How many BIOS profiles can you save with that board RGone? That is one thing I have been spoiled by on my 2 Asus boards, my Asrock Extreme 4 only has 3 .
I have my daily and my Video Edit profiles saved. The only thing that changes from one to the other is the Multiplier and the Vcore since all else remains the same to move from daily 4.2Ghz to Video Edit 4.8Ghz.
RGone...
So what is the deal with the Asrock Fatal1ty 990FX Professional so far in my my mind? Well it certainly does not make me wish my CHV was back on the bench at this time.
Once I got the knowledge in my head that changing the FSB would not be reflected by updating of the location where CPU Mhz was shown, well I don't change that very much anyway. One maybe 3 FSB at one time to vary my CPU Speed to Video Edit, but now that I have my normal and my Video Editting speeds saved to the Bios Profiles...well even that is not an issue for me. You get used to doing one thing and it takes breaking the habit. Humans are creatures of habit so they resist change with a passion in general.
I have rather settled into using the board to Edit all my SShots and captures and such has proceeded without the first hitch. Changing speeds up and down for the heck of it. No problem either since I am not running those speeds that require such ridiculous Vcore on the FX-6300.
My next move is to put my FX-8350 on this board and see if the CPU performs equally well as it did on my CHV. I expect that it will do so. That really says the Fatal1ty board is very nice for sure. Less pricey as well. On to the FX-8350 shortly.
RGone...
Rgone you tested the 6300 on the Asrock board only correct? Is it possible that the 6300 required more v core because of the board not the chip? Also can you explain to me how to do the show/hide button with pictures I cannot find out how to searching on the forum.
"Mandrake" man I doubt the board has anything to do with it since the FX-6300 just did not want to GO when I was testing it on my CHV and "knowing" what C_D meant when he said I was babying my VRMs on the CHV to put the FX-6300 on the FatLady he had sent with the cpu...well the rest is history of the 5.5Ghz cpu validation and everything else.
I know how we think as humans and I did have a brain wrestling match about aw crap man you are testing on a different board until I just set down and looked at the Vcore I was willing to use when the FX-6300 was on my CHV and I did nOt give it enough Vcore to run beyond about 5.1Ghz on the CHV since I did not use more than 1.6125V on the CHV and that was n0t enough to get on up there. I guess that is why good records keeping is such a must.
I was just thinking out loud, since the board was the only component that was different was such a difference caused by it. As we well know by now every chip, board, etc... is different so I'm sure there are more factors that come into play. Part of it being that chip feels all warm and toasty inside when it gets fed More V Core thanks to it's owner. Maybe it needed more V Core because you weren't feeding it everclear?
It was always a question mark for me about Vcore and Everclear. I suspected that you could not have a lot of the one without some of the other. Hehehe.
...Okay the FX-6300 is getting too homesick. She is getting restless and wants to go home. So I finally got around to doing some stressing with her and as I have said before she likes that Vcore so I give it to her. Otherwise she refuses stress test stability. Given her voltage she just stresses along.
I am still impressed with the Asrock Fatal1ty 990FX Professional motherboard. It hands out the voltage to this hungry cpu without a whimper and does not have Vcore bouncing all over the place. Load the processor and you can see the Vcore raise slightly and just go and continue to do the stress test. Got the LLC 'workings' sorted in my head after the CHV and although different the results are just as good as I see in my tests.
There maybe features like setting bios from laptop on the CHV that could appeal to LN2 users and the super serious overclockers. F-12 captures the bios settings to Usb stick just like CHV which is neat for either motherboard. There are only three saveable profiles on the Fatlady but I only use two anyway, so it is not a bother to me.
Everything said, the Fatal1ty is a board for use with any of the FX processors be they the BD or the PD types and can push the processor as hard as you can supply cooling and do so without VRM throttling and all that other crap associated with lesser boards and FX processors.
Next up is my own FX-8350 and the FatLady board since I am far more acquainted with my FX-8350 than the borrowed FX-6300. I think the owner of the FX-6300 is wanting to know how the FatLady deals with my FX-8350 also. So it is up next.
Okay just a screen capture to give a little indication of stressing out the CPU and Ram and Mobo.
She hummed right on thru once given enough cpu voltage.
RGone...
Wow 1.5 Cpu V core for 4.6 does seem like a lot, though I do needs to get me hands on that special cooling system you have there Gonster. 46 and 41c seem like pretty good temps for that V Core.
Yeah, I have been noticing that benching and stress test stable are not nearly the same thing. That is why we have users sail in the CPU forum all the time thinking they are lovely until it comes time to stress test the system. Then everything is hay-wire. Temps are out the window as voltages have to rise to become even what is considered moderately stable.
Perhaps I could have dropped the Vcore a little more; except I was tired of playing with it to stress it. Dang heavy duty cold and sinus flare-up had kicked me for 10 days and I was just ready to stress it and move on.
If you remember C_D said put the FX-6300 on the FatLady so I could up the Vcore since he wanted to 'just see' the 5.5Ghz if at all possible short of LN2 or DICE. I gave it to him but it was with a ton of Vcore. I already knew this FX-6300 and to hear him tell it any of the three he had would all act the same. That is the only thing that makes me unsure of this wholesale stamping of the FX-6300 as the way to go for most users. I just don't know if they 'all' or most do need so much Vcore to run high mhz and also to stress test toward stable. I may try to find another FX-6300 to test with a little later.
What is reasonable anymore? I saw what you responded to that poster in the AMD Mobo section when he called a 4.6Ghz to 4.9Ghz as a moderate overclock of an 8 core Vishera I believe it was where he had just burned up his M5A97 Asus mobo pushing his cpu. Said he knew the mobo would not stand-up to pushing an 8-core and he was correct.
I mean that paragraph right that tells all about why it is we see so many newbs come into the AMD cpu section with unrealistic expectations. You know you don't have a mini-closed loop W/cooling system. I don't. Neither does C_D. Neither does "ssjwizard". Neither does "Bassnut". None of us that really push on even the new Vishera cpus; rely on air or mini-W/Cooling.
I mean I don't have to worry about how my cpu scales really. This FX-8350 I have seems to be a pretty good one. The one you have seems the same. Does C_D have a couple of FX-6300s that are not so stellar? I don't know. I 'originally' did not think my FX-8350 was much good since it seemed hard to clock fast. But we now know the Vishera's don't clock as fast but do more per clock than the Bulldozers did. So it all is relative. But the truth is just the truth.
And really that is all this whole thread has been about anyway. What was the worth of the Asrock Fatal1ty 990FX Professional when called on to really push a 6 core or 8 core cpu to high cpu speeds? Well the board passed with flying colors in my eyes. A cheaper but sturdy alternative to some of the higher priced AMD AM3+ motherboards.
Plus we might have had another thing made more crystal clear and that is the Vishera FX cpus don't like to clock as high and may take more Vcore than we have been used to really push the cpu and yet remain stress test stable. I cannot say that for sure but it is a thing I remember now when new users come into the AMD CPU section wanting to overlcock their stuff. Vishera cpus do more in the same time than a Bulldozer did, but they "still" run hot when pushed. Never losing sight of that paradigm is one thing I just learned. I won't forget it easily either. All I have done is see that paradigm played out for real, again.
RGone...
Yeah, I have been noticing that benching and stress test stable are not nearly the same thing.
Perhaps I could have dropped the Vcore a little more; except I was tired of playing with it to stress it.
I just don't know if they 'all' or most do need so much Vcore to run high mhz and also to stress test toward stable. I may try to find another FX-6300 to test with a little later.
What is reasonable anymore? I saw what you responded to that poster in the AMD Mobo section when he called a 4.6Ghz to 4.9Ghz as a moderate overclock of an 8 core Vishera I believe it was where he had just burned up his M5A97 Asus mobo pushing his cpu. Said he knew the mobo would not stand-up to pushing an 8-core and he was correct.
I mean that paragraph right that tells all about why it is we see so many newbs come into the AMD cpu section with unrealistic expectations. You know you don't have a mini-closed loop W/cooling system. I don't. Neither does C_D. Neither does "ssjwizard". Neither does "Bassnut". None of us that really push on even the new Vishera cpus; rely on air or mini-W/Cooling.
Ya these CPUs are happy to boot and even run some benches on fairly standard vCore but if you want them 100% stable those voltage requirements can spiral out of control pretty fast.
1.5V stress test stable on a 6300 TBH is not all that bad. I have one 8320 here that needs 1.62V at 4.62Ghz to be load stable just to give you an idea what the range of these CPUs is.
IMO 4.9Ghz is not a "moderate OC" for a vishera unless your using DICE/LN2. As far as im concerned 4.2-4.5Ghz is doable on ever Vishera and that is the range I would consider moderate/reasonable. Anytrhing past 4.5ish GHZ is unknown territory and should be considered a heavy OC for one of these chips.
Final thought THANK YOU RGONE for your testing of this motherboard. I almost bought one myself but went with the sabertooth. I see now that I was right in my assumptions of what its capabilities are and am certain that I would not be disappointment if I had gone that route.
What is reasonable anymore? I saw what you responded to that poster in the AMD Mobo section when he called a 4.6Ghz to 4.9Ghz as a moderate overclock of an 8 core Vishera I believe it was where he had just burned up his M5A97 Asus mobo pushing his cpu. Said he knew the mobo would not stand-up to pushing an 8-core and he was correct.
I was actually laughing when I was reading this persons post as I was answering it. He's running a Bulldozer, not a Piledriver which makes it even more funny. I guess it's the way of the world these days. People hear someone doing "x" and they think that all they need to do is get that product and they can do "x" also.
As far as all your testing Gonster it does make interesting reads and I have learned a lot from them.
As far as the V Core goes I think they so called binned the 8350's very well. I compared my processor with yours in the scaling tests you did and mine takes the same V Core give or take 1 notch all the way up to 5.0. I haven't really put any effort into getting it stable at 5.2, due to lack of need but I'm sure I will one of these days, for kicks. I use 1.45 for 24/7 at 22.5*210 roughly 4.7. Also having the ram over @ 1960 is nice. I have also found that I start having instability above 2000mhz , though this might me due to my lack over ram overclocking skills.
Drake take the Read to Read Delay and set 1 setting larger than what the board auto sets itself to. That might give you just a little more on the ram. I think my board was setting Read to Read Delay to 6 and I set to 7 and when I did that my Ares DDR1600 in 'with' my DDR1866 would both then do DDR1900 easily for a total of 16gigs ram from two greatly dissimilar sets of ram.
I use only the G Skil, Ares series ram for the non-obtrusive heat spreaders and my DDR1866 will do DDR2040 pretty handily.
I have the GSkill Snipers 1866,9-10-9-28 they have similar heat spreaders. I'll try that thanks, it's those secondary and tertiary one's that I'm still trying to figure out.
...just got my own FX-8350 mounted in the socket of this loaner Asrock Fatal1ty 990FX Professional motherboard. Guessed at a Vcore since I could not see my own FX-8350 Vcore pics since they were on the un-booted puter. Hehehe. Sure do like the lesser Vcore the FX-8350 seems to like. Straight to 4.6Ghz as there is no reason to dork around with lesser speeds. 4.4Ghz and 4.8Ghz are the two numbers I need to run for daily use and then for Video Editting. So nailed it right in between. Not bad for a guess at Vcore. Testing to follow.
RGone...
Man oh man am I liking this FX-8350 and the Fatal1ty board. Been awhile since I ran the benches on my FX-8350 and the CHV and I am not looking to see the score comparison yets but this rig is just about gaining 0.2 points for every 100Mhz increase in speed of the cpu in the Cinebench cpu test. I am thinking that is pretty good.
Had to increase the CPU_NB to 1.325V in order to run the high CPU_NB speed and in lock step with the HT Speed but it is rokken along pretty darn spiffy. Got to see if the Fatal1ty will bench right on up to the 5.4Ghz I did on the CHV. Be a day or two before I get all the benches run and wish I had my new video card since I RMA'd that wimp 6850. I guess if wishes were horses then beggars would ride. Hehehe.
RGone...
EDIT:
Holy Bat Shettman this is freeking me out. It looks like the FX-8350 is going to run as fast on less Vcore on this board than on my CHV. That is reading from CPUz and HWMonitor. That will be a swift kick in the head. Hehehe. Still got more benching by a lot to do but had to take a quick look when I got to 5.0Ghz since I had written down the Vcores for the FX-8350 so I would not get caught with the voltages on an unbooted computer. Looked and I had 1.5Vcore for the FX-8350 on the CHV and am only at 1.48 on the Fatal1ty. Oh my gootness. Now to see what they all show.
RGone...Again.
END EDIT.
While I am thinking about it let me make note of an observation that I have after putting this Fatal1ty thru it's paces now with 2 different cpus. One an FX-6300 on loan with the mobo and my own FX-8350 that I am more familiar with after having had it in my CHV for maybe 2 months.
1. CPU_LLC does not respond the same on this Fatal1ty as it does on my CHV.
2. CPU_LLC on the Fatal1ty board, for lack of a better word, "clamps" the Vcore downward so that the VRM spikes do not stand a chance of damaging the CPU. Remember what I just said.
3. CPU_LLC on the Fatal1ty board has 6 options.
1. Disabled which I have had nO need to try.
2. Auto
3. 100%
4. 75%
5. 50%
6. 25%
4. So remembering #2 above 25% is less down-clamp of Vcore and the CPU gets more voltage.
5. When not pushing the CPU very hard I was generally running at 50% CPU_LLC since the down-clamping action would keep the Vcore lower overall.
6. When running at 4.6Ghz to 5.0Ghz, I dropped to 25% CPU_LLC so more voltage and current could be passed thru to the cpu and with it were a few higher voltages due to the VRM dialing up to supply current as the cpu did work.
7. When running above 5.0Ghz I set the CPU_LLC to AUTO and that worked to allow the cpu to be able to get enough voltage and power to run the big numbers.
8. You can look at the screen capture below and see what CPU_LLC at AUTO did with a bios Vcore setting of 1.500Vcore. CPU voltage dropped to 1.480V when I fired LinX up and then jumped to 1.52V and after that watching HWMonitor which was open on the desktop while running LinX, the VCore stayed within 0.05 of the 1.496V you see in the screen shot. Pretty good working CPU_LLC if you stop to think about it.
9. Actually unlearning CHV CPU_LLC and knowing that Asrock did their LLC circuit different and pretty good on the Fatal1ty motherboard was the only thing I had to pay any special attention to when using the Fatal1ty motherboard. For $160.00 Usd over my CHV at $230 Usd, this is certainly a keeper motherboard except it is on loan. Hehehe. Oh you have ROG connect and the like on the CHV and if you need it then spend the money and get it, otherwise this is one heck of a motherboard and quite capable of handling 8 core FX-type processors.
10. I have not looked at all the captures to check used Vcore yet since I have only text editted one screen capture, but I think in the 4.2 to 4.6Ghz range during the benching, the Fatal1ty motherboard actually benched with slightly less Vcore than my CHV. On the order of like ~0.03Volts. Not much but might be worth it to someone with less cooling than I have.
11. One other thing I did to keep the playing field as level as I could between the Fatal1ty and my CHV, besides the same ram, power supply and finally the same CPU was to keep the CPU_NB and the HT Frequency up high and run at the same speed one to the other. UP until I reached about 5.0Ghz the CPU_NB voltage was in the same range as my CHV at roughly 1.3V. After 5.0Ghz I did have to up the CPU_NB to 1.325V to continue my use of the high CPU_NB along with the HT Frequency at the same speed. I refused to drop CPU_NB so up with the CPU_NB voltage a little more and she stayed right there and giving all the performance one can get from running high CPU_NB.
No the owner of this mobo never ran the board like #11 above, at least not that I am aware of but I did on CHV and fully intended that I could on the Fatal1ty and could and did.
Okay the screen capture of LinX and some CPUz's and the HWMonitor. Until I get the other captures sorted.
RGone...
Not quite 2 hours LinX since the extra Gflops fooled me on how long to run 25 passes.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
.
Gonster it does make interesting reads and I have learned a lot from them. 
