AMD Ryzen 5000 Builders Thread

[B-Riz]

	Cores Threads	Base Freq	Turbo Freq	L3 Cache	TDP	MSRP
Ryzen 9 5950X	16c/32t	3400	4900	64 MB	105 W	$799
Ryzen 9 5900X	12c/24t	3700	4800	64 MB	105 W	$549
Ryzen 7 5800X	8c/16t	3800	4700	32 MB	105 W	$449
Ryzen 5 5600X	6c/12t	3700	4600	32 MB	65 W	$299

5600X comes with cooler, the rest do not.

above from https://www.anandtech.com/show/1614...en-3-on-nov-5th-19-ipc-claims-best-gaming-cpu

 

[lightmanek]

Benchmate info-snip with clocks
View attachment 56301

That makes sense then for these scores, but how?
Do you have a chilled water or just live in an Iglo?

 

[Det0x]

That makes sense then for these scores, but how?
Do you have a chilled water or just live in an Iglo?

Had opened the window to lower ambient temp for this bench session
Cooling is TechN AM4 block with liquid metal on custom watercooling = 60 degrees under load when running Cinebench r23 MT bench

 

[lightmanek]

Had opened the window to lower ambient temp for this bench session
Cooling is TechN AM4 block with liquid metal on custom watercooling = 60 degrees under load when running Cinebench r23 MT bench

My Corsair water block keeps it under 88C with water temp of 34C and pushing 4.7GHz all cores @1.35V (mining all the time on GPU heats up my room and water).

 

[Noid]

x570 chipset D/L

Everyone gets an update this time

AMD Ryzen™ Chipset Driver Release Notes 4.03.03.431
Article Number LRN-RYZEN-CHIPSET-4-03-03-431
Release Highlights

• Fixed PSP driver downgrade issues
• Fixed some text on Russian OS language pack

Known Issues

• Sometimes custom install fails to upgrade to latest drivers.
• Text alignment issues may be seen on Russian language.
• Manual system restart required on Non-English OS after the installation is complete.
• Windows® Installer pop-up message may appear during the installation.
• Uninstall summary log may incorrectly show uninstall status as fail on non-English OS.
• May observe a pop-up message "AMD Chipset Software is not responding" when the installer is launched and UI screen is clicked.

Chipset Support

	Windows 10	Windows 11
AMD WRX80	Yes	Yes
AMD TRX40	Yes	Yes
AMD X570	Yes	Yes
AMD B550	Yes	Yes
AMD A520	Yes	Yes
AMD X399	Yes	Yes
AMD X470	Yes	Yes
AMD B450	Yes	Yes
AMD X370	Yes	Yes
AMD B350	Yes	Yes
AMD A320	Yes	Yes

Processor Support

	Windows 10	Windows 11
AMD Ryzen™ Threadripper™ PRO Processors	Yes	Yes
AMD Ryzen™ 5000 Series Desktop Processors	Yes	Yes
2nd/3rd Gen AMD Ryzen™ Threadripper™ Processors	Yes	Yes
1st Gen AMD Ryzen™ Threadripper™ Processors	Yes	No
2nd/3rd Gen AMD Ryzen™ Desktop Processors	Yes	Yes
1st Gen AMD Ryzen™ Desktop Processors	Yes	No
AMD Ryzen™ 3000G/4000G/5000G Series Desktop Processors with Radeon™ Graphics	Yes	Yes
AMD Ryzen™ 2000G Series Desktop Processors with Radeon™ Graphics	Yes	No
AMD Ryzen™ Mobile 3000U/C, 4000U/H, 5000U/H/HS Series Processors with Radeon™ Graphics	Yes	Yes
AMD Ryzen™ Mobile 2000U Series Processors with Radeon™ Graphics	Yes	No
7th Gen AMD A-Series Processors	Yes	No

Package Contents
The AMD Ryzen™ Chipset Driver installation package contains various independent drivers designed to support the following Microsoft® Windows® platforms. Operating System support may vary depending on your specific AMD product

Driver Name	Windows 10	Windows 11	Change Details
AMD Ryzen Power Plan / AMD Processor Power Management Support	7.0.4.4	7.0.4.4	Updated processor power management settings for performance and power improvement
AMD PCI Device Driver	1.0.0.87	1.0.0.87	Stability improvement
AMD I2C Driver	1.2.0.118	1.2.0.118	No change
AMD UART Driver	1.2.0.113	1.2.0.113	No change
AMD GPIO2 Driver	2.2.0.130	2.2.0.130	No change
PT GPIO Driver	3.0.0.0	3.0.0.0	New program support added
AMD PSP Driver	5.18.0.0	5.18.0.0	New program support added
AMD IOV Driver	1.2.0.52	Not Applicable	No change
AMD SMBUS Driver	5.12.0.38	5.12.0.38	No change
AMD AS4 ACPI Driver	1.2.0.46	Not Applicable	No change
AMD SFH I2C Driver	1.0.0.86	1.0.0.86	No change
AMD USB Filter Driver	2.1.11.304	Not Applicable	No change
AMD SFH Driver	1.0.0.326	1.0.0.326	Fixed proximity sensor related issues Stability improvement on Ambient light sensor calibration
AMD CIR Driver	3.2.4.135	Not Applicable	No change
AMD MicroPEP Driver	1.0.33.0	1.0.33.0	Fixed ACPI related issues

 

[Ranulf]

300 series is getting win11 support?

 

[Stuka87]

300 series is getting win11 support?

AMD doesn't have control over MS giving official support. But that won't stop people from installing it anyway. So its good to see AMD tested those configs.

 

[Dave3000]

I noticed that the coolest core, core 0, on my 5800x reaches 77C in OCCT's Linepack 5-minute run and the hottest core, core 5, reaches 90C. The other cores temperatures are somewhere in between. All cores are around the same temperature during idle however, which is in the low 30's. I did reapply the thermal paste, this time with a bigger blob in the center instead of the 5-dot method, and tightened up the support bracket of my CPU cooler as there still was some play in the screws that hold the retention bracket onto the motherboard. Still my 5800x is getting a 13 degree difference between core 0 and core 5 during OCCT's Linepack 5-minute run and still the same temperatures. In Cinebench the difference is 10C between core 0 and core 5 with core 5 reaching 90C and core 0 reaching 80C and other cores somewhere in between in temperatures.

 

[Dave3000]

My Ryzen 5800x Cinebench 20 and 11.5 scores vary so much between runs or system boots. For example I sometime get a Cinebench 20 multi-core score as low as the upper 5700's (rare) and sometimes in the 5800's and sometimes as high as the upper 5900's to just above 6000 (rare), with the same CPU and case cooling, though usually it's in the mid-5900's. Is it normal for a score to vary this much in Cinebench 20, especially when it's from a fresh reboot of the system before a run? I'm using a Noctua U12A cooler for the CPU cooling with some Corsair thermal paste.

 

[DrMrLordX]

My Ryzen 5800x Cinebench 20 and 11.5 scores vary so much between runs or system boots. For example I sometime get a Cinebench 20 multi-core score as low as the upper 5700's (rare) and sometimes in the 5800's and sometimes as high as the upper 5900's to just above 6000 (rare), with the same CPU and case cooling, though usually it's in the mid-5900's. Is it normal for a score to vary this much in Cinebench 20, especially when it's from a fresh reboot of the system before a run? I'm using a Noctua U12A cooler for the CPU cooling with some Corsair thermal paste.

Are you running Win10 or Win11?

 

[Zenoth]

Hey guys, got my Ryzen 7 5800X just two days ago. Everything by default in the BIOS (ASUS TUF B550-Plus, latest BIOS update), ASUS Performance Enhancer is also Disabled (default). I don't overlock, it's just all defaulted.

But I noticed in most of the games I've tested so far that - actually quite often - the 'base' / lowest frequency can reach 3.6Ghz. Now, yes, it does boost itself up to anywhere between 4.6Ghz to even 4.8Ghz every now and then (usually the boost stays around 4.5Ghz though, even if it does boost higher sometimes). But... I feel concerned about the 'base' frequency. Isn't this CPU supposed to run at 3.8Ghz minimum? Or is that a "Sometimes runs to 3.8Ghz base but actually runs slower by about 200Mhz every now and then" sort of thing?

Is that normal behavior perhaps to keep the temps in check?

Although the temperatures seem to be within 'normals'. It idles between 54 to maybe 58 Celsius, can reach about 74 to 76 on load in most games. I've seen 84 to 85 Celsius in Cyberpunk 2077 though, although it doesn't go any higher than that. The 'average maximum' seems to be about 75 to 76 in most games (so far of the games I've tested, only Cyberpunk was the exception and increased the temps quite higher).

So anyway, just wanted to be sure. If it's all expected and yes the CPU can actually run at a lower-than-advertised-base frequency from time to time, then... alright I guess. Although it does boost itself up so I guess that part works well enough.

 

[Shmee]

Hey guys, got my Ryzen 7 5800X just two days ago. Everything by default in the BIOS (ASUS TUF B550-Plus, latest BIOS update), ASUS Performance Enhancer is also Disabled (default). I don't overlock, it's just all defaulted.

But I noticed in most of the games I've tested so far that - actually quite often - the 'base' / lowest frequency can reach 3.6Ghz. Now, yes, it does boost itself up to anywhere between 4.6Ghz to even 4.8Ghz every now and then (usually the boost stays around 4.5Ghz though, even if it does boost higher sometimes). But... I feel concerned about the 'base' frequency. Isn't this CPU supposed to run at 3.8Ghz minimum? Or is that a "Sometimes runs to 3.8Ghz base but actually runs slower by about 200Mhz every now and then" sort of thing?

Is that normal behavior perhaps to keep the temps in check?

Although the temperatures seem to be within 'normals'. It idles between 54 to maybe 58 Celsius, can reach about 74 to 76 on load in most games. I've seen 84 to 85 Celsius in Cyberpunk 2077 though, although it doesn't go any higher than that. The 'average maximum' seems to be about 75 to 76 in most games (so far of the games I've tested, only Cyberpunk was the exception and increased the temps quite higher).

So anyway, just wanted to be sure. If it's all expected and yes the CPU can actually run at a lower-than-advertised-base frequency from time to time, then... alright I guess. Although it does boost itself up so I guess that part works well enough.

What cooler are you using and what are you using to monitor temps? I think the cooling/temps can have an effect on boosts, but I would think the chip would stay at least at base clock, unless it is throttling or idle / not under sufficient load. Remember CPUs can downclock when idle to save power.

 

[igor_kavinski]

Although the temperatures seem to be within 'normals'. It idles between 54 to maybe 58 Celsius,

I don't know what's normal for 5800X but idling at above 50 degrees celsius doesn't seem normal. That's a LOT of leakage current if just idling spits out that much heat. What's your ambient temperature?

 

[Zenoth]

@Shmee and @igor_kavinski

I'm currently using my previous AMD 3700X Wraith cooler, although I cleaned it up before using it for my 5800X (removed the previous thermal paste too; and then applied brand new Arctic Silver 5 for the new setup). I'm aware, however, that a better cooler is recommended. But it's out of the question for the next 2 or 3 weeks, but I'll look into it at some point in a month or two from now.

Maybe the default fan speed for the Wraith cooler isn't good enough for a 5800X, but to be honest I actually had very similar temperatures with my 3700X anyway, really. But my 3700X never really had frequencies going below the 'base' minimum it should have had, which is what my 5800X does sometimes. According to AIDA64, my CPU Fans' RPM fluctuates between 1550 RPM to 1750 RPM on idle (would have been the same on my 3700X, I never touched that ever, not in BIOS and not from any programs in Windows to play around with Fan speeds).

For monitoring Temps I usually stick to either in-game OSD with MSI Afterburner (latest version), or in the Desktop when I do non-gaming benchmarks with AIDA64 (trial version). Now, an example is that - as of this typing - in the Desktop and currently idling (no background gaming going on) the reported temperatures are:

- With MSI Afterburner = 55º to 58º Celsius (fluctuating, doesn't seem constant).

- With AIDA64 = 47ºC to 52ºC (also fluctuating constantly) for the CPU, 56º to 60º for the CPU Diode and Package. Both the Motherboard and the Chipset's reported temps are exactly at 46ºC and constant, that one doesn't move (if it matters to mention).

And finally, my ambient room temperature is about 28ºC (just about to hit spring, but we do have VERY hot and extremely humid temps during the summer here sometimes; those days will be a real test).

So, MSI Afterburner does report higher CPU temps than AIDA64, there's that. Now, I did read around that the Ryzen 7 5800X is known to be a pretty 'spicy' CPU with temperatures anyway on a lot of air coolers (maybe less so with liquid cooling). Besides the Wraith cooler, maybe the voltage is the problem? As I mentioned, I don't overclock at all, and BIOS settings are defaulted. However, I've read reports that the 5800X can run hot and sometimes uses more voltage (maybe from BIOS settings) than it "should". Although, those articles and reviews are in most cases a couple of months old by now. Maybe those problems were from the early batches of the chip and it's been figured out better with BIOS updates from various manufacturers (I only assume that's the case).

 

[coercitiv]

Although the temperatures seem to be within 'normals'. It idles between 54 to maybe 58 Celsius, can reach about 74 to 76 on load in most games. I've seen 84 to 85 Celsius in Cyberpunk 2077 though, although it doesn't go any higher than that. The 'average maximum' seems to be about 75 to 76 in most games (so far of the games I've tested, only Cyberpunk was the exception and increased the temps quite higher).

So anyway, just wanted to be sure. If it's all expected and yes the CPU can actually run at a lower-than-advertised-base frequency from time to time, then... alright I guess. Although it does boost itself up so I guess that part works well enough.

AMD Zen CPUs operate with a target temperature as basis for allowing higher than base clocks. This means that aside from power limits, the CPU needs to stay under a certain temperature to boost up to max clocks, and there's also a temperature threshold after which the CPU may go bellow base clocks.

What's happening in your case is probably a combination of factors that lead to higher temps on the CPU than would normally allow for consistent boosting. These include:

• high ambient temps (and as you said, even higher temps to come in summer)
• possibly high case temps during gaming, as the 3080 dumps quite a bit of heat (but this is something you know better than anyone, as it relates to how much airflow your build has)
• an undersized cooler intended for 65W TDP operation that is now used on a 105W TDP CPU

My suggestion would be to move to a better cooling solution, there's plenty of cheap and decent tower coolers that will drastically improve this situation. The good news is there's no pressure on you to do this anytime soon other than the need for improved performance (and possibly bigger performance "loss" during summer heat), since the CPU power management is built in such a way as to limit power consumption when cooling isn't strong enuough.

You're essentially running the 5800X in "low power" mode. I did the same with my first gen Ryzen, the R5 1600X which I cooled using an old Scythe Ninja heatsink without a fan (semi-passive configuration as the case had some low speed fans). I also set the target temperature to 75C, which IIRC is lower than stock. The result was the 1600X boosted to max clocks as long as temps allowed, then progressively dropped those clocks to keep temps around the 75C target. The end result after continuous load was only 100-200Mhz above base clocks.

 

[Zenoth]

Yeah the cooler needs to be improved. I'll look into it at some point. I'm eyeing the Noctua NH-U9S at the moment, decent price. That'll wait a couple of weeks though. In the meantime I assume that having 75 to 80º on load in games isn't alarming per se. It's 'hot' for sure, but from official AMD documentation it seems the 5800X is designed to operate even up to 95ºC (but I wouldn't want to see that constantly, that much I know). The main issue is indeed probably just the cooler then, simply not being able to deal with the CPU for load temps (but is decent for idling or light Desktop use). I'll see how it goes when I get a new one.

Thanks for your time and help guys.

EDIT: Small update, I changed my Windows Power Options from 'Balanced (Recommended)' to Power Saver. It instantly drops the Idling temps to 45º to 46ºC, drops the CPU voltage down to about 0.945v (in AIDA64 anyway) and runs the CPU at an astonishing 1.75Ghz in the Desktop, lol. I mean I don't mind at all... it's just in the Desktop. It would go back up to normal operations in a game. But, yeah, doing this will help a lot for the next couple of weeks until I get a new cooler. And, still, I might just leave it on Power Saver since there's actually not much of a point having a CPU running circles at 4.0Ghz+ in the Desktop just browsing the web or listening to music. The Power Saver mode should probably be defaulted instead of Balanced, I'd argue.

 

[igor_kavinski]

I'm eyeing the Noctua NH-U9S at the moment, decent price.

https://www.amazon.com/Thermaltake-Universal-Ultimate-Over-Clocking-CLP0587/dp/B007UIX1RY

Normally $99 but on sale right now. May help you avoid hitting 90s in the summer months.

EDIT: Sorry, that's not compatible with AM4. Here's one that is: https://www.newegg.com/deepcool-gamer-storm-assassin-iii/p/13C-002K-000H4

 

[Noid]

uh .. del

 

[Noid]

Yeah the cooler needs to be improved.

Make sure your voltage does not go over " default ".
I limit my voltage to 1.25v. ( static OC of 4425 )
When I'm benching .. it's at 1.35v.
( BIOS manual over-ride )

Even though you might have everything at " Default "
" Default " is different for each MB manufacturer.

Lowering the voltage will lower your " boost ".

The " boost " mode default for my MB had voltages up to 1.45v depending on load.

I found that the numbers used in PBO had nothing to do with heat.

ALSO !

Check your " Load Line Calibration " ( LLC Mode = BIOS )
Usually in Power Options

Try " 3 " ( mode level 3 )

BOTH of these will limit your performance.
However, you will enjoy lower wattage, and heat.

 

[Noid]

MSI dropped a beta version of AMD ComboAm4v2PI 1.2.0.6c .

Some say that AMD put limits on the voltage now. ( hindering previous OC's )
I've also seen some improved benches though ...

Anyone tried it ?

 

[Det0x]

MSI dropped a beta version of AMD ComboAm4v2PI 1.2.0.6c .

Some say that AMD put limits on the voltage now. ( hindering previous OC's )
I've also seen some improved benches though ...

Anyone tried it ?

yes, i shared some data here

amd agesa 1.2.0.5 and above is limiting ST boost voltage to 1425mv when you set higher than default 140edc which can potentially limit boosting if you have good cooling.. If you have "normal cooling" you probably won't see any meaningful difference.

 

[Shmee]

I recently picked up a 5950X for a good price, and so far, what surprises me, is that it runs cooler on the cores than my 5900X did. It seems like some of the cores even boost higher as well, even though I have not even touched the curve optimizer yet.

 

[Shmee]

I scored 0 in CPU Profile

AMD Ryzen 9 5950X, NVIDIA GeForce RTX 3090 x 1, 32768 MB, 64-bit Windows 10}

www.3dmark.com

Here is the 5950X at stock with memory at XMP.

 

[Makaveli]

I recently picked up a 5950X for a good price, and so far, what surprises me, is that it runs cooler on the cores than my 5900X did. It seems like some of the cores even boost higher as well, even though I have not even touched the curve optimizer yet.

Not to surprised to be honest the 5950X has the highest quality silicon out of the Zen 3 line up.

 

[Shmee]

https://www.3dmark.com/compare/cpu/493356/cpu/492887 here is a comparison of my previous results at XMP, and the results with memory at 3733MHz. A fair bit of performance gain from the memory OC.

 

[moinmoin]

I recently picked up a 5950X for a good price, and so far, what surprises me, is that it runs cooler on the cores than my 5900X did. It seems like some of the cores even boost higher as well, even though I have not even touched the curve optimizer yet.

It makes perfect sense if you consider that for all cores to fit within the given TDP, per core the cores in 5950X should be binned to be more efficient than the ones in 5900X.