5600X Overclocking advice

SpongeBob

The Contraceptive Under the Sea
Joined
Jan 15, 2011
Messages
938
Hey folks, curious where everyone ended up with overclocking their 5600X these days? Are you OCing in your BIOS or in AMD Master software? I've been a little out of the loop and a buddy just gifted me a 5600X. What should I expect out of overclocking this chip or does AMD have their stuff together and I should just let it boost on it's own?
 
I used a 5600x when I built my son's PC and decided to leave it at stock. I didn't have much luck overclocking it and according to most of the reviews I've seen, the benefits are kinda negligible over leaving it at stock and letting its "turbo mode" do its job.

I tinkered with it in the BIOS as I've always believed that's the best way to overvlock. I've never been a fan of software based OC'ing.
 
I also tuned my ram to 4000 14-15-15 @1.6v to optimize it's FCLK.
hmm maybe im not takin enough voltage off. mine still peaks at 1.35ish.
i think thats some of the lowest timings ive seen at 4000, nice.
 
hmm maybe im not takin enough voltage off. mine still peaks at 1.35ish.
i think thats some of the lowest timings ive seen at 4000, nice.
What is your fclk limit? I think going 3800mhz 16-16-16 1.4v on ryzen 5000 series can help improve 1% low fps using high end gpu.
When I had 5600x, I only have gtx 1050 LMAO.
 
Well, get a board with 4 dimm's and run 4 x 8Gb or 4 x 16Gb setup, Ryzen 5000 works different is what Steve is saying, add that Thermalright Peerless Assassin 120 SE cooler for $36 if it will fit, everyone keeps asking about the 5800X 3D and how to make it faster!

 
Well, get a board with 4 dimm's and run 4 x 8Gb or 4 x 16Gb setup,...
Or a 2x16 setup with dual ranked DIMM's. I'm 99.9% sure that they were demonstrating the benefits of dual ranking, not some intrinsic benefit of 4 sticks. 4x8 is guaranteed dual rank, while these days 2x16 is not; and Zen 3 is the first Ryzen with a controller that can reliably run 4 sticks well. (I gather that a good number of Zen 2's can as well, but it's not a given.)
 
Or a 2x16 setup with dual ranked DIMM's. I'm 99.9% sure that they were demonstrating the benefits of dual ranking, not some intrinsic benefit of 4 sticks. 4x8 is guaranteed dual rank, while these days 2x16 is not; and Zen 3 is the first Ryzen with a controller that can reliably run 4 sticks well. (I gather that a good number of Zen 2's can as well, but it's not a given.)
Yeah I am going to be going for the 2 x Dual rank I believe. I am looking at ram options now and I'm having a hard time trying to figure out which to go with. DDR4 cl16 3200 or cl18 3600 (unlikely cl16 3600). I know some people are like 3600 is the sweet spot for Ryzen. Then others are saying go for the lower timings. Then I think well which is going to actually contribute to increasing my 1% lows in games? Or do I just grab 2 x 16GB of whatever is cheapest on my QVL list.
What do you guys think?
Basing it on this on the benchmark info below it looks like the difference between the cl16 3200 cl16 3600 (usually out of my price range where I live) and cl 18 3600. The difference is
only about 4.32-5.1% from what I can see.

Anyone use both with their 5000 series and did you even see a difference?

https://www.guru3d.com/articles_pages/amd_ryzen_ram_scaling_effect_in_games,9.html
 
Keep in mind that 3200CL16 and 3600CL18 are exactly the same CAS latency; you can't compare timing numbers in a vacuum, you need to apply the transfer rate as well, because timing numbers are in clock ticks. A 3600 (1800MHz) tick is faster / doesn't last as long as a 3200 (1600MHz) clock tick.

From what I've read, in most cases, the dual ranking benefit is larger than the transfer rate / latency differences, at least on Zen 3 and I think Zen 2 as well. At 3600 MT/s, CL 18 vs CL 16 is 10 ns vs 8.9 ns; a 10% occasional timing difference deep in the memory system is unlikely to matter much.

The best way to help 1% lows with Zen 3 is to trade for a 5800X3D, but I understand that that's an expensive way to do things.
 
Hey folks, curious where everyone ended up with overclocking their 5600X these days? Are you OCing in your BIOS or in AMD Master software? I've been a little out of the loop and a buddy just gifted me a 5600X. What should I expect out of overclocking this chip or does AMD have their stuff together and I should just let it boost on it's own?
Depends on the quality of your silicon. However, you can do a number of things. Play with your curve optimizer and undervolt the core to get max stable clocks so it boosts on it's own to max. You can crank the voltage and play with the frequencies manually. You should make sure you have at least 3600 Mhz DDR4 and that your Fabric Speed is set to 1:1. You can try squeezing more out of the RAM but maximum theoretical (not guaranteed) stable fabric Speeds end somewhere around 3733 (IIRC). Anything past that is trial and error.

About the best frequency I have seen on the 5000 Series is around 4650 (Stable). Anything more than that is hit or miss. My 5600X never really even got to 4600 Stable (it was my 5900X that got there). Don't get discouraged, a 4500 Mhz stable clock and good fabric speeds will yield you an excellent all around processor that can handle almost anything.
 
Keep in mind that 3200CL16 and 3600CL18 are exactly the same CAS latency; you can't compare timing numbers in a vacuum, you need to apply the transfer rate as well, because timing numbers are in clock ticks. A 3600 (1800MHz) tick is faster / doesn't last as long as a 3200 (1600MHz) clock tick.

From what I've read, in most cases, the dual ranking benefit is larger than the transfer rate / latency differences, at least on Zen 3 and I think Zen 2 as well. At 3600 MT/s, CL 18 vs CL 16 is 10 ns vs 8.9 ns; a 10% occasional timing difference deep in the memory system is unlikely to matter much.

The best way to help 1% lows with Zen 3 is to trade for a 5800X3D, but I understand that that's an expensive way to do things.
The 5800X3D is not miracle cure for everything. If you're gaming at 1080P, sure. It you're gaming at 1440P... maybe. If you game at 4K, the 5600X is just fine.

Single thread on the 5800X3D is poo.

It runs hot so you need to have ample cooling.
 
The 5800X3D is not miracle cure for everything. If you're gaming at 1080P, sure. It you're gaming at 1440P... maybe. If you game at 4K, the 5600X is just fine.

Single thread on the 5800X3D is poo.

It runs hot so you need to have ample cooling.

Isn't single thread performance the most important aspect for games? If it's poo does the cache compensate to that much of an extreme? I find this confusing seeing as sure it has extra cache however the chip can still out perform Intel 13th gen processors?
 
Isn't single thread performance the most important aspect for games? If it's poo does the cache compensate to that much of an extreme? I find this confusing seeing as sure it has extra cache however the chip can still out perform Intel 13th gen processors?
Single thread on the 5800X3D is hardly "poo". It's marginally slower than a 5800X, when you exclude cache effects, because the clock is a little lower. I personally don't consider a few percentage points, that you can't detect unless you're watching the fps meter, to be a big deal.

Whether the 5800X3D or a 13th gen wins will depend on what's executing. A lot of games seem to be memory intensive, and the extra cache will win. A game or game phase that's very cache friendly, and isn't multi-threaded, will probably run best on the 13th gen, or possibly a Zen 4.

Also, by the way, the 5800X / X3D doesn't run especially hot in the same sense as the Intel chips. It's a hard CPU to cool because all the heat is coming from a single off-center chiplet, so there's a small hot-spot and you need good heat transport from the cooler. The total heat dissipation is considerably less than say a 13900K or even a 13700K.
 
Isn't single thread performance the most important aspect for games? If it's poo does the cache compensate to that much of an extreme? I find this confusing seeing as sure it has extra cache however the chip can still out perform Intel 13th gen processors?
Single thread performance is more important in older titles and it's heavily leveraged in games like Civilization. The 5800X3D is slower in single threading then the 5600X, 5800X and the 5900X (all of which I have or had). My 13900K is 44% faster than my 5900X Overclocked in all aspects.

The 5800X3D shines in 1080P acceleration, where you really don't need any more FPS unless you are playing a competitive shooter. It loses ground at 1440P and is effectively a waste of money if you're planning on playing games at 4K. The 13900K beats the 5800X3D to DEATH in 4K. Anywhere from 10-20+ FPS faster.

I dwell almost exclusively in the 4K spectrum. I also do tend to use a lot of stuff that is single thread sensitive.

So, you purchase what you need based on your use case.

In my opinion, the only processor that was a shift up from the 5600X is my 5900X, the difference between the 5600 and 5800 is essentially indistinguishable since most games barely leverage it's 6 cores / 12 threads. The 5800X3D is a great CPU if you want to bolster the longevity of the AMD X3-500 platform. It's a great gaming CPU but it's not perfect for everyone. People can squak all they want about how damn good it (5800X3D) is but a 5900X is either faster or within 1-5 FPS in almost all titles. The 5900X is a better value to anyone with is a content creator / needs productivity speed.

13th Gen is arguably a much better processor than the AMD lineup in 5000 and 7000 series with the AMD processors taking certain wins here and there. The 7000 Series is a productivity beast.

I have encountered anomalous issues with programs and AMD's architecture since the my Ryzen 1700. The processors either run certain programs like dogshit or they cause the programs to crash at the worst possible times. I haven't experienced any of those issues on the myriad of my Intel platforms. Maybe this is because most programs are compiled with Intel Optimization in mind... maybe not. But My Affinity Suite of Publishing, Photo and Designer applications have all crashed randomly while using various AMD processors for the past 3-4 years now. Serif Page Plus X9 is rock solid on Intel and runs like ass on AMD. It will all come down to what you do with your computer.

Personally, I think it's almost embarrassing that AMD on 5nm can't even pull out a clear win against Intel at 10nm.

Up until I got burned twice on the 7000 Series platform I was AMD's biggest fanboy. I tried the 13900K out for a test drive and I sold my soul to team Blue.... Holy Shit! So much easier to use, to set up and just blew me out of the water.

If you're on the (AMD) platform and you are trying to save money, the 5600X is fine processor. It performs the same as a 5800X in Cyberpunk (initially there was a bug where it performed worse but that was fixed).
 
Single thread on the 5800X3D is hardly "poo". It's marginally slower than a 5800X, when you exclude cache effects, because the clock is a little lower. I personally don't consider a few percentage points, that you can't detect unless you're watching the fps meter, to be a big deal.

Whether the 5800X3D or a 13th gen wins will depend on what's executing. A lot of games seem to be memory intensive, and the extra cache will win. A game or game phase that's very cache friendly, and isn't multi-threaded, will probably run best on the 13th gen, or possibly a Zen 4.

Also, by the way, the 5800X / X3D doesn't run especially hot in the same sense as the Intel chips. It's a hard CPU to cool because all the heat is coming from a single off-center chiplet, so there's a small hot-spot and you need good heat transport from the cooler. The total heat dissipation is considerably less than say a 13900K or even a 13700K.
The single thread is not poo, I was being nice, it's dogshit slow. My 5900X is a couple percent faster in Single Thread than the 5800X3D and my 13900K beats both of them to death in that metric. Being nearly 50% faster in Single Thread than the 5800X3D.

The Cache memory is where the X3D parts shine but they only manage to pull out wins in the lower resolutions. Intel is generally better at 4K to insanely better.

The chip does run hot, that's a fact. Doesn't matter where it's coming from. It's coming from the same damn place it does on the 5600 and 5800 processors. It's not insurmountable, but it runs hot.

Comparing a 7nm design that runs at 105 Watts to a 10nm (inferior process node) that runs up to 250 Watts is a poor comparison.
 
I'm just running +100 auto overclock on the 5600x / Assassin is in balance fan mode, it has 4 sticks of single rank Muskin Redline running 3600Mhz cl 16 and that is the Sapphire Pulse RX 6700 on 22 .11 .2 and I play in 4K.

 
The single thread is not poo, I was being nice, it's dogshit slow. My 5900X is a couple percent faster in Single Thread than the 5800X3D and my 13900K beats both of them to death in that metric. Being nearly 50% faster in Single Thread than the 5800X3D.
No, this is false, or at best misleading. The 5800X3D is not half the speed, single thread, of a 13900K, unless you're looking at some very specific benchmark that favors the Intel CPU. They are much closer. One site (techpowerup) concludes that the 5800X3D is about 7% slower single thread than the 13900K, on average.

The 5800X3D is about 7% slower than a 5800X, single core, ignoring cache effects. That's hardly "dogshit slow".

The Cache memory is where the X3D parts shine but they only manage to pull out wins in the lower resolutions. Intel is generally better at 4K to insanely better.

The chip does run hot, that's a fact. Doesn't matter where it's coming from. It's coming from the same damn place it does on the 5600 and 5800 processors. It's not insurmountable, but it runs hot.
You are confusing total heat output with heat transport ability. It certainly does matter where it's coming from. X watts spread across a square inch is a lot easier to cool than X watts coming from an 80 mm2 area, and that's because there's more area for conduction to move heat away from. Total heat output of a 5800X3D is significantly less than the total heat output of a 13900K. That's simple fact.

Yes, the 5800X3D (and 5800X) are not easy to cool, but it's not because of heat total output, it's because of the combination of total dissipation and heat source area. The 13900K is hard to cool because it puts out a bloody great amount of heat, total.

Feel free to hate on the 5800X3D for your own use cases as much as you like, but please don't promulgate misinformation.
 
3440x1440 5600x and a 3090. Any reason to upgrade? For fps gaming mostly. The game I play a lot is more single thread. Planetside.
 
Last edited:
No, this is false, or at best misleading. The 5800X3D is not half the speed, single thread, of a 13900K, unless you're looking at some very specific benchmark that favors the Intel CPU. They are much closer. One site (techpowerup) concludes that the 5800X3D is about 7% slower single thread than the 13900K, on average.

The 5800X3D is about 7% slower than a 5800X, single core, ignoring cache effects. That's hardly "dogshit slow".


You are confusing total heat output with heat transport ability. It certainly does matter where it's coming from. X watts spread across a square inch is a lot easier to cool than X watts coming from an 80 mm2 area, and that's because there's more area for conduction to move heat away from. Total heat output of a 5800X3D is significantly less than the total heat output of a 13900K. That's simple fact.

Yes, the 5800X3D (and 5800X) are not easy to cool, but it's not because of heat total output, it's because of the combination of total dissipation and heat source area. The 13900K is hard to cool because it puts out a bloody great amount of heat, total.

Feel free to hate on the 5800X3D for your own use cases as much as you like, but please don't promulgate misinformation.
"The 5800X3D is about 7% slower than a 5800X, single core, ignoring cache effects. That's hardly "dogshit slow"."
That is dogshit slow compared to my 13900K which is 44% than my 5900X which is in turn, faster than the 5800X in Single Thread. That's 51+% SLOWER than my 13900K. That's simply reality.

Heat is heat. The 5800X3D can hit 90c, doesn't matter how it happens but it happens. Is it easier to cool than a 13900K? Likely. Can it outperform one? No. Possibly in 1080P gaming only but anything else not so much.

If you want to help people make informed decisions, help them. Do not provide ambiguous bullshit based off of "feels". AMD is the clear looser here this generation. 5000 Series is old news. The 7000 Series trades blows with 13th Gen but does not pull ahead with a significant win. Depends heavily on use case.

"Feel free to hate on the 5800X3D for your own use cases as much as you like, but please don't promulgate misinformation."
I don't hate the processor, just the people that spread misinformation about it. It's good for low end gaming, not for the high end.

Also, if I am not mistaken this thread is about how to eek performance out of the 5600X... NOT about why the OP should purchase the 5800X3D. Help the OP and stop fighting a losing battle.
 
3440x1440 5600x and a 3090. Any reason to upgrade? For fps gaming mostly. The game I play a lot is more single thread. Planetside.
At 1440P you start to have less impact with a CPU and more with your choice of GPU. You will likely see a moderate bump in performance with a 5800X3D or 5900X. What that means is you will see a "Couple" Frames per Second. The 5800X3D may deliver better 1% lows. The 3090 TI is a robust GPU that should perform rather well. The only upgrade you should consider at this point is a move to a new platform, eventually. If you want a performance boost wait until the 5900X and 5800X3D are cheap and then drop one in. You use the 5900X if you are into productivity and the 5800X3D if you primarily game on your pc.

I personally have the 5900X and it was my upgrade from the 5600X. I generally don't like to comment on feelings but it was more responsive and much faster than my 5600X. I could "feel" the difference.

Likewise, when I moved from the 5900X to my 13900K I was astonished! The performance leap was staggering!

Reality: 5600X is a fine CPU that delivers excellent performance on a budget. Only upgrade to the 5900X or 5800X3D if the price dips below 300 dollars. Otherwise you are wasting your money.
 
Last edited:
What I question is where you get that 51% number from. It must be a scenario specific to something that you are doing, because it doesn't match up with any reviews or benchmarks that I've seen anywhere else.

In any case I'm not interested in arguing with you. You make one post bashing "feelings" (which I said nothing about) and your next post extols how much better the 13900K "feels" to you. Sigh.
 
Well dudes....... So I tracked down a set of 32GB 2 x 16GB CL16 3600 sticks that I could actually afford. These are not on my QVL list but I'm told many 32GB sticks aren't on QVL lists. These are Kingston Renegades. What are my chances I'll be able to run theses without issue do you think? I currently run some crappy Oloy in my system without issue that aren't on my QVL list but these also run at dingleberry speeds.
 
Well dudes....... So I tracked down a set of 32GB 2 x 16GB CL16 3600 sticks that I could actually afford. These are not on my QVL list but I'm told many 32GB sticks aren't on QVL lists. These are Kingston Renegades. What are my chances I'll be able to run theses without issue do you think? I currently run some crappy Oloy in my system without issue that aren't on my QVL list but these also run at dingleberry speeds.
If its 16gb x 2 stick, I'm pretty sure it will run the xmp (3600).
Kingston usually used Hynix C/J Die or Micron E-Die for their 3600 rated kits so no worries about the QVL.
Ryzen 5000 optimized for ram speed at 3600++
 
Well dudes....... So I tracked down a set of 32GB 2 x 16GB CL16 3600 sticks that I could actually afford. These are not on my QVL list but I'm told many 32GB sticks aren't on QVL lists. These are Kingston Renegades. What are my chances I'll be able to run theses without issue do you think? I currently run some crappy Oloy in my system without issue that aren't on my QVL list but these also run at dingleberry speeds.
worse case is xmp wont load right and youll have to manually set speed/timing/voltage. its what i had to do on my sig rig.
 
Well dudes....... So I tracked down a set of 32GB 2 x 16GB CL16 3600 sticks that I could actually afford. These are not on my QVL list but I'm told many 32GB sticks aren't on QVL lists. These are Kingston Renegades. What are my chances I'll be able to run theses without issue do you think? I currently run some crappy Oloy in my system without issue that aren't on my QVL list but these also run at dingleberry speeds.
You will probably be fine. Otherwise as pendragon1 said. Should be easy enough either way.
 
Does this mean it will run in dual rank at 3600 but not higher? From my mobo's website.

  • 4x DDR4 memory slots, support up to 128GB 1
    • Supports DDR4 1866/ 2133/ 2400/ 2667/ 2800/ 2933/ 3000/ 3066/ 3200 MHz by JEDEC
    • Supports DDR4 2667/ 2800/ 2933/ 3000/ 3066/ 3200/ 3466/ 3600/ 3733 / 3866/ 4000/ 4133/ 4266/ 4400+ MHz by A-XMP OC MODE
      • 1DPC 1R max speed 4400 MHz
      • 1DPC 2R max speed 3866 MHz
      • 2DPC 1R max speed 4000 MHz
      • 2DPC 2R max speed 3600 MHz
 
Does this mean it will run in dual rank at 3600 but not higher? From my mobo's website.

  • 4x DDR4 memory slots, support up to 128GB 1
    • Supports DDR4 1866/ 2133/ 2400/ 2667/ 2800/ 2933/ 3000/ 3066/ 3200 MHz by JEDEC
    • Supports DDR4 2667/ 2800/ 2933/ 3000/ 3066/ 3200/ 3466/ 3600/ 3733 / 3866/ 4000/ 4133/ 4266/ 4400+ MHz by A-XMP OC MODE
      • 1DPC 1R max speed 4400 MHz
      • 1DPC 2R max speed 3866 MHz
      • 2DPC 1R max speed 4000 MHz
      • 2DPC 2R max speed 3600 MHz
1. Not all 32gb kit (16gb x2 sticks) now are dual rank kit. My previous 2 friend's build have 2 of those kits but both are using single rank kits.
2. It depends on your procie's IMC & the ram IC's quality
> IMC --> Usually for FCLK 1:1 UCLK, ryzen 5000 can go as far as 4000 or more on better IMC, at least 3800/3933 is achievable.
> IC's quality --> if it's A2 / B2 PCB, then it can clock 4000++ easier compared to the other PCB.
 
Interesting, waiting to hear from the seller since the ram is on backorder apparently now........ Out of curiosity do ram manufacturers make DDR5 in single rank or does DDR5 only come in dual rank kits?
 
Interesting, waiting to hear from the seller since the ram is on backorder apparently now........ Out of curiosity do ram manufacturers make DDR5 in single rank or does DDR5 only come in dual rank kits?
if I'm not wrong, the 8gb stick is single rank, and 16gb stick and up are dual rank.
 
if I'm not wrong, the 8gb stick is single rank, and 16gb stick and up are dual rank.
Interesting that would be a positive since DDR4 kits are all over the place. First of all amazing info from Kingston. Most manufacturers are complete piss with their ram details. You might have to hop around to websites looking for reviews. Corsair is kind meh for info, team group good luck. Kingston though most impressive they have pdf's on most sets.

IE: the DDR4 set I ordered KF436C16RB1K2_32
1674662096239.png



However another common DDR4 3600 kit from Kingston: KF436C18BBK2/32 single.
1674662313057.png


Anyway, this has been super annoying in searching for kits to find dual rank and c16 kits at an affordable price. What would be the equivalent in DDR5 to cl14 3200 kits which seem to be the holy grail of DDR4, CL36 6000 or something? I know this is sort of off topic but I've only recently started looking at ram more.
 
I'll post an update on OC'ing/settings with this chip when I can work on my PC it might be a while. Feel free to chime in any additional tips for the 5000 series.
 
Interesting that would be a positive since DDR4 kits are all over the place. First of all amazing info from Kingston. Most manufacturers are complete piss with their ram details. You might have to hop around to websites looking for reviews. Corsair is kind meh for info, team group good luck. Kingston though most impressive they have pdf's on most sets.

IE: the DDR4 set I ordered KF436C16RB1K2_32
View attachment 544260


However another common DDR4 3600 kit from Kingston: KF436C18BBK2/32 single.
View attachment 544261

Anyway, this has been super annoying in searching for kits to find dual rank and c16 kits at an affordable price. What would be the equivalent in DDR5 to cl14 3200 kits which seem to be the holy grail of DDR4, CL36 6000 or something? I know this is sort of off topic but I've only recently started looking at ram more.
You can't get the AMD platform to run past a certain level of 1:1 fabric Speed. You can run faster RAM but you will get diminishing returns or performance hits unless it's sufficiently fast enough with the Fabric Clock unsynced.

You're likely looking for something like DDR4 3600 CL15 for 1:1 fabric and affordable price (I think you can actually get it in CL14). Ryzen can handle up to 4000 Mhz RAM at 1:1 Fabric if you are ULTRA LUCKY. AMD's posted 1:1 Fabric Limit is 3733 for the 5000 series. Some users on here are reporting up to 3866 in 1:1.

DDR5 6000 parity is generally around DDR4 4133 with tight timings (on Intel). I am doing it with DDR4 4366 with Gear 1 and CL 19 timings. I'm within a couple (1-2.4) percent of DDR5 7400.
 
I'll post an update on OC'ing/settings with this chip when I can work on my PC it might be a while. Feel free to chime in any additional tips for the 5000 series.
Undervolt with the curve optimizer if you want a hands off OC approach. The system will boost the Processors to nearly max clocks if they are running cool.
 
Is there a tutorial or walkthrough for undervolting, curve optimizer, and PBO etc for the 5000 series? I haven't looked a whole lot but I'm used to Core 2 Duo and Core 2 Quad whatever OCing.
 
Is there a tutorial or walkthrough for undervolting, curve optimizer, and PBO etc for the 5000 series? I haven't looked a whole lot but I'm used to Core 2 Duo and Core 2 Quad whatever OCing.
Sort of, I'm fairly certain I watched a Youtube video about it from some rando tech dude. You have to play with the undervolt and curve settings but it works and it's like getting a free stable frequency boost out of the CPU. I can't really call it overclocking because these chips were rated to run at certain speeds and I never really saw them hit them on a consistent basis. It's really not until the 7000 series chips that AMD takes off some of the breaks and lets their CPUs run to their max limit.

When I OC'ed my 5900X the maximum stable, all core, clocks I could get were 4650 Mhz. That's hardly an OC. I couldn't even go higher disabling CCXes. Couldn't get a single core stable at 5Ghz. Anything past 4650Mhz required an assload of power and cooling and my processor started thermal throttling and the system was shutting down to avoid damage. But it would run 4650, all core, all day, stable.
 
Ahh well overclocking probably is the wrong word. This bitch gets pretty hot, I wouldn't mind keeping performance but lowering heat which seems to be pretty easy to do with Zen3 but I don't know what I'm doing. :D
 
Ahh well overclocking probably is the wrong word. This bitch gets pretty hot, I wouldn't mind keeping performance but lowering heat which seems to be pretty easy to do with Zen3 but I don't know what I'm doing. :D
if you dont want to mess with it a bunch just set a cpu offset voltage of -.03. should be fine and knock the temp down a bit.
 
Back
Top