• 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 "Zen 4" thread (inc AM5/APU discussion) ***

It only happens when CPU boost is active (at least, when I tested it, it seemed fine for hours with CPU boost off).

I'm just gonna check if it could be related to a setting I turned off weeks ago, called PSS (on ASRock boards). This changes how CPPC works, which determines how cores spread out the load across the available cores. I turned this off before, because I thought it would be better to spread out the load more evenly across each core.
 
Last edited:
Has anyone else tried the PBO Enhanced Temp limits on the Asus mobos? I've benchmarked a 7600x with these settings..
80 Deg is faster than stock, 70 is in the same ballpark as stock. Am I missing something here.. These chips are incredible. Back in the day we were heating these things up, now we're trying to cool them down :D
 
bfar said:
Has anyone else tried the PBO Enhanced Temp limits on the Asus mobos? I've benchmarked a 7600x with these settings..
80 Deg is faster than stock, 70 is in the same ballpark as stock. Am I missing something here.. These chips are incredible. Back in the day we were heating these things up, now we're trying to cool them down :D
Click to expand...

Yea many have; this info is new to Zen4 but has been known since Zen4 launched and was covered by tech sites already. I've only looked into the 7950x myself but I imagine other models could experience similar behaviour, which is that enabling PBO and manually lowering the temp limit would result in extra performance and lower temps
 
Last edited:
Grim5 said:
Yea many have; this info is new to Zen4 but has been known since Zen4 launched and was covered by tech sites already. I've only looked into the 7950x myself but I imagine other models could experience similar behaviour, which is that enabling PBO and manually lowering the temp limit would result in extra performance and lower temps
Click to expand...

It's all well and good for AMD to claim that the 95 deg stock limit is fine, but that heat has to go somewhere - they're effectively offloading the problem onto the user, and it throws fan control up in the air as most motherboards measure fan speed against CPU temp.

When I completed the build and ran the fan optimisation in bios, the result was a fast hairdryer. The PBO enhanced temp config let me reach a very satisfactory outcome with an hour's worth of benchmarking and stability testing.
 
bfar said:
It's all well and good for AMD to claim that the 95 deg stock limit is fine, but that heat has to go somewhere - they're effectively offloading the problem onto the user, and it throws fan control up in the air as most motherboards measure fan speed against CPU temp.

When I completed the build and ran the fan optimisation in bios, the result was a fast hairdryer. The PBO enhanced temp config let me reach a very satisfactory outcome with an hour's worth of benchmarking and stability testing.
Click to expand...

This definitely put me off in addition to the platform costs. Improved my AM4 system for about 1/3rd of the price instead.
 
Can't wait for the 3D chips, the 7700X is perfectly fine but nothing quite like having the 'best' for your use case. That said, if there are 4 models I am not sure I'd go for the top one.
 
Has anyone else been getting intermittent failures in AIDA64 with a Zen 4 CPU?

Sometimes it reports as stable for hours, then I restart the PC and it fails in minutes.

I'm starting to suspect the memory controller or SOC.
 
Last edited:
bfar said:
95 deg stock limit is fine, but that heat has to go somewhere
Click to expand...
It is the watts that matter, not degrees.
Zen4 at 95C will still heat up your system less than Raptor Lake K models, whatever their temperature.
Besides temp limit setting in PBO section solves the "problem" with barely any drop in performance.
 
Although I have to admit, the weird shape thick IHS on AM5 was not a good choice. Means we will be stuck with sub-par cooling on AM5 for next 3 gens.
Hopefully we see a return of direct-die cooling to mainstream.
 
Last edited:
g67575 said:
Has anyone else been getting intermittent failures in AIDA64 with a Zen 4 CPU?

Sometimes it reports as stable for hours, then I restart the PC and it fails in minutes.

I'm starting to suspect the memory controller or SOC.
Click to expand...

I swear you had it stable weeks ago. You buy a PC just to run stability tests?
 
Noxia said:
I swear you had it stable weeks ago. You buy a PC just to run stability tests?
Click to expand...
I hate it actually. But the damn thing was crashing randomly in games. It should just work, but it's rarely that easy.

It seems to be stable if I increase the Curve optimiser setting to +30, which increases the voltage supplied to each core.

I'm testing to see if running the LLC at the highest setting for the Vcore and SOC will be enough to keep it stable, which would be a better option.

Unfortunately, I've had some crap advice, about not increasing the load line calibration, which for some reason, some people think is a risky or detrimental thing to do.
 
Last edited:
g67575 said:
I hate it actually. But the damn thing was crashing randomly in games. It should just work, but it's rarely that easy.

It seems to be stable if I increase the Curve optimiser setting to +30, which increases the voltage supplied to each core.

I'm testing to see if running the LLC at the highest setting for the Vcore and SOC will be enough to keep it stable, which would be a better option.

Unfortunately, I've had some crap advice, about not increasing the load line calibration, which for some reason, some people think is a risky or detrimental thing to do.
Click to expand...

Ah damn ok. Also, I know it's hard to get a tone for text but I realise my comment may have sounded a bit arsey which I didn't mean it to be!
Strange isn't it. I have stability using my PC but not booting up and you have the opposite. Though, as usual, no boot issues with everything at default and the problem is that performance doesn't really make a difference either, so I don't see the point in changing it.
Strangely though, it doesn't get hotter than about 80° at the moment despite everything being at default, whereas before it would go to 95. Maybe it's because I'm playing games though not using benchmark.

I have identified my GPU is faulty just by using my onboard one and not experiencing the same issues so I'm currently trying to RMA it but I don't know how accepting OCUK will be. I guess it could be my PSU but it seems less likely.
 
Last edited:
g67575 said:
I have boot/post issues too, if I don't set the DRAM voltages to 1.45v.
Click to expand...

Oddly, whenever I get them and I get the blue "windows must be repaired" it's always a different error code, I've googled and one was indicating SSD issue but I don't think this is the case, especially as this only happens when I use DOHC OR any non stock cpu settings. Like I said, makes sod all difference in games but theres that annoying feeling of it's not running as fast as it should be. I only have buyers remorse for the RAM tbh, didn't need 6000mhz.
 
Last edited:
Wow, that's a really high voltage you guys have being applied there, mine (a fully populated board 4 x 32GB) is still at defaults of 1.1v each (just checked). And no issues experienced thus far, boot or in Windows 11, or during gamging or benching. And all my settings are on Auto (for now, waiting on new BIOS before I start manually tuning stuff downwards).
 
You must log in or register to reply here.
Back
Top Bottom