Ryzen - Tweaks?

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
I've already sold my 7700k system to a friend so I'm staying with the 3700x. I'm just going to have to come to terms with it being more of a sidegrade than an upgrade to a 3 year old CPU. It's close enough in most games although I've tried League of Legends and it's WAY worse than the 7700k was (220fps vs 350+fps with the 7700k). I'll just hope that there are some games that come out that I enjoy where it's a real improvement. Hopefully the 4000 series will build upon it.

I don't really know what I expected. I "upgraded" from an i5-6600k to a 1700x when they first came out and it was mostly a downgrade so I returned it for the 7700k which was a big improvement in most games. Now I've gotten rid of the 7700k for a 3700x and it again isn't really an upgrade. I miss the old days when I could buy new hardware and get meaningful improvements.
 
Last edited:

Jamie Marsala

Limp Gawd
Joined
Mar 9, 2016
Messages
293
I've already sold my 7700k system to a friend so I'm staying with the 3700x. I'm just going to have to come to terms with it being more of a sidegrade than an upgrade to a 3 year old CPU. It's close enough in most games although I've tried League of Legends and it's WAY worse than the 7700k was (220fps vs 350+fps with the 7700k). I'll just hope that there are some games that come out that I enjoy where it's a real improvement. Hopefully the 4000 series will build upon it.

I don't really know what I expected. I "upgraded" from an i5-6600k to a 1700x when they first came out and it mostly a downgrade so I returned it for the 7700k which was a big improvement in most games. Now I've gotten rid of the 7700k for a 3700x and it again isn't really an upgrade. I miss the old days when I could buy new hardware and get meaningful improvements.

May I ask why you did not bump up to the 3800x? That most likely would have been a slight upgrade that way.
 

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
May I ask why you did not bump up to the 3800x? That most likely would have been a slight upgrade that way.

They were sold out at my local MC and all of the reviews I checked out showed that there is basically no difference in performance anyway.
 

Jamie Marsala

Limp Gawd
Joined
Mar 9, 2016
Messages
293
Well I decided to do a windows re-install today, swapped through three motherboards and 4 different GPU's and two CPU's on that install. Anyway fresh install and I got better scores/performance all around. Also decided not to install Dragon Center because it seemed to screw up Ryzen Master, which I use for monitoring sometimes. The only reason I used Dragon Center was for some monitoring, can do the same with HWInfo64, and to control the one set of LED's on my MSI X570 Ace MB. But it seems to have held the color I had set so I un-installed it after I tried to open RM and it failed.
 

Jamie Marsala

Limp Gawd
Joined
Mar 9, 2016
Messages
293
Well that sounds like a good time LOL! I guess we can feel blessed its not as painful and time consuming as it was with Windows 95 or 98. ;)

Nope, took all of about 3 hours total and that was with updating the three gaming apps I use and downloading some 70GB of a game update. All windows updates and software were back on in like an hour and a half. Definetely better than the past versions, also helps that my two drives are both NVMe M.2. Which is also why I re-installed, I upgraded my boot drive from a 240 to a 512GB and swapped my data drive from a 1TB SATA SSD to a 1TB M.2. Figured it was time to do a fresh install.
 

Jamie Marsala

Limp Gawd
Joined
Mar 9, 2016
Messages
293
This is what I get after a clean install of Windows and the tweaks I previously posted.
 

Attachments

  • R20-02192020.jpg
    R20-02192020.jpg
    148.4 KB · Views: 0

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
I returned the Asus B450 board and picked up the Gigabyte X570 I AORUS PRO WIFI. One nice thing about the Gigabyte board is that the CPU Voltage when set to Auto is only 1.22V while the Asus board's Auto setting would always be at 1.4-1.44V. I also like the BIOS layout better than the Asus board but it's still pretty annoying that no ITX-compatible AM4 boards other than the $430 Asus Crosshair VIII Impact have an internal USB-C header when $150 Z390 boards have them.
 

kilroy67

[H]ard|Gawd
Joined
Oct 16, 2006
Messages
1,131
I returned the Asus B450 board and picked up the Gigabyte X570 I AORUS PRO WIFI. One nice thing about the Gigabyte board is that the CPU Voltage when set to Auto is only 1.22V while the Asus board's Auto setting would always be at 1.4-1.44V. I also like the BIOS layout better than the Asus board but it's still pretty annoying that no ITX-compatible AM4 boards other than the $430 Asus Crosshair VIII Impact have an internal USB-C header when $150 Z390 boards have them.
I read a post from last summer of AMD-Robert talking about setting the CPU VDDP voltage from auto to normal. It made a world of difference for me in boosting my clocks by allowing the voltage to ramp up higher.
 

Mauler87

Weaksauce
Joined
Dec 25, 2014
Messages
75
Make sure you're using the latest Ryzen 2 Chipset drivers, the 1usmus power plan (i found 1usmus universal settings is best for myself with gaming) as well as Process Lasso to set affinity of background apps to slower cores and threads.

Process lasso also helps to induce performance mode just make sure performance mode doesn't change the power plan, that can be changed in the options.

Performance mode helps greatly with foreground apps and the latest chipset drivers combined with the most recent BIOS updates allows 1usmus' power plan to boost the correct threads based on your workload instead of praying the scheduler plays nice with the standard ryzen power plans (they don't as much and 1usmus' plan forces correct core boosting).

You should find after doing all that that your ryzen CPU performs a bit better. Below is my test I did over the last few days, also my 3900x beats my 8700k in CBR20 by a few points ST with background apps loaded and by 20+ points with a clean boot.

Video for comparison in a few games.
 

Shikami

Gawd
Joined
Apr 5, 2010
Messages
792
Don't know if this provides a point of reference since I am usually the opposite around here. That is, no overclocking, no multi-core enhancement, no uber RAM (dual rank XMP 15-15-15 2666), and balanced power. I get 4900/497 on my i7-9900K.
 

Attachments

  • Cinebench20.png
    Cinebench20.png
    1.1 MB · Views: 0

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
1582845647039.png


That's with all of the tweaks discussed in this thread. While the single core test was running I watched Ryzen Master and saw Core 8 boosting up to a max of 3,386 Mhz (probably an average high of 3,350) and then handing off to Core 6 and boosting up to a max of 3,354 Mhz (average of around 3,345 Mhz) and back and forth; temps never went over 56C.
 

Mauler87

Weaksauce
Joined
Dec 25, 2014
Messages
75
Don't know if this provides a point of reference since I am usually the opposite around here. That is, no overclocking, no multi-core enhancement, no uber RAM (dual rank XMP 15-15-15 2666), and balanced power. I get 4900/497 on my i7-9900K.

Totally different architecture, Without a cross CCX communication ryzen tears through the intel sku's in nearly all synthetics. It's not surprising that intel does not need "uber ram" to come out at that score. Plus you would hope their performance didn't tank after being stuck on 14nm for so long. I love my 8700k don't get me wrong, and overclocking it's ram provided a decent bump in games (tfaw, tcke and trfc clocks mostly for fps in games), but productivity it just is miles ahead. I would rather cop a hatchet to the back of my skull and spend the rest of my days using crayons to color books than to use the igpu for hardware accelerated exporting in adobe. x264 is just so much better, smaller file size, better quality at the same bitrate (pixel peeping of course though it can get blocky on quicksync) and is comparatively as fast at times.

Power is also a thing here though, in Australia we pay 25 american cents per kilowatt hour used. I think it may be more now even....sad days....anyway between dynamic boosting and sustained boosts I love the efficiency and performance ryzen now provides with a more mature ecosystem in relation to BIOS/Drivers/Windows updates.

Great things ahead hopefully with the ryzen 3 announcements.
 

Shikami

Gawd
Joined
Apr 5, 2010
Messages
792
Totally different architecture, Without a cross CCX communication ryzen tears through the intel sku's in nearly all synthetics. It's not surprising that intel does not need "uber ram" to come out at that score. Plus you would hope their performance didn't tank after being stuck on 14nm for so long. I love my 8700k don't get me wrong, and overclocking it's ram provided a decent bump in games (tfaw, tcke and trfc clocks mostly for fps in games), but productivity it just is miles ahead. I would rather cop a hatchet to the back of my skull and spend the rest of my days using crayons to color books than to use the igpu for hardware accelerated exporting in adobe. x264 is just so much better, smaller file size, better quality at the same bitrate (pixel peeping of course though it can get blocky on quicksync) and is comparatively as fast at times.

Power is also a thing here though, in Australia we pay 25 american cents per kilowatt hour used. I think it may be more now even....sad days....anyway between dynamic boosting and sustained boosts I love the efficiency and performance ryzen now provides with a more mature ecosystem in relation to BIOS/Drivers/Windows updates.

Great things ahead hopefully with the ryzen 3 announcements.

Oh, I'm with you, not against you. This was only me making a reference using this system for his tweaks and scores, and for me to see the score base for Cinebench 20 because I never ran it till now due to this thread. He did manage to get a saucy %'s. I saw that many in my above range were overclocking with LN2 for 9900K's and not getting much at all. 100/200 points difference is not anything in comparison to my system for cost and work of the overclock. I am bit of an anti-overclocker because there is a lot of mis-direction with it, I feel. For example, what is 500 points more in Cinebench if you crash in a few or many applications, and experience nuances with system use?

I'm actually very excited about Ryzen 3, and this system would have been a Ryzen 2 if actually it wasn't going to be way cheaper to go Intel at the time (ya I know). I was personally waiting for the Ryzen's FPU to finally have 256bit AVX to purchase. Finally it did, but then I saw total cost, and there was some good pricing at the time for me with Intel. Would of rather a Threadripper, and again, I am glad I did purchase instead of waiting. MY GOD I NEED TO SALE ONE OF ME KIDNEYS, LUNGS, AND PARTIAL LIVER TO AFFORD!

Stay safe Down Under
 

Shikami

Gawd
Joined
Apr 5, 2010
Messages
792
That's with all of the tweaks discussed in this thread. While the single core test was running I watched Ryzen Master and saw Core 8 boosting up to a max of 3,386 Mhz (probably an average high of 3,350) and then handing off to Core 6 and boosting up to a max of 3,354 Mhz (average of around 3,345 Mhz) and back and forth; temps never went over 56C.

Have you tried getting a baseline of your stock performance first? With the system doing ACPI CPPC, CStates enabled, XFR, balanced power (or whatever optimal Ryzen power setting). RAM at XMP or manual config for its timings, as it should be, for all of the tests. Then doing PBO off/on?

You are very close to my system. You should have a 3.6 base clk that boosts a bit higher, no? And a single core ~4.3? With typical baseline configuration....


And to note, there may be an issue with single core boosting from Cinebech. I can easily and do hit 5ghz single/dual core (as you can see in HWMonitor, and that 5Ghz is not from Cinebench) but it clocks at 4.9 all cores and seems to toss the clock around even though one core is pegged for work? Although, IIRC, the 5GHz is only to last a short time, and not full clock for single and dual core operation due to threads and OS handling.
 

Attachments

  • 5GHz.png
    5GHz.png
    285.9 KB · Views: 0
Last edited:

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
Have you tried getting a baseline of your stock performance first? With the system doing ACPI CPPC, CStates enabled, XFR, balanced power (or whatever optimal Ryzen power setting). RAM at XMP or manual config for its timings, as it should be, for all of the tests. Then doing PBO off/on?

You are very close to my system. You should have a 3.6 base clk that boosts a bit higher, no? And a single core ~4.3? With typical baseline configuration....


And to note, there may be an issue with single core boosting from Cinebech. I can easily and do hit 5ghz single core (as you can see in HWMonitor, and that 5Ghz is not from Cinebench) but it clocks at 4.9 and seems to toss the clock around even though one core is pegged for work. Although, IIRC, the 5GHz is only to last a short time, and not full clock for single and dual core operation.

IIRC at stock settings Multi Core was in the realm of 4700 points; I didn't test Single Core. I have my current config saved as a BIOS profile so maybe I'll try resetting the system and doing some more testing when I get home tonight.
 

Starrbuck

2[H]4U
Joined
Jun 12, 2005
Messages
2,727
Make sure you're using the latest Ryzen 2 Chipset drivers, the 1usmus power plan (i found 1usmus universal settings is best for myself with gaming) as well as Process Lasso to set affinity of background apps to slower cores and threads.

Process lasso also helps to induce performance mode just make sure performance mode doesn't change the power plan, that can be changed in the options.

Performance mode helps greatly with foreground apps and the latest chipset drivers combined with the most recent BIOS updates allows 1usmus' power plan to boost the correct threads based on your workload instead of praying the scheduler plays nice with the standard ryzen power plans (they don't as much and 1usmus' plan forces correct core boosting).

You should find after doing all that that your ryzen CPU performs a bit better.

Does Process Lasso need any special configurations or does it do this on its own?
 

CyberJunk

Supreme [H]ardness
Joined
Nov 13, 2005
Messages
4,097
Hey everyone, so I just built a 3700x system and while the synthetic benchmarks are amazing the gaming performance isn't what I expected (with a 2080ti @ 1440p it's anywhere from a bit slower to equal to my old 7700k@4.9Ghz). So I'm looking for things that I may need to do to help it stretch it's legs.

Some of the things I've tried:

1) Set RAM to 3600Mhz and IF to 1800Mhz. Used the Ryzen Mem Calc and tweaked the RAM timings. This helped quite a bit (AC:Odyssey gained like 3fps on the minimum framerate and 14fps on the maximum).
2) Installed Ryzen Master but haven't messed with it that much yet as it seems to reduce the desktop refresh rate to like 20hz? As soon as it's open the mouse movement on the windows desktop just dies.
3) Installed the AMD Ryzen drivers and switched to the Ryzen power plan.

I haven't messed with PBO/overclocking yet as the Asus ROG Strix B450-I board's BIOS is convoluted as all hell and I haven't looked up any guides on it. I mean most motherboards would have have an XMP setting to enable but on this thing the memory timings are enabled through D.O.C.P...

What I've been used to with Intel systems is basically just, "Set FSB, set Multiplier, install chipset drivers, done" but it seems like there's quite a bit more to getting Ryzen working optimally. Is there anything that I'm missing?


Check out Buildzoids youtube channel he did a video on how to set PBO limits for the 3700x in Bios
 

Mauler87

Weaksauce
Joined
Dec 25, 2014
Messages
75
I've already sold my 7700k system to a friend so I'm staying with the 3700x. I'm just going to have to come to terms with it being more of a sidegrade than an upgrade to a 3 year old CPU. It's close enough in most games although I've tried League of Legends and it's WAY worse than the 7700k was (220fps vs 350+fps with the 7700k). I'll just hope that there are some games that come out that I enjoy where it's a real improvement. Hopefully the 4000 series will build upon it.

I don't really know what I expected. I "upgraded" from an i5-6600k to a 1700x when they first came out and it was mostly a downgrade so I returned it for the 7700k which was a big improvement in most games. Now I've gotten rid of the 7700k for a 3700x and it again isn't really an upgrade. I miss the old days when I could buy new hardware and get meaningful improvements.


If you want extra performance from your chip i suggest you dont run wallpaper engine (it dropped my max boost by 100mhz on my 3900x). Install process lasso, and .ove the permanent affinity for all non windows system processes and apps like steam etc... to the last chiplet or whichevet isnt your fastest. Do NOT reassign affinity for games you play as it can tank pefformance. Take hunt: showdown, only uses my physical cores by default across both chiplets. Changing its affinity cut my fps by 20, also do not allow bitsum highest performance power plan to run.

Go into options as well and enable forced mode so affinity's stick and are updated on the fly. And for apps that induce performance mode, change performance modes default power plan to the 1usmus universal (if you dont have it i suggest you get it).

Hopefully these tips net you a sizeable performance boost, they allowed my 3900x to fly past my 4.9ghz 8700k even further (20fps on average in farcry 5). Also got me 1k more gpu score in firestrike somehow.
 

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
I reset the BIOS to Optimized defaults + XMP and received the below. So about a 100 point difference in Multicore and a 2 point difference in single core. No point in even spending any time in the BIOS adjusting memory and PBO settings....

1583818026281.png
 

tunatime

Well...OK
Joined
Sep 15, 2011
Messages
4,479
What are you bosting to? My 3960x stays at 4.475-4.6 with peaks to 4.675 95% time over 4.5 gaming and about 4.3 allcore.

It destroys my old 6850k in everything it might just be you had unrealistic expectations going form a Intel at 4.9 to and at 4.5
 

bobzdar

[H]ard|Gawd
Joined
Jun 6, 2003
Messages
1,835
You would have run into the same thing with the newer Intel CPU's as well - they aren't really any faster in gaming if the game doesn't need more than 4c/8t. However, there are a bunch that gain on 1% lows by having the extra threads and they allow you to do stuff like stream without impacting game play or stream quality. Even an intel 9900KS only gets 508 single thread, or 1-2% more than you get with your vastly cheaper 3700X. Most new AAA games are using more than 8t if available.

https://www.cgdirector.com/cinebench-r20-scores-updated-results/
 

gerardfraser

[H]ard|Gawd
Joined
Feb 23, 2009
Messages
1,113
I reset the BIOS to Optimized defaults + XMP and received the below. So about a 100 point difference in Multicore and a 2 point difference in single core. No point in even spending any time in the BIOS adjusting memory and PBO settings....

I do not understand why you do not want free performance by change a couple things in BIOS.It is fine but here are my settings with 3800X note temperatures and score,that could be you in all the Cinebench20 glory lol.

Single Thread-530 score
Multi Thread -5260 score
Idle temperature CPU-28°C
Max temperature CPU-71°C
 

pendragon1

Fully [H]
Joined
Oct 7, 2000
Messages
27,365
I do not understand why you do not want free performance by change a couple things in BIOS.It is fine but here are my settings with 3800X note temperatures and score,that could be you in all the Cinebench20 glory lol.

Single Thread-530 score
Multi Thread -5260 score
Idle temperature CPU-28°C
Max temperature CPU-71°C
maybe they arent doing anything were it matters. my sig system is the first system ever that i havent bothered to oc, wasnt much of a point pbo works fine.
 

gerardfraser

[H]ard|Gawd
Joined
Feb 23, 2009
Messages
1,113
PBO always worked fine me also.I do get more performance for free with changing one setting,I am fine with that.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,617
What are you bosting to? My 3960x stays at 4.475-4.6 with peaks to 4.675 95% time over 4.5 gaming and about 4.3 allcore.

It destroys my old 6850k in everything it might just be you had unrealistic expectations going form a Intel at 4.9 to and at 4.5

lol that's like comparing a Lambo V12 LM002 to a Honda CRV.
 

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
I do not understand why you do not want free performance by change a couple things in BIOS.It is fine but here are my settings with 3800X note temperatures and score,that could be you in all the Cinebench20 glory lol.

If you read the thread I've tried just about everything out there and there just doesn't seem to be much to be gained on this chip. Also, your score is probably being influenced by your 1900Mhz IF; I can't seem to get above 1800Mhz though I'm not sure if it's the CPU or Memory that is at fault. The temps on mine are similar and the all core boosts the same as yours.
 

trick0502

Supreme [H]ardness
Joined
Apr 17, 2006
Messages
5,435
I reset the BIOS to Optimized defaults + XMP and received the below. So about a 100 point difference in Multicore and a 2 point difference in single core. No point in even spending any time in the BIOS adjusting memory and PBO settings....

View attachment 228984

enabled xmp, the ram to 3800mhz, IF 1900mhz and set the multy to 43. i lost about 10 points off single thread but picked up 400 points multy. games play a lot smoother like this.

43.jpg
 

gerardfraser

[H]ard|Gawd
Joined
Feb 23, 2009
Messages
1,113
If you read the thread I've tried just about everything out there and there just doesn't seem to be much to be gained on this chip. Also, your score is probably being influenced by your 1900Mhz IF; I can't seem to get above 1800Mhz though I'm not sure if it's the CPU or Memory that is at fault. The temps on mine are similar and the all core boosts the same as yours.
Not trying to trick you,I have read the thread and posted way to much in thread.I was trying to help you.Your chasing and spending cash on nothing that does not make a difference in performance ,it is all just numbers.Stop ,return your X570 and get your money back.The 450 board is/was fine. You can always go back to your old Intel setup or upgrade Intel system.It appears the AMD system may be to much to deal with,it happens ,you have a what you think should be the performance and that was not met by the new AMD PC.

Memory has nothing to do with it.Here is ,oh god I hate cinebench but 2133Mhz is still same result.Even PC gaming with 2080ti at 2560x1440 with 2133Mhz Ram has nothing to do with your perceived under performance problems.

Check Ram speeds
49380259732_0bfa63122d_k.jpgMemory-Chart by gerard fraser, on Flickr
49648510421_812ea21b65_k.jpg2133Mhz cinebench20
 

Bankie

[H]ard|Gawd
Joined
Jul 27, 2004
Messages
1,770
As you can see on the post above you he picked up 400 points from setting the IF to 1900mhz....

You act like I haven't done any tweaks. I've done everything in this thread and more from 50 different PBO configs to all-core overclocks; this includes your settings on the first page. Tweaking PBO and using Ryzen Memory Calculator gets me 100 points in CB and like +/- 2 fps in games. If you read up you'll see that I reset back to the Optimized Defaults + XMP just to get some baseline numbers....

I sold the 7700k guts to a friend before I sidegraded to Ryzen so there's no going back. I just didn't like the B450 ITX boards; they all felt lacking compared to my old Z270 ITX board and were more limited with m.2 slots so I went with a nicer x570 board. I can deal with it but will hold out hope that the next AM4 Ryzen CPUs are an upgrade.
 

gerardfraser

[H]ard|Gawd
Joined
Feb 23, 2009
Messages
1,113
Yeah I can see he picked up 400 points,does not mean he is correct or knows how to use the Ryzen CPU.
My 1067 FCLK has a higher score than 1933FCLK. I am not here to argue with you all I can say I know how it works for me.
 

trick0502

Supreme [H]ardness
Joined
Apr 17, 2006
Messages
5,435
As you can see on the post above you he picked up 400 points from setting the IF to 1900mhz....

You act like I haven't done any tweaks. I've done everything in this thread and more from 50 different PBO configs to all-core overclocks; this includes your settings on the first page. Tweaking PBO and using Ryzen Memory Calculator gets me 100 points in CB and like +/- 2 fps in games. If you read up you'll see that I reset back to the Optimized Defaults + XMP just to get some baseline numbers....

I sold the 7700k guts to a friend before I sidegraded to Ryzen so there's no going back. I just didn't like the B450 ITX boards; they all felt lacking compared to my old Z270 ITX board and were more limited with m.2 slots so I went with a nicer x570 board. I can deal with it but will hold out hope that the next AM4 Ryzen CPUs are an upgrade.

the biggest boost you’ll get in games will be higher clock speed. I was only hitting like 4.05-4.1 in games. Going up to 4.3 has helped a lot in 1% and .1% lows (at least in fortnite).

I would try to set your CPU to 4.3-4.4. I really didn’t change anything other than the multiplier and (maybe?) llc.

I have a friend with at 8700k and the exact same gpu. He gets the same frame rates as me in fortnite. His cpu is stock though.
 

trick0502

Supreme [H]ardness
Joined
Apr 17, 2006
Messages
5,435
Not trying to trick you,I have read the thread and posted way to much in thread.I was trying to help you.Your chasing and spending cash on nothing that does not make a difference in performance ,it is all just numbers.Stop ,return your X570 and get your money back.The 450 board is/was fine. You can always go back to your old Intel setup or upgrade Intel system.It appears the AMD system may be to much to deal with,it happens ,you have a what you think should be the performance and that was not met by the new AMD PC.

Memory has nothing to do with it.Here is ,oh god I hate cinebench but 2133Mhz is still same result.Even PC gaming with 2080ti at 2560x1440 with 2133Mhz Ram has nothing to do with your perceived under performance problems.

Check Ram speeds
View attachment 229294Memory-Chart by gerard fraser, on Flickr
View attachment 2292952133Mhz cinebench20

memory has a lot to do with gaming performance. look at your memory chart. there's a 10-20% performance increase in 1% lows going from 2133 to 3733. if you're only looking at average fps you're missing the point. also, that chart is 1440p, at 1080p the difference can be bigger.
 

gerardfraser

[H]ard|Gawd
Joined
Feb 23, 2009
Messages
1,113
Everyone is entitled to there opinion

AMD Ryzen Fabric Clock Vs Tuned Ram Battlefield 5 and Grand Theft Auto V 1920x1080

Tuned Ram AMD Ryzen Fabric Clock 1467Mhz (DDR4 2933Mhz) vs Fabric Clock 1933Mhz (DDR4 3866Mhz) Red Dead Redemption 2 and Sleeping dogs 1080p/1440p
 

kamikazi

Gawd
Joined
Jan 19, 2006
Messages
963
Everyone is entitled to there opinion

AMD Ryzen Fabric Clock Vs Tuned Ram Battlefield 5 and Grand Theft Auto V 1920x1080

Tuned Ram AMD Ryzen Fabric Clock 1467Mhz (DDR4 2933Mhz) vs Fabric Clock 1933Mhz (DDR4 3866Mhz) Red Dead Redemption 2 and Sleeping dogs 1080p/1440p


Still love to know how some people get above 1866 fclk. I've had two 3900x chips and one 3950x and have never gotten 1900 fclk. I've set the SOC voltage as high as I'm comfortable with at 1.11250 and it just doesn't work.
 

gerardfraser

[H]ard|Gawd
Joined
Feb 23, 2009
Messages
1,113
Still love to know how some people get above 1866 fclk. I've had two 3900x chips and one 3950x and have never gotten 1900 fclk. I've set the SOC voltage as high as I'm comfortable with at 1.11250 and it just doesn't work.
Well ,it just may be that you have 2 x CCD and one sucks .With 1 CCD much better chance to hit 1900Mhz FCLK /1933Mhz FCLK.That is the best I got,no problems though.If you tune your timings on Ram ,the FCLK does not make a real difference in performance.
 

kamikazi

Gawd
Joined
Jan 19, 2006
Messages
963
Well ,it just may be that you have 2 x CCD and one sucks .With 1 CCD much better chance to hit 1900Mhz FCLK /1933Mhz FCLK.That is the best I got,no problems though.If you tune your timings on Ram ,the FCLK does not make a real difference in performance.
Yeah, I quit chasing the fclk. I set it back down to 1800 and roll with that.
 
Top