AMD Ryzen 5000 Builders Thread

Page 56 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

B-Riz

Golden Member
Feb 15, 2011
1,482
612
136
Last edited:

DrMrLordX

Lifer
Apr 27, 2000
21,582
10,785
136
Yeah that's a high tRFC setting. Typical for Intel memory controllers (at least of the Skylake-and-earlier generations). XMP 2.0 under any other name is still XMP 2.0. One wonders how long it will be before the DRAM OEMs finally start making alternate values for AMD CPUs? It's about time they got around to it.
 

bigboxes

Lifer
Apr 6, 2002
38,485
11,942
146
Yeah that's a high tRFC setting. Typical for Intel memory controllers (at least of the Skylake-and-earlier generations). XMP 2.0 under any other name is still XMP 2.0. One wonders how long it will be before the DRAM OEMs finally start making alternate values for AMD CPUs? It's about time they got around to it.

The timings are very tight. I suppose I could waste a lot of time trying to get that maximum performance out of the ram. I'm sure you could loosen the timings and get it to 3800 with good timings. Or you can set it and forget it and not worry about stability. I used to go for max o/c some years back. When it mattered. Now, it's so damn fast. I mean, I'm lightly o/cing the CPU to 4.1Ghz on all cores. It still boosts to 5050. What's not to love. The computer is dead quiet when not under load. The only thing that would make this better is a new video card. I'm still waiting on that. I'm not a big gamer so not that big of loss. I'm just going to be patient until prices come down on the new cards. No point in being in a hurry having a 1080 Ti.
 
  • Like
Reactions: Tlh97 and Leeea

DrMrLordX

Lifer
Apr 27, 2000
21,582
10,785
136
The timings are very tight.

Not really. tRFC north of 400 is pointless on just about every Ryzen memory controller ever made. It does nothing to contribute to stability to set that timing that high. The primary timings might be okay but the secondaries are probably all over the place.

I suppose I could waste a lot of time trying to get that maximum performance out of the ram.

You can, and it really doesn't take very long with the calculator, but the other point I was making is that DRAM OEMs need to start accommodating owners of the fastest enthusiast desktop CPUs on the market with DIMMs targeted at those machines. They're still selling RAM tuned for Coffeelake/Comet Lake. I think even Rocket Lake-S has problems with some of the higher-binned kits.
 

lixlax

Member
Nov 6, 2014
183
150
116
I just wanted to share that my 5800x's both are runaway hot with Gigabyte boards.

I can easily temp throttle on my noctua in the giant HAF case with lots of airflow when leaving the CPU with it's "stock" settings. It was really drawing a lot of power according to Ryzen Master.

My work PC with a AMD RGB cooler and PBO off and ecomode on at least doesn't crank the fans up so quickly after some aggressive fan curve tuning but temps can still change very dynamically.

Even with eco mode enabled on my bigger PC it is crazy beast, my 3600 was much easier to keep temps under control. It's actually really annoying because the 12cm fan will be really peaky doing normal things like logging me into windows.
Your boards are probably feeding the processor much more power than what is actually reported. I have the same "problem" with my Asrock board. I know it's nothing dangerous, but I just put on an all core overclock since I don't like seeing high temps (I also know its not optimal but likely I'll play around a little with the settings in the future) .

Are the board vendors really doing it to look couple of percent better in benchmarks/reviews?
 

cytg111

Lifer
Mar 17, 2008
23,047
12,715
136
Yea, went through that a dozen or so other threads on the same subject. Doesnt sound too good.

I have not swapped the CPU yet but I do seem to have gotten rid of it for the past two days, and I have been trying to replicate it.

What did I do? Open the case and put a big a* ventilator on it.. we are talking a one meter fan that should have been operating in a wind tunnel, this thing moves AIR.

So maybe something heat after all?
So swapped the CPU with the spare, did not connect the fan on the heatsink on purpose, closed it up and went to work. Works like a charm, the thing throttles like a champ and continues to trawl away.
This crappy case has a glass side so you can see through it (did t even know that until I unpacked the thing).. and that glass gets so hot you will burn yourself touching it. The 3080ti gets insanely hot! And the way the board is laid out means it sits 2cm’s away from the cpu heatsink.
I guess something is off with the other chips thermal/throttling mechanic.
I even went DOCP without problems.

I am gonna test the chip in the other setup before returning it, just to make sure its not some weird odd combo that sets it off.

Thanks everyone for the help!!!

Btw, auto cpu voltage in bios ranging from 1.35 to 1.48 normal? seems like a lot.
 
  • Like
Reactions: Tlh97 and Ajay

Dave3000

Golden Member
Jan 10, 2011
1,343
91
91
When I did a manual fan curve. I set it to spin both fans at around 1200-1250 rpm (54% duty cycle) for 20C to 60C CPU temperatures (not CPU Package temperature) and 100% duty cycle for CPU temperatures of 70C and up. If I set the fan profile to standard or turbo, the CPU fans will rev up and down frequently even during light usage or some background tasks happening and I find that annoying. On my 5800x, a CPU temperature of 70C is equivalent to around a CPU package temperature of 81C, so at around 81C CPU package temperature, my CPU fans are set to hit full blast. Also today I moved my stock rear case fan to the front of the case as a 2nd intake fan and use a spare 120mm Coolermaster Silent that I had lying around as a rear exhaust fan. CPU temperatures didn't change, but I think my 5800x boosts slightly higher now at full load, but it's still what I would consider within margin and not really a big deal. Would a 360mm AIO mounted in the front of the case as intake be worse for CPU cooling than a 240mm AIO mounted at the top of the case as exhaust. Also my GPU is runs hot enough as it is during heavy GPU load, and I don't want to bring hot air into my case and I heard the front mounting an AIO brings hot air into a case because of the radiator and it may be better for CPU temperatures than a top mounted AIO but worse for other components such as video card, is there any truth to that?

Last week I removed my 250GB 970 EVO Plus NVMe from the top M.2 slot and removed my 2TB 970 EVO Plus NVMe from the bottom M.2 slot and installed it in the top M.2 slot and in the process I had to remove my CPU heatsink to give more clearance for removing/installing the NVMe in that slot. This time I used the X-method application on the CPU for the thermal paste. No difference in temperatures or all-core boost speeds. One thing that does make a little difference is that when I have the air conditioner on, my CPU runs a bit cooler and boosts a bit better. If my room temperature is in the low 70's I get around 6040 in CB20 multicore and around 1-2C cooler CPU package temperature, but if my room is in the upper 70's, I get around 100 points less and hit 90C midway through the run. Also I had a hard time removing the heatsink from the CPU as it was literally glued on there. I did the twist method to try to break apart the bond before trying to pull it off rather than straight pulling it off the CPU and even then it was difficult as it took a few twists and still some force to pull the heatsink off. I would like to avoid repeatedly removing and installing my heatsink due to this, especially since the methods of application of the thermal paste I tried so far made practically no difference in CPU temperatures or boost speeds. So far I tried the pea size in the middle, a large blob in the middle (ended up making a mess using this method), 5-dot, and the X-method. I wonder if it's just my CPU cooler or if I just have a naturally hot running chip. If I can't get descent cooling to my 5800x using the Noctua U12A, I won't get descent cooling to it on a Corsair H100i CUe unless something is wrong with my Noctua U12A?
 

LikeLinus

Lifer
Jul 25, 2001
11,518
670
126
My 5950x has an Artic II 280 AIO and I've never seen it go beyond 58C. That is with it OC to 5.0 and a 30k (mult-core) Cinebench. Ryzen Master says I have a Golden CPU though, so that may help. I previously never tried to overclock it but decided to try a week or so ago, but turns out that my RAM needed to be in certain slots. Very odd, but I finally (using MemTest86) determined what sticks needed which slots. I'm running 64GB of ram in 4 x 16GB configuration. It was a pain, but I got it all working.

I was going to use a Noctua DH-15s, but I've since moved that to my 5800X. Highly recommend the Artic II 280 AIO though. I have Noctua 140mm fans on the other side of the radiator in a push/pull fashion with the stock fans. Keeps the CPU temps in the mid to low 50's at all times!!
 

Dave3000

Golden Member
Jan 10, 2011
1,343
91
91
My 5950x has an Artic II 280 AIO and I've never seen it go beyond 58C. That is with it OC to 5.0 and a 30k (mult-core) Cinebench. Ryzen Master says I have a Golden CPU though, so that may help. I previously never tried to overclock it but decided to try a week or so ago, but turns out that my RAM needed to be in certain slots. Very odd, but I finally (using MemTest86) determined what sticks needed which slots. I'm running 64GB of ram in 4 x 16GB configuration. It was a pain, but I got it all working.

I was going to use a Noctua DH-15s, but I've since moved that to my 5800X. Highly recommend the Artic II 280 AIO though. I have Noctua 140mm fans on the other side of the radiator in a push/pull fashion with the stock fans. Keeps the CPU temps in the mid to low 50's at all times!!

Is your RAM faster than 3200? If so that may be the true reason why you were having RAM issues and you may have to lower it to 3200 in the BIOS. Ryzen 5xxx supports up to 3200 officially and higher than that is considered an overclock to the memory controller of the CPU even though the RAM itself might be running within spec.
 

LikeLinus

Lifer
Jul 25, 2001
11,518
670
126
Is your RAM faster than 3200? If so that may be the true reason why you were having RAM issues and you may have to lower it to 3200 in the BIOS. Ryzen 5xxx supports up to 3200 officially and higher than that is considered an overclock to the memory controller of the CPU even though the RAM itself might be running within spec.

Yes, I have G.SKILL Trident Z Neo DDR4 DDR4 3600 (4x16GB) Samsung B-dies. I originally ran it without XMP and it was running at something like 2800 or 2600? I ran everything at stock and was getting BSOD when running games. I then started MemTest86 on ever single stick of memory in the first slot. Then I ran 2 slots, 3 slots and then 4, while populating it with certain sticks until I had full stability over 8 passes. I have not had a single BSOD since I populated it in a certain order. I also have it running at 3600 now with XMP. It's odd, but it totally worked. This was after I used other memory, another GPU and such. I was at the point that I was going to throw a 5800x and see if the CPU was the problem. My last try was going to be ordering a new motherboard. Thankfully, I got everything stable after the RAM fix.
 

lixlax

Member
Nov 6, 2014
183
150
116
Is that still a thing? Do boards with modern UEFI revisions and current AGESA versions still pull "phantom power" that isn't measured accurately by the on-board sensors?
At least mine does, according to HW info the value is ~85% and can go as low as 80%. When I was trying to figure it out I took couple of reviews as reference (CB20 multi score and system power consumtion). I kept lowering the PPT until I hit around the reference power consumtion and "accidentally" I was hitting the same score in CB as a stock 5900X should (the system total power draw was ~40W less compared to before).
The reason why I opted for the all core overclock was a situation when several cores got heavily loaded (BF5 for example)- it tried to clock those to the moon and beyond since it had still the coefficent of the power budget in hand even though the full CPU PPT was lowered (it was still plenty for several cores/threads).
I just couldn't think of a way to make the processor act normally under these conditions as well.
 

Dave3000

Golden Member
Jan 10, 2011
1,343
91
91
I noticed in one of the reviews online, that the 5700G has full bandwidth memory write bandwidth due to it's monolithic design. However, in the reviews I read about the 5700G compared to the 5800x, the 5700G still slower in everything else except the memory write performance. Is there be a situation in gaming where the 5700G would have the advantage due to twice the memory write bandwidth of the 5800x despite that all game benchmarks in the reviews I seen on it show that it's slower than the 5800x except for the memory write benchmark in Sandra? Also is a monolithic design suppose to be technically better for gaming than a chiplet design? Also does anyone here know how the 5700G compares to the 5800x in FS2020?
 

Makaveli

Diamond Member
Feb 8, 2002
4,715
1,049
136
I noticed in one of the reviews online, that the 5700G has full bandwidth memory write bandwidth due to it's monolithic design. However, in the reviews I read about the 5700G compared to the 5800x, the 5700G still slower in everything else except the memory write performance. Is there be a situation in gaming where the 5700G would have the advantage due to twice the memory write bandwidth of the 5800x despite that all game benchmarks in the reviews I seen on it show that it's slower than the 5800x except for the memory write benchmark in Sandra? Also is a monolithic design suppose to be technically better for gaming than a chiplet design? Also does anyone here know how the 5700G compares to the 5800x in FS2020?

I don't think memory write performance matters that much for gaming. The 16MB L3 vs 32MB L3 will have a much bigger difference in performance for gaming.
 

thigobr

Senior member
Sep 4, 2016
231
165
116
The CCD's memory write bandwidth has no impact in performance as it it still allows the core to performance at full speed... L3 on the other hand has a big impact on Zen3 designs. One would say the on die IMC could bring advantages but that doesn't seem to be the case here as well.
 

Noid

Platinum Member
Sep 20, 2000
2,373
183
106
AMD ComboAM4PIV2 1.2.0.3c
( MSI has released a BETA of this version )
I will be flashing later,,, when the sun is up ...

The AMD AGESA 1.2.0.3 Patch C BETA BIOS is an extension to the recently released Patch B, helping to create better compatibility with the new 5000G (Cezanne) APUs that are imminently due. Further to this, the patch will also enable X570 motherboards to have better compatibility with Ryzen 5000 CPUs (Vermeer B2 stepping) and should also resolve some of the USB issues users are experiencing with Ryzen 5000 systems.

The Ryzen 5000G APU lineup is set to launch in the coming weeks, with ASUS seemingly releasing this AGESA firmware patch just at the right time. The AMD Ryzen 7 5700G and 5600G will both be hitting DIY desktop shelves by the 5th of August. With the arrival of AGESA 1.2.0.3 patch, users should see an increase in reliability and yield issues for Cezanna APUs. That’s about all we know at this early stage, with more BIOS improvements to come.
We got some new information from MSI on the latest releases and the key features it the updated SMU. For Cezanne (Ryzen 5000G), the SMU has been updated to 64.50.0 while for Vermeer (Ryzen 5000), the SMU has been reverted to 56.52.0. For ASUS X570 ROG motherboards, the SMU version sticks to 56.53.0 for an unknown reason but we have asked ASUS & AMD to comment on this. Also, it is reported that the BIOS Firmware will be recognized as Patch B because the source code for the Patch C iBIOS is labeled as Patch B. To really make sure that you have updated to the right Patch, you will have to match the SMU version with Cezanne 64.50.0.
 

moinmoin

Diamond Member
Jun 1, 2017
4,933
7,619
136
Further to this, the patch will also enable X570 motherboards to have better compatibility with Ryzen 5000 CPUs (Vermeer B2 stepping)
That's odd. I thought the B2 stepping didn't bring any functional changes that would need tweaks of compatibility? Or is that disguised preparation for Zen 3D? But why only X570?
 
  • Like
Reactions: lightmanek

aigomorla

CPU, Cases&Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
20,839
3,174
126
Sigh...

So i hear the next gen threadripper will not support DDR5... or is this false?

Because my next upgrade will probably be a next gen threadripper with ddr5 + PCI-E 4.0 (although i would love 5.0), as we can safely say intel lost this one again, and i am not paying for a HEDT without at least 128 lanes and have overclocking. (Rules out Threadripper Pro)