive, what are your temps? PBO is also needed to push the higher scores, along with a good curve in the curve optimiser
That's with PBO at 270/168/220 and it's not even the highest out there. 32k has been broken on a 5950x. 33k might even have as well with extreme cooling.
But yeah, you could literally open your window in winter, drop your ambient a few degrees and then see +500 in CB23 lol.
80c in r23 when stock (HWinfo shows 130-140watts)
82c in r23 with pbo (HWinfo shows 230-260watts)
Air-cooling? Not too bad temp wise.
Have you got a curve setup with curve optimiser? A lot of performance can come from it. It results in the cores boosting higher.
You should definitely be able to squeeze over 30k with PBO, a curve and those temps with a 5950x.
At least 29xxx if anything else.
Custom water loop but it's middle of summer with 28-31c temps. CPU temps will be 10c lower in Winter. Nah it's just auto or stock on everything. Tried curve optimizer a couple times but doing any negative offset results in crashes
AMD Ryzen 7000 Zen4 “Raphael” desktop 170W SKU rumored to feature 16 cores
The latest rumors suggest that the top 16-core configuration might require up to 170W of power.
https://videocardz.com/newz/amd-ryzen-7000-zen4-raphael-desktop-170w-sku-rumored-to-feature-16-cores
Are AMD developing this CPU in partnership with another company? Nvidia?
With superior node on that wattage performance should be extremely good. That wattage alone doesn't tell us anything about efficiency, we have yet to see performance and how it scales with power consumption.You know how with the 5950x it says 105w but actually in certain loads it will reach 140w, well for Zen 4 it's 170w TDP and under certain loads it goes up to 230w - 230watts for a stock 6950x!
AMD is quickly catching up to Intel's Wattage
AMD Ryzen 7000 Zen4 “Raphael” desktop 170W SKU rumored to feature 16 cores
The latest rumors suggest that the top 16-core configuration might require up to 170W of power.
https://videocardz.com/newz/amd-ryzen-7000-zen4-raphael-desktop-170w-sku-rumored-to-feature-16-cores
Is AMD developing this CPU in partnership with another company? Nvidia?
Ah, it'll be not using curve optimizer then, a lot of performance comes from that.
If you ever feel like trying again this will help https://github.com/sp00n/corecycler/releases
It's a slow process but worth it in the end. The above tests the cores on a cycle and will report any that crash.
Unfortunately all core undervolts at big numbers never work. People who say things like -30 all core stable are talking nonsense. Per core is the way to do it, unless you want to start with a -10 all core and see what cores crash.
My lowest core is -1 and my highest -30. Your 4 best cores will likely be somewhere between -1 and -20. The rest probably -20 to -30, a good few possibly hitting -30.
Maybe. But if the sweet point of 5nm is, say, 1.00V and 4.5GHz and they want to run it at 1.25V and get 4.7GHz then that is not a good thing.With superior node on that wattage performance should be extremely good. That wattage alone doesn't tell us anything about efficiency, we have yet to see performance and how it scales with power consumption.
Maybe. But if the sweet point of 5nm is, say, 1.00V and 4.5GHz and they want to run it at 1.25V and get 4.7GHz then that is not a good thing.
Or even worse, say at 1.00V 5nm gets 4.9Ghz and uses 65W, while at 1.25V it gets 5.0GHz but uses 170W. Crazy stuff like that is possible.