It should combine them. CO first, the CS on top.
Otherwise, the CS can't take account of the different ccd frequency limits.
I have suspicions about whether it actually works like that, but it's supposed to
-
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.
**AMD 9950X3D Owners Club**
- Thread starter The Asgard
- Start date
Otherwise, the CS can't take account of the different ccd frequency limits.
I have suspicions about whether it actually works like that, but it's supposed to
I think so. An offset on top of CO. So say you have CO -20 and CS set at 0 low, 10 med, 5 high, when conditions meet medium the -20 CO becomes -30.
Also for people that set high CO but freeze in light loads due to lack of power, you can set CS to positive for low loads, which boosts the voltage in light threaded applications to prevent freezing
- Joined
- 27 Jan 2003
- Posts
- 6,107
- Location
- Chesterfield, UK
That’s pretty clever. I assumed it was just an override. Cheers for explaining how it works.
It'd be really helpful if there was a max frequency option, preferably by CS range but I'd settle for an absolute max.
My all-core sits around 5300, it can go higher, temp and power are fine. Increasing fmax increases it, but also increases the single core max to unmanageable levels
I'd settle for disabling single core boost levels, i don't care about single core workouts at all
My all-core sits around 5300, it can go higher, temp and power are fine. Increasing fmax increases it, but also increases the single core max to unmanageable levels
I'd settle for disabling single core boost levels, i don't care about single core workouts at all
Last edited:
- Joined
- 27 Jan 2003
- Posts
- 6,107
- Location
- Chesterfield, UK
Perfect, thanks.
It'd be really helpful if there was a max frequency option, preferably by CS range but I'd settle for an absolute max.
My all-core sits around 5300, it can go higher, temp and power are fine. Increasing fmax increases it, but also increases the single core max to unmanageable levels
I'd settle for disabling single core boost levels, i don't care about single core workouts at all
There is a max frequency - it is hard fused to 5750Mhz (even says so in the bios). If you want to disable "low core workouts" you do this by setting curve shaper to 0 for the min and low frequency settings
Last edited:
Just ordered the 9950X3D, but now have to get the rest of the bits. The processor was what I was waiting for before pulling the trigger on the new build. Upgrading from a 5900X, mainly for VR gaming, but also for converting video. I didn't want to lose those extra cores.
I've had cores up at 5.9, I think with fmax but could have been eclk. But yes, fused limit per CCD.
Pushing Fmax up moves the curves and gives higher all core speeds(good), but means I get those 5.9Ghz speeds (bad) and then things error out.
I'll have a look at the CS options, but not sure how/why that would affect single core max boost clocks. I'll give it a shot though!
The AMD overclocking section has more options to tinker with. It is designed by AMD and is for people who know what they're doing. The Extreme Tweaker section is designed by the motherboard manufacturer and is supposed to be a bit more user friendly with only essential options. Also I believe the Extreme Tweaker section feeds data to Ryzen Master, but I don't think the AMD section does (I could be wrong on this). If there's a conflict, Extreme Tweaker takes precedence.
Now this explains why I couldn’t see anything taking effect in Ryzen Master when using it today. I thought my settings weren’t working for some reason when trying the AMD overclocking section.
Thanks for the info
What max clocks are people getting? Although my CCD0 can boost to 5750Mhz on most cores, my CCD1 only gets to 5925Mhz on 2 cores and I've been trying to max it for days.
I have noticed in my Asus bios the AI Tweakers section (which is greyed out so it's for information only) indicates that the AI is only boosting the CCD1 to 5925. Any thoughts on how I might get it to 5950Mhz?
Obvs the missing 25Mhz is neither here nor there, but I'm OCD and won't rest until I see it running at the max clock it should
I have noticed in my Asus bios the AI Tweakers section (which is greyed out so it's for information only) indicates that the AI is only boosting the CCD1 to 5925. Any thoughts on how I might get it to 5950Mhz?
Obvs the missing 25Mhz is neither here nor there, but I'm OCD and won't rest until I see it running at the max clock it should
Mines exactly the same 5,925MHZ.
Fitted a new Kraken Elite 360 v2 AIO today, was previously running a Kracken Z73 360, but that was 4 years old now and I'm convinced the pump was on the way out, as it made a couple of odd noises recently, then on Sunday, I came in to my office after a few hours away to see that the PC had shut down and upon powering it back up, the display had a reading of 105c so obviously pump had stopped and it shut itself down.
Temps seem a tiny bit better, the contact area on the cold plate is better than the old cooler, which had a circular plate as opposed to a square one with the V2. Idle is still 40-42, gaming around 65 average, and cinebench after 30 minutes was 76. Only 2-3 degrees difference across all 3 scenarios but at least I don't have to worry about the pump on the old one. Fan curve on all 9 fans is the same, fixed at 950RPM to keep it quiet
Also have gotten mine down to -25 CO, I could try for -30 but find the long stability tests each time I adjust it a massive PITA
Temps seem a tiny bit better, the contact area on the cold plate is better than the old cooler, which had a circular plate as opposed to a square one with the V2. Idle is still 40-42, gaming around 65 average, and cinebench after 30 minutes was 76. Only 2-3 degrees difference across all 3 scenarios but at least I don't have to worry about the pump on the old one. Fan curve on all 9 fans is the same, fixed at 950RPM to keep it quiet
Also have gotten mine down to -25 CO, I could try for -30 but find the long stability tests each time I adjust it a massive PITA
Last edited:
I’m currently doing a per core CO which is taking a while due to previous test methods I used last gen not being so good this time around. Looks like AIDA 64 CPU, FPU and cache is finding instability better than per core testing.
I’m also using the exact same AIO and my idle temps are exactly the same as yours around 42 degrees and it’s quite warm here for a change.
I’m also using the exact same AIO and my idle temps are exactly the same as yours around 42 degrees and it’s quite warm here for a change.
What application and sensor are you using to monitor temperature?Fitted a new Kraken Elite 360 v2 AIO today, was previously running a Kracken Z73 360, but that was 4 years old now and I'm convinced the pump was on the way out, as it made a couple of odd noises recently, then on Sunday, I came in to my office after a few hours away to see that the PC had shut down and upon powering it back up, the display had a reading of 105c so obviously pump had stopped and it shut itself down.
Temps seem a tiny bit better, the contact area on the cold plate is better than the old cooler, which had a circular plate as opposed to a square one with the V2. Idle is still 40-42, gaming around 65 average, and cinebench after 30 minutes was 76. Only 2-3 degrees difference across all 3 scenarios but at least I don't have to worry about the pump on the old one. Fan curve on all 9 fans is the same, fixed at 950RPM to keep it quiet
Also have gotten mine down to -25 CO, I could try for -30 but find the long stability tests each time I adjust it a massive PITA
HWInfo for the package temp, which is I’m guessing what CAM reads as they seem to match up