Ryzen Owners Zen Garden

[AusWolf]

Nope. Wondering about any actual difference between XMP and EXPO. TPU memory reviews seem to point to no difference.

They're basically two approaches in your RAM telling your motherboard what speed works with what latencies. I'd be very surprised if there was any difference between them assuming that the two kits are otherwise the same.

 

[sethmatrix7]

They're basically two approaches in your RAM telling your motherboard what speed works with what latencies. I'd be very surprised if there was any difference between them assuming that the two kits are otherwise the same.

While I also understand that to be the case- a direct comparison would tell us for sure.

I'll even test it myself if somebody wants to send an XMP/EXPO capable set

 

[Klemc]

While I also understand that to be the case- a direct comparison would tell us for sure.

I'll even test it myself if somebody wants to send an XMP/EXPO capable set

You'll need three :

XMP
EXPO
XMP/EXPO

 

[AsRock]

Here's some thing odd, while playing BeamNG i noticed a temp jump to too 101c, so thought maybe for what ever reason after weeks of not noticing any thing like this maybe a bad contact so thought i would see if i could do it again same game same challenges same cars and tried to see if it would say it again and it just will not do it.

Question is the CPU trying to kill it self or the motherboard or maybe a software glitch. As you can see from HWinfo the right side the temps are no were near what it supposed to of done.

Damn do I wish that G.Skill would mention what kind of memory dies they use on their modules.

Not even the packaging says, i have some here been a none issue.

 

[AusWolf]

While I also understand that to be the case- a direct comparison would tell us for sure.

I'll even test it myself if somebody wants to send an XMP/EXPO capable set

I'll offer the same if someone sends me a 6000 MHz XMP kit.

 

[Klemc]

How can i switch CPU-Turbo from windows ?.. possible ?..

Even if i need logoff/on but not reboot.

B650E-E
7800X3D
DOCP OFF

 

[AusWolf]

How can i switch CPU-Turbo from windows ?.. possible ?..

Even if i need logoff/on but not reboot.

B650E-E
7800X3D
DOCP OFF

Maybe in Ryzen Master, but I'm not sure. Relying on Windows programs for CPU-related stuff is not really good, imo.

 

[Klemc]

Can't try it in VM, so can't be sure option is there !

 

[Count von Schwalbe]

You can set maximum processor speed under advanced power options.

 

[Nordic]

What speed does a 7800x3d boost upto under full load at stock settings?
What Curve Optimizer values are common for the 7800x3d?

 

[Klemc]

AMD AGESA 1.0.0.7B BIOS Delivers Improved DDR5 Memory Support of Up to 8,200 MT/s on AM5 Motherboards

AMD has begun rolling out AGESA 1.0.0.7B, and according to CPU-Z screenshots shared by early testers and manufacturers, the new BIOS is set to deliver improved DDR5 memory support to AM5 motherboard users. MSI's MEG X670 ACE motherboard can now support up to DDR5-8200 thanks to the new firmware...

www.thefpsreview.com

 

[tabascosauz]

Damn do I wish that G.Skill would mention what kind of memory dies they use on their modules.

Here's some thing odd, while playing BeamNG i noticed a temp jump to too 101c, so thought maybe for what ever reason after weeks of not noticing any thing like this maybe a bad contact so thought i would see if i could do it again same game same challenges same cars and tried to see if it would say it again and it just will not do it.

Question is the CPU trying to kill it self or the motherboard or maybe a software glitch. As you can see from HWinfo the right side the temps are no were near what it supposed to of done.


View attachment 305610

Not even the packaging says, i have some here been a none issue.

Y'all know that G.skill is by far the most transparent vendor.....? Lol

Each G.skill packaging has a window. The window allows you to see the sticker. The sticker shows you the 042 code. The 042 code always shows you what ICs the kit has, no exceptions. S820A for good DDR5.

As to the temp spike seems like it's just a bug. AM5 AGESA still seems a lil buggy. Seems to be related to the 1.6 (even 2.0V iirc) VSOC spikes that some folks have recorded in HWInfo but clearly aren't true because that sort of VSOC would cause instant death even with no current draw.

Happens on AM4 too. I still have those screenshots lying around over the years. Vega 8 running at terahertz core clock, 3700X at -10,048C degrees, 5900X at 8GHz effective clock, etc.

 

[AsRock]

Y'all know that G.skill is by far the most transparent vendor.....? Lol

Each G.skill packaging has a window. The window allows you to see the sticker. The sticker shows you the 042 code. The 042 code always shows you what ICs the kit has, no exceptions. S820A for good DDR5.

As to the temp spike seems like it's just a bug. AM5 AGESA still seems a lil buggy. Seems to be related to the 1.6 (even 2.0V iirc) VSOC spikes that some folks have recorded in HWInfo but clearly aren't true because that sort of VSOC would cause instant death even with no current draw.

Happens on AM4 too. I still have those screenshots lying around over the years. Vega 8 running at terahertz core clock, 3700X at -10,048C degrees, 5900X at 8GHz effective clock, etc.

Yes they do but there in my system, is a shame that the packing don't too.

 

[tpa-pr]

Y'all know that G.skill is by far the most transparent vendor.....? Lol

Each G.skill packaging has a window. The window allows you to see the sticker. The sticker shows you the 042 code. The 042 code always shows you what ICs the kit has, no exceptions. S820A for good DDR5.

As to the temp spike seems like it's just a bug. AM5 AGESA still seems a lil buggy. Seems to be related to the 1.6 (even 2.0V iirc) VSOC spikes that some folks have recorded in HWInfo but clearly aren't true because that sort of VSOC would cause instant death even with no current draw.

Happens on AM4 too. I still have those screenshots lying around over the years. Vega 8 running at terahertz core clock, 3700X at -10,048C degrees, 5900X at 8GHz effective clock, etc.

Yup, I'm seeing the same thing from time to time in HWiNFO myself. Occasionally my stock VDD and VSOC will just double. Obviously it's a software bug since I'm not getting any smoke or fire. I've heard mention that a beta version fixes the misreads but I haven't changed to it myself yet.

 

[AusWolf]

What speed does a 7800x3d boost upto under full load at stock settings?

5.05 GHz single core, 4.8-4.85 GHz multi-core, or 89 °C, whichever is lower with your cooling. (Measured in Cinebench R23)

 

[Klemc]

Arctic Freezer 280 works well on 7800X3D

 

[trparky]

Yes they do but there in my system, is a shame that the packing don't too.

It's also a shame that HWInfo doesn't show it too.

 

[Nordic]

Arctic Freezer 280 works well on 7800X3D

I have custom water cooling with the AMD specific optimus block. This loop was able to keep my 5950x at 200w very cool. I am having a yard time gaging what is good for a 7800x3d. The chip doesn't run hotter than 77c under full load at stock settings or with -30 CO on all cores. Stock settings will boost upto 4800mhz for all cores and -30 CO will boost upto 4900mhz for all cores. Is that good? It is hard to say. My loop doesn't even get warm.

 

[tabascosauz]

I have custom water cooling with the AMD specific optimus block. This loop was able to keep my 5950x at 200w very cool. I am having a yard time gaging what is good for a 7800x3d. The chip doesn't run hotter than 77c under full load at stock settings or with -30 CO on all cores. Stock settings will boost upto 4800mhz for all cores and -30 CO will boost upto 4900mhz for all cores. Is that good? It is hard to say. My loop doesn't even get warm.

You might just have to repaste well and tighten it down real hard like I did on my Optimus. Skin on my fingers bout ready to come off back then. One of the reasons I haven't busted out the loop again in like a year lol

Still, there's only so much you can do. It is an AMD optimized block and the die orientation hasnt changed from AM4 to AM5, but heat density is much higher compared to 5950X. On my 5900X it was evident that while all-core cooling capability significantly increased, ST temps were not much different.

On 5800X3D it was noted there was an outrageously large sample-to-sample variation in avg temps. Seeing as AMD is still AMD and the IHS is even thicker now, might still be the case.

 

[Nordic]

You might just have to repaste well and tighten it down real hard like I did on my Optimus. Skin on my fingers bout ready to come off back then. One of the reasons I haven't busted out the loop again in like a year lol

Still, there's only so much you can do. It is an AMD optimized block and the die orientation hasnt changed from AM4 to AM5, but heat density is much higher compared to 5950X. On my 5900X it was evident that while all-core cooling capability significantly increased, ST temps were not much different.

On 5800X3D it was noted there was an outrageously large sample-to-sample variation in avg temps. Seeing as AMD is still AMD and the IHS is even thicker now, might still be the case.

It is really hard to say. I cannot tighten it anymore. I usually do a quality job of applying paste but sometimes mistakes are made. I expected the thicker IHS to make cooling harder. I don't know what is normal for these chips. Is a 4900mhz all core boost good for water cooling? W1zards review showed a 4800mhz boost at 89c with air cooling. I have better results than that which makes sense with custom water cooling.

I got one of those offset kits to move the waterblock down a bit. My motherboards top most M2 slot gets in the way.

I might end up delidding this chip. I'm going to wait a few months and see how I feel then.

 

[AusWolf]

It is really hard to say. I cannot tighten it anymore. I usually do a quality job of applying paste but sometimes mistakes are made. I expected the thicker IHS to make cooling harder. I don't know what is normal for these chips. Is a 4900mhz all core boost good for water cooling? W1zards review showed a 4800mhz boost at 89c with air cooling. I have better results than that which makes sense with custom water cooling.

I have better results than that with the Dark Rock 4 - 84-ish °C at 4800 MHz. But the Silent Loop 2 280 mm is absolute garbage with this chip. 4400-ish MHz at 89 °C. It seems to me that the 7800X3D is just generally picky with the cooler for some unknown reason. Maybe the IHS isn't as thick as on other AM5 CPUs because of the stacked cache, leading to hotspots? Dunno...

I got one of those offset kits to move the waterblock down a bit. My motherboards top most M2 slot gets in the way.

Same here, although I managed to do a 45° offset instead of straight down. It didn't improve my thermals, though, so I'm not using it. Unnecessary hassle.

 

[tabascosauz]

It is really hard to say. I cannot tighten it anymore. I usually do a quality job of applying paste but sometimes mistakes are made. I expected the thicker IHS to make cooling harder. I don't know what is normal for these chips. Is a 4900mhz all core boost good for water cooling? W1zards review showed a 4800mhz boost at 89c with air cooling. I have better results than that.

I got one of those offset kits to move the waterblock down a bit. My motherboards top most M2 slot gets in the way.

I might end up delidding this chip. I'm going to wait a few months and see how I feel then.

I don't know if the offset is necessary. Optimus is optimized for Ryzen layout. Maybe I'm wrong and it still makes a difference though.

As long as inlet/outlet aren't mixed up, sounds great already.

 

[Nordic]

My process for overclocking this 7800x is to decrease all core Curve Optimizer in increments of 5 until Core Cycler finds an unstable core. At that point, stable cores will be decreased by 5 again and unstable cores will be increased by 4. Cores that failed once will then go up in increments of 1. Rinse repeat until I have found optimal per core values.

For my 5950x Prime95 found errors at almost any setting when the system was otherwise stable. I started using Y-cruncher instead because it did find errors but was a lot more lenient. I never had stability issues using this method.

The 7800x3d has passed 60 minutes per core with Y-cruncher for -25, -30, and -35. I now have it running at -40. I am not sure if this is going well or if Y-cruncher isn't identifying errors that it should. I may need to double check with Prime95 or another test.

 

[freeagent]

Even OCCT has become very good at finding problems.

Too good sometimes.. grr..

 

[Nordic]

Even OCCT has become very good at finding problems.

Too good sometimes.. grr..

At -40 CO I am getting ~5010mhz on all cores with a CB23 MT of 19076.

What settings are you using for OCCT?
-40 passed the default 30 minutes with these settings.

https://imgur.com/8bzMGHj

EDIT: I got all the way upto -50 CO. OCCT's stability test, CoreCycler with Prime95, and CoreCycler with Y-Cruncher were not finding ANY errors. Some light gaming showed had no issues. I was beginning to think I might have some golden sample or something. Then it happened. The computer restarted during OCCT's Multiple threads - AVX benchmark. Testing continues!