- last edited on by
ROGBot
Ok guys... for months, I have been attempting to achieve 5ghz stable and Ive come so close, yet so far.
I have gotten 4.816 @ 1.4875v 24/7 stable. It passes all my stress tests ie; Sony Vegas H.264 Rendering of 1hr video at 96k, OCCT, Prime95, BurnInTest, Linpack etc and all is great. Rock solid. Ive achieved this clock with the following settings.
Now Ive tried the go to method of 25x200mhz and gone as high as 1.55v and cannot achieve stability. I have tried total FSB overclocking, combo of FSB and Multi overclocking, ive tried the brute voltage approach (more on that in a minute). I have played with the VDD voltage, NB Voltage, RAM voltage, VID, etc.
Now, last night I was livestreaming my efforts for 5.0ghz and I had about 30 people present from several forums all trying to help me get 5.0 stable, but after 3.5 hours of crashing (usually about 30sec in to OCCT) and complete system lockups, I loaded my 4.8 again and called it a night.
I have a feeling that my memory is holding me back some how. I need some advice or to be pointed in the right direction of reseatching the coorolation of RAM timing and how to tune RAM for high overclocks.
Here are my specs.
CPU- FX8350
MB - Crosshair V Formua (non Z)
RAM - Patriot Viper XTreme 1866 (4x8gb) Ive tried 2 sticks instead of 4 with 0 change in stability
PSU - Corsair AX750 80Plus Gold
Cooling - Full custom loop 1x360mm rad 1x240mm rad
Now I DID achieve 5.0ghz stable on accident, but here is why I dont consider this a success. For some reason, my LLC allowed my 1.55v VCore setting to jump as high as 1.608v - Thats the first time Ive seen it move THAT far above the VCore set value. I was prettu frustrated and at that point didnt care if the chip fried to death so I let it go and to my amazment, passed 20min of OCCT before I got scared and shut it off. Temps maxed out at 61*c Socket 58*c core. Looks like my watercooling is doing its job.
This was all on live stream so of course, people were chanting for it to explode lol...
Anyway, Im at my wits end and reaching out to the experts. Im a youtuber with thousands of people asking me to make a 5ghz stable guide, but of course when I cant reach it myself, I cant even consider making the video... ugh.
I have gotten 4.816 @ 1.4875v 24/7 stable. It passes all my stress tests ie; Sony Vegas H.264 Rendering of 1hr video at 96k, OCCT, Prime95, BurnInTest, Linpack etc and all is great. Rock solid. Ive achieved this clock with the following settings.
Now Ive tried the go to method of 25x200mhz and gone as high as 1.55v and cannot achieve stability. I have tried total FSB overclocking, combo of FSB and Multi overclocking, ive tried the brute voltage approach (more on that in a minute). I have played with the VDD voltage, NB Voltage, RAM voltage, VID, etc.
Now, last night I was livestreaming my efforts for 5.0ghz and I had about 30 people present from several forums all trying to help me get 5.0 stable, but after 3.5 hours of crashing (usually about 30sec in to OCCT) and complete system lockups, I loaded my 4.8 again and called it a night.
I have a feeling that my memory is holding me back some how. I need some advice or to be pointed in the right direction of reseatching the coorolation of RAM timing and how to tune RAM for high overclocks.
Here are my specs.
CPU- FX8350
MB - Crosshair V Formua (non Z)
RAM - Patriot Viper XTreme 1866 (4x8gb) Ive tried 2 sticks instead of 4 with 0 change in stability
PSU - Corsair AX750 80Plus Gold
Cooling - Full custom loop 1x360mm rad 1x240mm rad
Now I DID achieve 5.0ghz stable on accident, but here is why I dont consider this a success. For some reason, my LLC allowed my 1.55v VCore setting to jump as high as 1.608v - Thats the first time Ive seen it move THAT far above the VCore set value. I was prettu frustrated and at that point didnt care if the chip fried to death so I let it go and to my amazment, passed 20min of OCCT before I got scared and shut it off. Temps maxed out at 61*c Socket 58*c core. Looks like my watercooling is doing its job.
This was all on live stream so of course, people were chanting for it to explode lol...
Anyway, Im at my wits end and reaching out to the experts. Im a youtuber with thousands of people asking me to make a 5ghz stable guide, but of course when I cant reach it myself, I cant even consider making the video... ugh.
hmm i'd say up the multi drop the nb ht to 2400 bring your bus speed down ..
mem cycle time is wrong should be 32
whats all the other settings at .. take a pic and post em ..
mem cycle time is wrong should be 32
whats all the other settings at .. take a pic and post em ..
Jaysvfx wrote:
Ok guys... for months, I have been attempting to achieve 5ghz stable and Ive come so close, yet so far.
I have gotten 4.816 @ 1.4875v 24/7 stable. It passes all my stress tests ie; Sony Vegas H.264 Rendering of 1hr video at 96k, OCCT, Prime95, BurnInTest, Linpack etc and all is great. Rock solid. Ive achieved this clock with the following settings.
Now Ive tried the go to method of 25x200mhz and gone as high as 1.55v and cannot achieve stability. I have tried total FSB overclocking, combo of FSB and Multi overclocking, I've tried the brute voltage approach (more on that in a minute). I have played with the VDD voltage, NB Voltage, RAM voltage, VID, etc.
Now, last night I was livestreaming my efforts for 5.0ghz and I had about 30 people present from several forums all trying to help me get 5.0 stable, but after 3.5 hours of crashing (usually about 30sec in to OCCT) and complete system lockups, I loaded my 4.8 again and called it a night.
I have a feeling that my memory is holding me back some how. I need some advice or to be pointed in the right direction of reseatching the coorolation of RAM timing and how to tune RAM for high overclocks.
Here are my specs.
CPU- FX8350
MB - Crosshair V Formua (non Z)
RAM - Patriot Viper XTreme 1866 (4x8gb) Ive tried 2 sticks instead of 4 with 0 change in stability
PSU - Corsair AX750 80Plus Gold
Cooling - Full custom loop 1x360mm rad 1x240mm rad
Now I DID achieve 5.0ghz stable on accident, but here is why I dont consider this a success. For some reason, my LLC allowed my 1.55v VCore setting to jump as high as 1.608v - Thats the first time Ive seen it move THAT far above the VCore set value. I was prettu frustrated and at that point didnt care if the chip fried to death so I let it go and to my amazment, passed 20min of OCCT before I got scared and shut it off. Temps maxed out at 61*c Socket 58*c core. Looks like my watercooling is doing its job.
This was all on live stream so of course, people were chanting for it to explode lol...
Anyway, Im at my wits end and reaching out to the experts. Im a youtuber with thousands of people asking me to make a 5ghz stable guide, but of course when I cant reach it myself, I cant even consider making the video... ugh.
It was a success. Some poorly binned or leaky 8350's like mine won't reach 5.0 GHZ without extra voltage. No you wouldn't want to run it 24/7 at 1.608 v. But for an overclock and stress test like OCCT or IBT -AVX it is fine to do so. Just don't go higher than that even for a temporary overclock. There is a wide variation in cpu VID's. So it's a bit of a craps shoot. You have very good cooling. 20 minutes of OCCT is good enough to know it is stable. Your ok.
