• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

***official amd 83x0 overclocking thread***

farmerboy-86 said:
Is sabertooth on par with crosshair? And is it better vrm over my board?
Click to expand...

Both of them have much better VRMs than your board. They both can supply power stably to beyond what your CPU can handle without bellow ambient cooling. The Crosshair has more overclocking features and by the look of the components, has slightly better phase design (Formula-Z revision only) but both can easily handle anything even an ambitious overclocker with throw at it. The formula-Z revision's phase design is 8+2+2 as opposed tot he normal 8+2 we see on the Sabertooth and on the original Crosshair. This means an extra 2 phases dedicated just to memory for supposedly better tuning and to reach higher overclocks. While i am skeptical to how important it is, i did find that the fine tuned overclocks i ended up with on the Formula-Z would not even boot on the Sabertooth. Your board i think sports a 4 phase dedicated to the CPU and also a much more modest heatsink that you will need for high voltage 8320/8350 clocks. Those VRMs can become hot and make your overclock either unstable or throttle, if it was an Asrock board i would even mention that there would be a chance of fireworks.

I have used both your current board (the usb 3 revision), the sabertooth and i currently own two Crosshair formula-Z and have clocked a number of my own FX CPUs and a few friends. I very much doubt the 9590 and other higher end fx chips are speed binned from the results i have seen. Though you might not get the same clock on the same settings (solely a multiplier overclock with a 1.5V voltage), i am confident most if not all 8350s are capable of hitting 5GHz with a mixture of both fsb and multiplier clock and an increase in voltage. the 9590 has a default voltage of 1.5V, so its not all that surprising that it has that sort of frequency, it is essentially just a pre-overclocked chip with a lot of voltage added as an overhead for stability and a warning label telling you to accommodate for it with a decent cooler.

If you have trouble getting stable clocks after upping bus speed, up your nb voltage and knock your RAM down to the next multiplier or up the voltage on the RAM. Could be that instability of your overclock could be due to the RAM failing due to it being overclocked aswell. I have never tried using memtest to check if it is the problem of an overclock is the RAM (since its quicker to change settings and use trial and error) but i know others that swear by it when fine tuning a high clock.

Currently i own 3 fx CPUs two 8350's and one 8150. All three have hit 5GHz on water and achieved prime stable for at least an hour without workers shutting down. The 8150 is an absolutely golden chip i got lucky with and it can actually be under volted to 1.25 and the fsb pumped up to 240 (with some adjustment to nb voltage and playing with RAM timings to get them to agree). I did find the heatsink get a bit hot when trying to really push one of my 8350s, so i went with an EK VRM waterblock for CHFV-Z. Eventually after hours of trial and error, i managed to fine tune it to 5.3 before i hit an impossible wall. The other 8350 i gave up on just shy of 5.1GHz. With time, and a lot of trial and error, you can really build up some high clocks.
 
Last edited:
farmerboy-86 said:
Putting my gpu under water should help my ambient temps or not?
Click to expand...

MOving to water would be a good learning curve if your into your overclocking. It's not cheap investing in all the gear but you will learn a lot by doing so.

You see with the investment into the equipment will set you back enough to get a quality GPU, so if your CPU is stable just under 5Ghz I would upgrade graphics.

Tinker with water though if you want the next step into hardcore. :D
 
I have water on cpu with d5 vario so all I need for gpu is block and extra rad wouldn't cost to much just wondered if it will help temps as an overall?

Looks like from Avenged7fold post if I want 5+ I need crosshair board I may hold off on this and stic with 4.8 for now and invest in putting gpu under water.

I will post my gaming benchmark results tonight at 4.8 as you guys requested im keen to see results my self
 
farmerboy-86 said:
I
Looks like from Avenged7fold post if I want 5+ I need crosshair board I may hold off on this and stic with 4.8 for now and invest in putting gpu under water.
Click to expand...

I would go with GPU, upgrade board later some day (or you can sell on).

farmerboy-86 said:
I will post my gaming benchmark results tonight at 4.8 as you guys requested im keen to see results my self
Click to expand...

Sounds like a plan!
 
I currently run 278MHz bus clock on my 8350 with 1860MHz memory speed (naked venom and patriot 2133 sticks).

I have run up to 302Mhz bus with a 4832Mhz CPU and 2012Mhz memory clock on your board, the M5A99X evo which I thought a great board.

I also had no problems running cinebench, aida64 stress test, x264HD stress test, ARMA 2 benchmarks and several others at 5GHz on the M5A99X.

The crosshair formula V is a bit disappointing in comparison IMHO. It runs quite stably at 4.8GHz, it does not feel quite as comfortable at 5GHz as my previous board, an upgrade too far.

I use a phanteks air cooler
 
Last edited:
Th0nt said:
Which was?
Click to expand...

The M5A99X Evo rev 1.0.

Surprising I know, I suspect that the bios for that board is better than the piledriver bios for the crosshair. Using the same or very similar settings, I was experiencing restarts on the crosshair. I have found stable settings now but the M5A99x seemed to go much higher in bus clock.

Another learning curve, I did experiment with water cooling the CPU only with a single thick 240 rad and found that the temperatures were very comparable within a degree or two to the phanteks high end air. I was unfortunate to experience a pump failure and have reverted to air.
 
Last edited:
Very interesting reading this, will be clocking my 8320 tonight probably.

If I tweak the FSB, should I be reducing the memory, HT and NB dividers to bring them back down to standards levels? so that the only effect is on the CPU?
 
teppic said:
Are you using the older Crosshair Formula or the Formula Z?
Click to expand...

The older one, I have just started tweaking a couple of items again, notably the NB (ICH) (1.15V) and HT (1.25V) voltages and am currently running as below.

301 Mhz bus x 16 and 2408 NB and HT clocks with a 2000 MHz memory clock.

I am running an offset voltage of +0.116V giving 1.452V on load, 0.996V on idle at 2.1GHz (cool and quiet enabled).

This gives 8.23 rendering points on Cinebench 11.5. System is stable currently.
 
Yes as you fsb goes up so does everything else so bring them back to near as possible to standard for now untill you find max overclock for cpu. Then try taking cpu/nb mhz up may have to raise voktage on cpu/nb slightly. Its hard to determine what is causing instability so do one at a time
 
nkata said:
The M5A99X Evo rev 1.0.

Surprising I know, I suspect that the bios for that board is better than the piledriver bios for the crosshair. Using the same or very similar settings, I was experiencing restarts on the crosshair. I have found stable settings now but the M5A99x seemed to go much higher in bus clock.

Another learning curve, I did experiment with water cooling the CPU only with a single thick 240 rad and found that the temperatures were very comparable within a degree or two to the phanteks high end air. I was unfortunate to experience a pump failure and have reverted to air.
Click to expand...

I thought at first that water wasn't great compared to my old akasa venom medusa air cooler*but at these high clock I found my air cooler struggled where as the water hit a point where it didnt rise anymore adding gentle typhoons to my case and rad made a huge difference, also there are many factors as to why water may not seem as good as I found out very quickly.
first off decent fans a must (typhoons dropped couple degrees) air flow through case, heat generated from components, also adding fan to rear of mobo helped loads for me personally, rad position (some cases only have room in roof) I changed my loop position 3 times in last month roof was worst temps as heat rises simple physics lol
Bottom of case I dont think there was enough air being drawm in on my case im not saying all
And currently I run my rad in front giving me personslly best cooling.

So to recap as anything with overclocking its trial and error as we know. I think u shouldn't write water off so easy
 
farmerboy-86 said:
Yes as you fsb goes up so does everything else so bring them back to near as possible to standard for now untill you find max overclock for cpu. Then try taking cpu/nb mhz up may have to raise voktage on cpu/nb slightly. Its hard to determine what is causing instability so do one at a time
Click to expand...

Makes perfect sense. Thanks
 
Whilst I'm thinking about it, what are the benefits of adjusting both multiplier and FSB, as apposed to just the multiplier? I've read people have clocked higher, but why?
 
Th0nt said:
With just the multiplier increase you are only pushing the CPU harder. When you alter the FSB it pushes the rest of the system (RAM, NB, SB) which is where you would lock off or apply dividers should your BIOS allow for it.

Purchasing better quality memory and motherboard allows to push the FSB higher.
Click to expand...

Thanks
 
You must log in or register to reply here.
Back
Top Bottom