FX giving bad FPS? Here's how to tune your old FX to give better in game performance... - Page 23 - Overclock.net - An Overclocking Community

Forum Jump: 

FX giving bad FPS? Here's how to tune your old FX to give better in game performance...

Reply
 
Thread Tools
post #221 of 612 (permalink) Old 12-01-2017, 12:02 PM
New to Overclock.net
 
radisto's Avatar
 
Join Date: Dec 2017
Posts: 1
Rep: 1 (Unique: 1)
I read every page and tried this overclock.
I have
ASRock 970 Pro3 R2.0
2 x 4GB DDR3 1600MHz 10-10-10-30 HyperX Fury
AMD FX-6300
GTX 1060 3GB Zotac
650W EVGA B3
240GB SSD Team Group
be quiet! Pure Rock
My old OC (only tested game stable)
4.6GHz 1.3625V CPU (23x 200MHz auto FSB, it crashes even in games on 1.35V)
1866MHz 10-11-11-32 Auto Voltage (1.585V it can't go lower timings)
CPU NB and HT Link were on auto (2000MHz)
My new OC (stress tested in several programs, I'm pretty sure I can go higher FSB but this is what I tested and at this point I don't care about 20 or 40 MHz more)
FSB 235 MHz (even 240MHz turns on with 4680MHz and 1.35V with almost 2000MHz RAM and crazy high HT Link and CPU SB, there's no way to run 2.7MHz on 1.35V with the default 200MHz FSB so I'm glad I found this thread to realize my OC was ****)
PCIE 107 MHz (I read somewhere it's the best frequency not sure though)
4582 MHz 1.35V CPU (19.5 x 235 I haven't tried lower voltage since this is lower than old OC and more stable)
2585 MHz 1.275V CPU NB
2585 MHz auto voltage (1.215V) HT Link
1880MHz 10-11-11-32 Auto Voltage (1.585V)
1.2V NB (default is 1.1V I read that it can help with stability)
Now the benchmarks
_______________________4582 MHz CPU_______4.6 GHz CPU
_______________________1880MHz RAM_______1866MHz RAM
_______________________2585 MHz NB________2000MHz NB
_______________________2585 MHz HT Link____2000MHz HT Link
Geekbench
Single-Core Score_________2868_______________2830
-Memory Latency_________6862 / 63.1 ns________6557 / 66.0 ns
-Memory Bandwidth_______2037 / 10.9 GB/sec____1877 / 10.0 GB/sec
Multi-Core Score_________8205 _______________7909
-Memory Latency________6691 / 64.7 ns ________6305 / 68.7 ns
-Memory Bandwidth______2785 / 14.9 GB/sec ____2543 / 13.6 GB/sec
OpenGL Score___________120742______________122596 (why is this higher?)
AIDA64
CPU Queen _____________30737_______________30874
Memory read____________25902 MB/s__________24960 MB/s
Memory write___________17351 MB/s___________14710 MB/s
Memory copy___________21046 MB/s___________20414 MB/s
Memory latency_________62.2 ns_______________65.9 ns
CPU-Z
Single Thread___________266_________________267
Multi Thread____________1356________________1354
7-Zip (stopped on 2 min)
Compressing___________15958 KB/s___________15878 KB/s
Decompressing_________240586 KB/s__________237855 KB/s
Total Rating____________19815 MIPS__________19648 MIPS
Cinebench
CPU __________________542 cb_______________526 cb
OpenGL_______________86.05fps_____________82.05 fps
radisto is offline  
Sponsored Links
Advertisement
 
post #222 of 612 (permalink) Old 12-01-2017, 01:01 PM - Thread Starter
Useful idiot
 
Minotaurtoo's Avatar
 
Join Date: Jun 2013
Location: Florence, AL
Posts: 2,365
Rep: 80 (Unique: 46)
Quote:
Originally Posted by radisto View Post

I read every page and tried this overclock.
I have
ASRock 970 Pro3 R2.0
2 x 4GB DDR3 1600MHz 10-10-10-30 HyperX Fury
AMD FX-6300
GTX 1060 3GB Zotac
650W EVGA B3
240GB SSD Team Group
be quiet! Pure Rock
My old OC (only tested game stable)
4.6GHz 1.3625V CPU (23x 200MHz auto FSB, it crashes even in games on 1.35V)
1866MHz 10-11-11-32 Auto Voltage (1.585V it can't go lower timings)
CPU NB and HT Link were on auto (2000MHz)
My new OC (stress tested in several programs, I'm pretty sure I can go higher FSB but this is what I tested and at this point I don't care about 20 or 40 MHz more)
FSB 235 MHz (even 240MHz turns on with 4680MHz and 1.35V with almost 2000MHz RAM and crazy high HT Link and CPU SB, there's no way to run 2.7MHz on 1.35V with the default 200MHz FSB so I'm glad I found this thread to realize my OC was ****)
PCIE 107 MHz (I read somewhere it's the best frequency not sure though)
4582 MHz 1.35V CPU (19.5 x 235 I haven't tried lower voltage since this is lower than old OC and more stable)
2585 MHz 1.275V CPU NB
2585 MHz auto voltage (1.215V) HT Link
1880MHz 10-11-11-32 Auto Voltage (1.585V)
1.2V NB (default is 1.1V I read that it can help with stability)
Now the benchmarks
_______________________4582 MHz CPU_______4.6 GHz CPU
_______________________1880MHz RAM_______1866MHz RAM
_______________________2585 MHz NB________2000MHz NB
_______________________2585 MHz HT Link____2000MHz HT Link
Geekbench
Single-Core Score_________2868_______________2830
-Memory Latency_________6862 / 63.1 ns________6557 / 66.0 ns
-Memory Bandwidth_______2037 / 10.9 GB/sec____1877 / 10.0 GB/sec
Multi-Core Score_________8205 _______________7909
-Memory Latency________6691 / 64.7 ns ________6305 / 68.7 ns
-Memory Bandwidth______2785 / 14.9 GB/sec ____2543 / 13.6 GB/sec
OpenGL Score___________120742______________122596 (why is this higher?)
AIDA64
CPU Queen _____________30737_______________30874
Memory read____________25902 MB/s__________24960 MB/s
Memory write___________17351 MB/s___________14710 MB/s
Memory copy___________21046 MB/s___________20414 MB/s
Memory latency_________62.2 ns_______________65.9 ns
CPU-Z
Single Thread___________266_________________267
Multi Thread____________1356________________1354
7-Zip (stopped on 2 min)
Compressing___________15958 KB/s___________15878 KB/s
Decompressing_________240586 KB/s__________237855 KB/s
Total Rating____________19815 MIPS__________19648 MIPS
Cinebench
CPU __________________542 cb_______________526 cb
OpenGL_______________86.05fps_____________82.05 fps

Glad it helped you : ) That one that was higher on the previous OC was likely dependent on the core speed more... I've noticed that in some benches the cpu/nb OC doesn't help much, but in others it makes a huge difference... the ones where memory are involved usually are helped the most...

One thing I'm glad that you posted this for is that it confirms that even with slower RAM than mine it still helps... I've noticed a larger difference than most I think due to my 2400mhz ram.... but I suspect that even down to 1600mhz it would help and since yours is at 1880mhz it shows it likely will...

Intel who?


Minotaurtoo is offline  
post #223 of 612 (permalink) Old 12-05-2017, 12:48 PM
Oldtimer
 
gapottberg's Avatar
 
Join Date: Dec 2006
Posts: 794
Rep: 54 (Unique: 38)
Alright guys, played around with Stock settings Vs. OCed settings on my rig today, specifically looking at RAM, CPU/NB, and HT speeds. Some very interesting results.

Here is the run down of settings I compared and what I found.



FX-8320e
Set to stock FX-8350/8370 spec w/turbo disabled, a favorite use case scenario for TechTubers benchmarking of FX.
CPU 4.0Ghz, 1.30v

ASrock 970 Fatal1ty
CPU/NB 2200mhz, 1.2v
HT 2000mhz, 1.2v
NB 1.2v
16GB DDR3 1600mhz, C11, 1T, Duel channel, 1.65v

MSI R9-390X
GPU 1000mhz (-50mv, +50% power limit)
GDDR5 1500mhz
Crimson 17.7

I ran a total of 5 benchmarks and then ran the exact same set up but with 3 changes.
  • Ram + 2133mhz (with same timings)
  • CPU/NB + 2600mhz, 1.40v
  • HT + 2400mhz, 1.20v

The benchmarks I ran were the CPUZ bench and CBR15, as well as TombRaider 2013, Dirt Rally, and Ashes of the Singularity DX12
Games were run at 720p with details on Ultra except for AoS which used Standard detail preset. The main idea behind this was to show a scenario where the load was mostly shifted to the CPU, and where these subsytems might have the most impact. The actual numbers are not important as much as the differences between the two tests, as they do not reflect actual ingame perfomance for real use scenarios, but what I did find was very informative.

Both the CPU Synthetics showed no differences in the test results. Had me discouraged from even moving forward, but I decided to do a few games just to see. I really need to add Aida64 to my test suite soon I think. It would be most helpful in seeimg memory improvements my current synthetics miss. Anyways, on to the gaming benchmark results.

Warning: Spoiler! (Click to show)
Tombraider 2013 showed a +3% increase in perfomance across min/avg/maxs. Nothing too exciting so far.

Warning: Spoiler! (Click to show)
Ashes of the Singularity showed a more impressive +8-10% increase in avgs across all batch types in the GPU test.

Warning: Spoiler! (Click to show)
Dirt Rally canned benchmark was the eye opener. +39% min / +32% avg / +9% max


Clearly there are some workloads and games that benifit more than others from the improved CPU/NB and L3 cache speeds, increased HyperTransport speeds, and better Ram speed and bandwidth. Tasks that bottleneck one of these three systems can greatly benfit from some fine tuning and Overclocking, and can greatly improve FX overall experince and perfomance by unbelievable margins at times.

So much so that I am now extreamly suspicious of any benchmarking data I see on YouTube or elswhere that looks perticularly dreadful. Most of the time they either inentionally or ignorantly handicap the FX by not tuning these very important subsytems.

#REDTEAM
Serenity
(11 items)
Dat_Ashe
(12 items)
CPU
AMD FX-8320E OCed @ 4.5Ghz
Motherboard
ASRock 970 Fatality
GPU
MSI 390X
RAM
Gskill Ripjaw 16GB @ 2133MHz
Hard Drive
Samsung EVO 250GB SSD
Optical Drive
ASUS Blue Ray
Power Supply
Thermaltake toughpower 650watt
Cooling
Thermaltake Water 3.0
Case
Thermaltake T81
Operating System
Windows 10
Monitor
Asus VX248
CPU
FX-6300
Motherboard
BIOSTAR TA970
GPU
SAPPHIRE Radeon R9 270 2GB
RAM
G.SKILL Ripjaws X Series 8GB DDR3 1866mhz
Hard Drive
Seagate Barracuda 1TB Hard Drive
Optical Drive
ASUS Blu-ray Drive
Power Supply
Corsair 750watt PSU
Cooling
ZALMAN CNPS9900MAX-B 135mm CPU Cooler
Case
Antec GX500
Operating System
Windows 10 Technical Preview
Monitor
ASUS VX248 24" 1080p LCD Moniter
Mouse
Logitech G602 wireless gaming mouse
▲ hide details ▲


gapottberg is offline  
Sponsored Links
Advertisement
 
post #224 of 612 (permalink) Old 12-05-2017, 01:14 PM
New to Overclock.net
 
jaredismee's Avatar
 
Join Date: Mar 2016
Posts: 363
Rep: 12 (Unique: 9)
well since this thread got bumped i figured i would add this update to my OC here and the before/after aida 64 results too.

difference in clock is almost nothing, overall changes were a bump up of 9 to the fsb and -.5 to the cpu multiplier.

but even the small change of 2240MHz - 2312MHz ram and 2520 to 2600MHz NB with almost no change to cpu changed results in aida 64, and i think was interesting enough to include here.
Warning: Spoiler! (Click to show)
NEW


OLD

edit: also the post above shows pretty interesting results, and i also have trouble trusting any results of benchmarking these chips on youtube without seeing clearly the specs they are being run at in a lot more detail than just cpu clocks.
jaredismee is offline  
post #225 of 612 (permalink) Old 12-05-2017, 03:31 PM - Thread Starter
Useful idiot
 
Minotaurtoo's Avatar
 
Join Date: Jun 2013
Location: Florence, AL
Posts: 2,365
Rep: 80 (Unique: 46)
Quote:
Originally Posted by gapottberg View Post

Alright guys, played around with Stock settings Vs. OCed settings on my rig today, specifically looking at RAM, CPU/NB, and HT speeds. Some very interesting results.

Here is the run down of settings I compared and what I found.



FX-8320e
Set to stock FX-8350/8370 spec w/turbo disabled, a favorite use case scenario for TechTubers benchmarking of FX.
CPU 4.0Ghz, 1.30v

ASrock 970 Fatal1ty
CPU/NB 2200mhz, 1.2v
HT 2000mhz, 1.2v
NB 1.2v
16GB DDR3 1600mhz, C11, 1T, Duel channel, 1.65v

MSI R9-390X
GPU 1000mhz (-50mv, +50% power limit)
GDDR5 1500mhz
Crimson 17.7

I ran a total of 5 benchmarks and then ran the exact same set up but with 3 changes.
  • Ram + 2133mhz (with same timings)
  • CPU/NB + 2600mhz, 1.40v
  • HT + 2400mhz, 1.20v

The benchmarks I ran were the CPUZ bench and CBR15, as well as TombRaider 2013, Dirt Rally, and Ashes of the Singularity DX12
Games were run at 720p with details on Ultra except for AoS which used Standard detail preset. The main idea behind this was to show a scenario where the load was mostly shifted to the CPU, and where these subsytems might have the most impact. The actual numbers are not important as much as the differences between the two tests, as they do not reflect actual ingame perfomance for real use scenarios, but what I did find was very informative.

Both the CPU Synthetics showed no differences in the test results. Had me discouraged from even moving forward, but I decided to do a few games just to see. I really need to add Aida64 to my test suite soon I think. It would be most helpful in seeimg memory improvements my current synthetics miss. Anyways, on to the gaming benchmark results.

Warning: Spoiler! (Click to show)
Tombraider 2013 showed a +3% increase in perfomance across min/avg/maxs. Nothing too exciting so far.

Warning: Spoiler! (Click to show)
Ashes of the Singularity showed a more impressive +8-10% increase in avgs across all batch types in the GPU test.

Warning: Spoiler! (Click to show)
Dirt Rally canned benchmark was the eye opener. +39% min / +32% avg / +9% max


Clearly there are some workloads and games that benifit more than others from the improved CPU/NB and L3 cache speeds, increased HyperTransport speeds, and better Ram speed and bandwidth. Tasks that bottleneck one of these three systems can greatly benfit from some fine tuning and Overclocking, and can greatly improve FX overall experince and perfomance by unbelievable margins at times.

So much so that I am now extreamly suspicious of any benchmarking data I see on YouTube or elswhere that looks perticularly dreadful. Most of the time they either inentionally or ignorantly handicap the FX by not tuning these very important subsytems.

I linked this in the OP... very good research, thanks, repped up

Intel who?


Minotaurtoo is offline  
post #226 of 612 (permalink) Old 12-08-2017, 10:51 AM
New to Overclock.net
 
umeng2002's Avatar
 
Join Date: Jul 2010
Location: Florida
Posts: 3,083
Rep: 167 (Unique: 105)
Can't get my CPU/NB stable at 2600 MHz. I don't want to set it higher than 1.32v

CPU
AMD Ryzen 2700X
Motherboard
Asus Prime X470-Pro
GPU
EVGA GeForce RTX 2070 XC Ultra
RAM
TeamGroup T-Force 16 GB (2x8) Pro Dark (B-die TDPGD416G3200HC14ADC01)
Hard Drive
Samsung 840 EVO 250 GB
Power Supply
Seasonic Focus Plus Platinum SSR-750PX
Cooling
Corsair H80i (not V2 or GT)
Monitor
LG 34UC80-B
Keyboard
Logitech G413
Mouse
Logitech G503 RGB
Audio
Creative SoundBlaster Z (OEM)
▲ hide details ▲
umeng2002 is offline  
post #227 of 612 (permalink) Old 12-08-2017, 11:05 AM
New to Overclock.net
 
jclafi's Avatar
 
Join Date: Sep 2008
Location: Earth
Posts: 788
Rep: 55 (Unique: 37)
Do you have active cooling in the N.B ? If yes you can run it @ 1.4v w/ no problem.

Do a test boot the machine with 1.4v and run some games. While gaming touch the dissipator and see if is too hot.

thumb.gif
Quote:
Originally Posted by umeng2002 View Post

Can't get my CPU/NB stable at 2600 MHz. I don't want to set it higher than 1.32v

R5 2600 - MSI X470 Gaming Pro - nVIDIA GTX1060 6GB O.C - Corsair Vengeance LPX 16GB @ 3200 - Cooler Master Storm Scout - Western Digital Green 240GB M.2 - Kyngston A400 480GB - Seagate Barracuda 1Tb - Cooler Master Hyper 212 Turbo - SamSung 32in LED - eVGA 750W - Razer Abyssus V2 - Logitech G27
jclafi is offline  
post #228 of 612 (permalink) Old 12-08-2017, 11:08 AM
New to Overclock.net
 
umeng2002's Avatar
 
Join Date: Jul 2010
Location: Florida
Posts: 3,083
Rep: 167 (Unique: 105)
Quote:
Originally Posted by jclafi View Post

Do you have active cooling in the N.B ? If yes you can run it @ 1.4v w/ no problem.

Do a test boot the machine with 1.4v and run some games. While gaming touch the dissipator and see if is too hot.

thumb.gif

Isn't the CPU/NB voltage for the memory controller mainly? My CPU temps go up a few degrees when it set it over 1.3v

First Windows didn't boot. Then I upped the voltage, but IBT AVX won't get past 2 runs.

CPU/NB at 2400 MHz is stable even at 1.2v

With the Sabertooth, the 990FX has a sink and heat-pipe, but I don't have a fan right on it.

CPU
AMD Ryzen 2700X
Motherboard
Asus Prime X470-Pro
GPU
EVGA GeForce RTX 2070 XC Ultra
RAM
TeamGroup T-Force 16 GB (2x8) Pro Dark (B-die TDPGD416G3200HC14ADC01)
Hard Drive
Samsung 840 EVO 250 GB
Power Supply
Seasonic Focus Plus Platinum SSR-750PX
Cooling
Corsair H80i (not V2 or GT)
Monitor
LG 34UC80-B
Keyboard
Logitech G413
Mouse
Logitech G503 RGB
Audio
Creative SoundBlaster Z (OEM)
▲ hide details ▲
umeng2002 is offline  
post #229 of 612 (permalink) Old 12-08-2017, 11:26 AM
Oldtimer
 
gapottberg's Avatar
 
Join Date: Dec 2006
Posts: 794
Rep: 54 (Unique: 38)
Be very careful when dealing with CPU/NB...it is not the same thing as your NB chipset.

The CPU/NB is on your CPU die...and mainly affects the memory controler and the L3 chache speed. It has its own independant voltage delivery that is seperate from the CPU and from the actual NB chipset.

The NB chipset is on your motherboard and usually is near the CPU socket covered by a heatsink. It has its own voltage control so do not confuse them when tweaking voltages.

Typically stock voltage on the actual NB vhipset is fine with very small bumps sometimes needed at extream overclocks. I have found increasing voltages can very quickly increase the temp of the heatsink to untouchable, so be careful and never bump this voltage up to 1.40v...you should never need it that high and it could damage your board if left too long in that state.

The CPU/NB does however require voltages around or above 1.40v to hit higher clock speeds. Increasing the clock speed and voltages on this will increase your CPU temps due to it being on the CPU die itself. How far you can go is directly deoendant on your CPU cooling solution, your power delivery, and the quality of the memory/cpu/mobo.

As has been stated several times. Getting both the CPU/NB and HT speeds up to 2600mhz or higher and keeping them close to the same speed seems to yield the most stability and best performance.

#REDTEAM
Serenity
(11 items)
Dat_Ashe
(12 items)
CPU
AMD FX-8320E OCed @ 4.5Ghz
Motherboard
ASRock 970 Fatality
GPU
MSI 390X
RAM
Gskill Ripjaw 16GB @ 2133MHz
Hard Drive
Samsung EVO 250GB SSD
Optical Drive
ASUS Blue Ray
Power Supply
Thermaltake toughpower 650watt
Cooling
Thermaltake Water 3.0
Case
Thermaltake T81
Operating System
Windows 10
Monitor
Asus VX248
CPU
FX-6300
Motherboard
BIOSTAR TA970
GPU
SAPPHIRE Radeon R9 270 2GB
RAM
G.SKILL Ripjaws X Series 8GB DDR3 1866mhz
Hard Drive
Seagate Barracuda 1TB Hard Drive
Optical Drive
ASUS Blu-ray Drive
Power Supply
Corsair 750watt PSU
Cooling
ZALMAN CNPS9900MAX-B 135mm CPU Cooler
Case
Antec GX500
Operating System
Windows 10 Technical Preview
Monitor
ASUS VX248 24" 1080p LCD Moniter
Mouse
Logitech G602 wireless gaming mouse
▲ hide details ▲


gapottberg is offline  
post #230 of 612 (permalink) Old 12-08-2017, 12:45 PM - Thread Starter
Useful idiot
 
Minotaurtoo's Avatar
 
Join Date: Jun 2013
Location: Florence, AL
Posts: 2,365
Rep: 80 (Unique: 46)
Quote:
Originally Posted by umeng2002 View Post

Isn't the CPU/NB voltage for the memory controller mainly? My CPU temps go up a few degrees when it set it over 1.3v

First Windows didn't boot. Then I upped the voltage, but IBT AVX won't get past 2 runs.

CPU/NB at 2400 MHz is stable even at 1.2v

With the Sabertooth, the 990FX has a sink and heat-pipe, but I don't have a fan right on it.

cpu/nb can be finicky... mine has a voltage wall at 2700 and it takes 1.4 to get the, but 1.25 will get me close to 2600..... actually I think it's exactly 1.25 for 2600... but I run mine a bit higher than that usually around 2650 ish with 1.3v... I have found that getting 2600 stable is worth sacrificing some core clocks as I can get 4.7ghz to do better in many games than 5ghz with 2400mhz... with my cooling all I sacrificed actually was 5.1ghz... to 5ghz but my in game performance was actually better than the 5.1ghz

Intel who?


Minotaurtoo is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off