ASUS M5A97 R2.0 FSB max @ 300 Mhz?

[PAPADOC]

Hello,

I have managed to overclock the ASUS M5A97 R2.0 to a bus speed (FSB) of 300 MHz
using the phenom x6 1045T.

But the second I try and set the bus speed (FSB) a bit higher like 305 MHz,
windows keeps crashing!

Would this be a limitation of the board since it's a 4+2 power phase design?

Or are there other options I can try to stabilize the higher bus speeds

examples:

1- would increasing the mobo related voltages in bios help? i tried a bit of that didn't seem to

2- Remove the stock thermal pads of the default mobo heatsinks and replace them with AS5?

3- Add more heatsinks on the mobo to strategic components?

Any ideas will be appreciated.

thanks

[natr0n]

You need a fan blowing on the northbridge.

[PAPADOC]

Originally Posted by natr0n

You need a fan blowing on the northbridge.

Thanks,

I will try that tonight woo hoo !!!

[drdeathx]

Yes, that would be the max which is awesome but overclocking through the HTT(FSB) is pretty much history today. Now see how high you can get with 200 on the FSB and use the multi. Next play around with different multi's and HTT(FSB) to max the overclock. Overclocking through multi will usually require less voltage.......

[PAPADOC]

The Phenom X6 1045T has a locked multiplier can't raise it,
FSB is the only way to go for this chip !!!

[drdeathx]

Originally Posted by PAPADOC

The Phenom X6 1045T has a locked multiplier can't raise it,
FSB is the only way to go for this chip !!!

Ahhh, forgot about the locked multi. 300 on the HTT is awesome though.....

[de.das.dude]

thats the limit with phenoms i think

[PAPADOC]

I placed a 4000rpm fan right on top of the northbridge
but it didn't make any difference.

i also raised voltages to the northbridge but still not stable

[Aquinus]

What is your HT link and CPUNB running at? You might get more if you're pushing either of them too high. I would try to keep the CPU-NB at 2400Mhz and HT at 2000Mhz. Overclocking HT won't give you any gains and will only work both your CPU and NB (on the motherboard,) harder. The CPUNB on the other hand can improve memory bandwidth, but it depends on your core clock. There comes a point where higher CPUNB speeds actually give you worse performance. For 3.7-4.0Ghz, 2400-2600Mhz is the sweet spot, but you have to test a little bit to get it perfect.

[de.das.dude]

Originally Posted by PAPADOC

I placed a 4000rpm fan right on top of the northbridge
but it didn't make any difference.

i also raised voltages to the northbridge but still not stable

it wont happen. 300 is the limit.

also why on earth would you need 300? you CPU will not be able to handle that much!

also i dont think your nothbridge wants to go till 3Ghz either.

[Aquinus]

Originally Posted by de.das.dude

also i dont think your nothbridge wants to go till 3Ghz either.

That's when you drop the CPUNB and the HT multiplier. Just because the base clock is high doesn't mean other components need to be.

[PAPADOC]

Originally Posted by Aquinus

What is your HT link and CPUNB running at? You might get more if you're pushing either of them too high. I would try to keep the CPU-NB at 2400Mhz and HT at 2000Mhz. Overclocking HT won't give you any gains and will only work both your CPU and NB (on the motherboard,) harder. The CPUNB on the other hand can improve memory bandwidth, but it depends on your core clock. There comes a point where higher CPUNB speeds actually give you worse performance. For 3.7-4.0Ghz, 2400-2600Mhz is the sweet spot, but you have to test a little bit to get it perfect.

Ok my cpu-nb is around 2450Mhz as my HT

I will try to just lower the HT closer to 2000Mhz see what happens

[OneMoar]

300 is the limit of that motherboard I am supprised you got more then 250
also that motherboard is gonna blow up if you keep trying to OC on it the VRM's can't handle it
set the FSB to 250 and pray it doesn't blow up you are not gonna get 4Ghz out of that Chip EVER no way no how its not possible it WILL break Stop now before you wreak something
that motherboard CAN NOT HANDEL THE POWER DRAW OF A 6 CORE CPU AT 4GHZ IT WILL BLOW UP

[Aquinus]

Originally Posted by OneMoar

also that motherboard is gonna blow up if you keep trying to OC on it the VRM's can't handle it
set the FSB to 250 and pray it doesn't blow up you are not gonna get 4Ghz out of that Chip EVER no way no how its not possible it WILL break Stop now before you wreak something
that motherboard CAN NOT HANDEL THE POWER DRAW OF A 6 CORE CPU AT 4GHZ IT WILL BLOW UP

+1: He is right. Don't work the VRMs too hard or you'll do permanent damage to both the motherboard and the CPU.

[PAPADOC]

Originally Posted by de.das.dude

it wont happen. 300 is the limit.

also why on earth would you need 300? you CPU will not be able to handle that much!

also i dont think your nothbridge wants to go till 3Ghz either.

To get higher benchmarks really.

My CPU is already handling 300Mhz for a sweet 4Ghz

But u may be right i may be at the limit right now !!!

[OneMoar]

Originally Posted by PAPADOC

To get higher benchmarks really.

My CPU is already handling 300Mhz for a sweet 4Ghz

But u may be right i may be at the limit right now !!!

i Give the motherboard a month tops before the VRM's melt
I hope you have money for another board AND cpu

[PAPADOC]

Originally Posted by OneMoar

300 is the limit of that motherboard I am supprised you got more then 250
also that motherboard is gonna blow up if you keep trying to OC on it the VRM's can't handle it
set the FSB to 250 and pray it doesn't blow up you are not gonna get 4Ghz out of that Chip EVER no way no how its not possible it WILL break Stop now before you wreak something
that motherboard CAN NOT HANDEL THE POWER DRAW OF A 6 CORE CPU AT 4GHZ IT WILL BLOW UP

My CPU is already handling 300Mhz for a sweet 4Ghz

But u may be right i may be at the limit right now !!!

I hope it doesn't blow up

[OneMoar]

your cpu is currently drawing about a 140 to 160 watts(160 if you vcore is >1.50) that motherboard is rated for a MAX of 95 Watts its a cheap board
when IT blows you will be lucky if it doesn't take the CPU outwith it
NEVER oc a high-TDP chip on a cheap board thats just asking for trouble

[cadaveca]

HTT scaling is a function of the PLL chip used that actually generates those clocks. This is the one from my M5A97 EVO R1.0 board:




It generates clocks for both the CPU and the SB(since both connect using HTT)

I've seen x6 chips hit over 350 HTT. PLL used is a big part of that, but as mentioned, it's probably a chip limit, although 300 MHz is by no means the maximum for those chips.

Originally Posted by OneMoar

your cpu is currently drawing about a 140 to 160 watts(160 if you vcore is >1.50) that motherboard is rated for a MAX of 95 Watts its a cheap board
when IT blows you will be lucky if it doesn't take the CPU outwith it
NEVER oc a high-TDP chip on a cheap board thats just asking for trouble

No. Just no. 95W.....you are very much wrong. Where did you get THAT from?


The board supports all AM3+ chips(including 140W x4 BE 965), plus OC.

I mean really, just check the QVL:

http://www.asus.com/Motherboards/AMD..._EVO_R20/#CPUS

[OneMoar]

Originally Posted by cadaveca

HTT scaling is a function of the PLL chip used that actually generates those clocks. This is the one from my M5A97 EVO R1.0 board:

http://www.techpowerup.com/reviews/A...controller.jpg


It generates clocks for both the CPU and the SB(since both connect using HTT)

I've seen x6 chips hit over 350 HTT. PLL used is a big part of that, but as mentioned, it's probably a chip limit, although 300 MHz is by no means the maximum.

the 'speed' of the host-clock is just that a CLOCK its what the cpu uses to keep in'in step
200 x 13.5 = 2700 (proc speed)
200 x 10 = 2000 (HT/NB speed)
dave is right. tho the problem IS two fold the cpu is probably hitting a power supply/stability wall and the motherboard isn't liking that hostclock set so high because it has a affect on other components
the R2.0 is a garbage com-paired to the 1.0 they cut a bunch of function from it to lower costs

[OneMoar]

Originally Posted by cadaveca

HTT scaling is a function of the PLL chip used that actually generates those clocks. This is the one from my M5A97 EVO R1.0 board:

http://www.techpowerup.com/reviews/A...controller.jpg


It generates clocks for both the CPU and the SB(since both connect using HTT)

I've seen x6 chips hit over 350 HTT. PLL used is a big part of that, but as mentioned, it's probably a chip limit, although 300 MHz is by no means the maximum for those chips.




No. Just no. 95W.....you are very much wrong. Where did you get THAT from?


The board supports all AM3+ chips(including 140W x4 BE 965), plus OC.

I mean really, just check the QVL:

http://www.asus.com/Motherboards/AMD..._EVO_R20/#CPUS

Orly so they added support because I had a R2.0 for about a week before I bought this and was told by ASUS support that it didn't support 125TDP because my Phenom II would't post in it so someone someplace is lieing ...
I quit reading the QVL's when vendors starting listing 125TDP on boards with 4+1 phase VRM's
and the Op Says his board is 4+2 so its either NOT a m5A97 EVO or hes wrong
even assuming that the board is capable of outputting a sustained 140 the Op is exceeding that by at least 15-40W(depending on what his vcore is) granted not a whole lot but on a "4+2/6Phase" thats pushing your luck in my book hell I keep a eye on my 8+2's on this board if its one thing you don't wanna go chancing its the VRM's
edit the board is indeed 4+2 tho the vrm database on overclock.net says it should handle the strain they don't say what chip and at what speed they tested at

[PAPADOC]

Originally Posted by OneMoar

your cpu is currently drawing about a 140 to 160 watts(160 if you vcore is >1.50) that motherboard is rated for a MAX of 95 Watts its a cheap board
when IT blows you will be lucky if it doesn't take the CPU outwith it
NEVER oc a high-TDP chip on a cheap board thats just asking for trouble

This board supports max 140 watts TDP

source:

http://www.asus.com/Motherboards/AMD...specifications

http://www.overclock.net/t/946407/am...-info-database

but yeah i forgot about going over 140 watts , hehe got to check that tonight

[cadaveca]

Originally Posted by OneMoar

Orly so they added support because I had a R2.0 for about a week before I bought this and was told by ASUS support that it didn't support 125TDP because my Phenom II would't post in it so someone someplace is lieing ...
I quit reading the QVL's when vendors starting listing 125TDP on boards with 4+1 phase VRM's
and the Op Says his board is 4+2 so its either NOT a m5A97 EVO or hes wrong
even assuming that the board is capable of outputting a sustained 140 the Op is exceeding that by at least 15-40W(depending on what his vcore is) granted not a whole lot but on a "4+2/6Phase" thats pushing your luck in my book hell I keep a eye on my 8+2's on this board if its one thing you don't wanna go chancing its the VRM's
edit the board is indeed 4+2 tho the vrm database on overclock.net says it should handle the strain they don't say what chip and at what speed they tested at

I am not sure why ASUS would have told you that, since this board has supported all CPUs since it was launched, on the first BIOS.

The number of VRM phases hardly dictates how good a board is, unfortunately. What matters is what those phases are made with, since some easily support of 60a each. With four phases that support 60a, that's far more than most would ever need when running a 24/7 OC. Four 25a phases...sure, noise may limit things a bit, but not that much.

Originally Posted by PAPADOC

This board supports max 140 watts TDP

source:

http://www.asus.com/Motherboards/AMD...specifications

http://www.overclock.net/t/946407/am...-info-database

but yeah i forgot about going over 140 watts , hehe got to check that tonight

That VRM has no problem pushing 280W++. I've done it myself. VRM is 6+2+2 phases, not 4+ whatever.

[PAPADOC]

Originally Posted by OneMoar

Orly so they added support because I had a R2.0 for about a week before I bought this and was told by ASUS support that it didn't support 125TDP because my Phenom II would't post in it so someone someplace is lieing ...
I quit reading the QVL's when vendors starting listing 125TDP on boards with 4+1 phase VRM's
and the Op Says his board is 4+2 so its either NOT a m5A97 EVO or hes wrong
even assuming that the board is capable of outputting a sustained 140 the Op is exceeding that by at least 15-40W(depending on what his vcore is) granted not a whole lot but on a "4+2/6Phase" thats pushing your luck in my book hell I keep a eye on my 8+2's on this board if its one thing you don't wanna go chancing its the VRM's
edit the board is indeed 4+2 tho the vrm database on overclock.net says it should handle the strain they don't say what chip and at what speed they tested at

My Vcore is at 1.5 v with LLC on

I am currently at work , but this is why i turn it off when i am not there don't want any suprise fires

[cadaveca]

Originally Posted by PAPADOC

My Vcore is at 1.5 v with LLC on

Which doesn't matter. What TPU needs is age verification.


I pushed 1.65V @ 4.8 GHz no problem on one of those boards, using an 1100T. nearly 6 months later, the board is fine, as is the chip, and it's not my rig. The board is not part of what's limiting you. It's most likely your ram.