BCLK Jumped to 137.6 - Accidently Joined the 6.6 GHz Club - Is my Motherboard Bad?

[apv507]

I have a Asrock Z97 Annivesary Motherboard and i7 4790k which I have stable at 4.8. I expect small fluctuations in bus clock speed but from 100 to 137.6 seems a little much. I haven't seen this happen before and I don't know if I should chalk it up to a fluke or if my Motherboard is having issues. Could this be a HWiNFO misread? Did I accidently break the 4790k OC record at 1.32v with an Evo 212 keeping it under 68c? Has anyone dealt with anything like this before? Thanks in advance!

 

 

 

[MrIceCremeLollipop]

I mean....The numbers make sense lol

[MartinKweh]

well. hey, unofficial record. YAY?

[MrIceCremeLollipop]

CPU-Z?

[apv507]

Just now, MrIceCremeLollipop said:

CPU-Z?

99.98, The 137.6 seems to have been a misread or momentary spike. Either a software error or hardware error. Hoping for software!

[MrIceCremeLollipop]

5 minutes ago, apv507 said:

99.98, The 137.6 seems to have been a misread or momentary spike. Either a software error or hardware error. Hoping for software!

ah you said in the original post. My bad...Wierd one 

[Guest]

Welcome to the 6ghz club!!!

 

this one's real.

 

[apv507]

3 hours ago, ShrimpBrime said:

Welcome to the 6ghz club!!!

 

this one's real.

 

Nice Overclock! I find AMD likes bus clock overclocking more than Intel but that's just my experience. What's your RAM and PCIE running at? 

[svmlegacy]

5 hours ago, apv507 said:

I have a Asrock Z97 Annivesary Motherboard and i7 4790k which I have stable at 4.8. I expect small fluctuations in bus clock speed but from 100 to 137.6 seems a little much. I haven't seen this happen before and I don't know if I should chalk it up to a fluke or if my Motherboard is having issues. Could this be a HWiNFO misread? Did I accidently break the 4790k OC record at 1.32v with an Evo 212 keeping it under 68c? Has anyone dealt with anything like this before? Thanks in advance!

Your system isn't 100% stable. I've got a record breaker here, on an i7-860 once upon a time. A whiff more voltage cleaned it up for me.

[apv507]

6 hours ago, svmlegacy said:

Your system isn't 100% stable. I've got a record breaker here, on an i7-860 once upon a time. A whiff more voltage cleaned it up for me.

Were  you regularly getting spikes like that?

Out of curiosity what is the reason a little more voltage prevents these spikes?  

 

Maybe Ill add .005 to the core, I currently have it set on 1.31 and adaptive so it ends up taking that extra .01 when it needs it.

 

I do have a 10 year old PSU with a replacement on the way as of a few days ago. Is there a chance that caused something weird to happen if voltage might have been the culprit?

 

Thanks for your help, the in depth specifics of hardware electrical functions is my area of least knowledge.

 

(I know there's a bit of a debate on overclocking and adaptive voltage, but I just love saving power when it's not needed and I tend to leave my machine on 24/7 as I connect from my phone for various reasons throughout the day. Plus 21c to 25c Idle temp while drawing only .7 volts is just beautiful to me.)

[Guest]

14 hours ago, apv507 said:

Nice Overclock! I find AMD likes bus clock overclocking more than Intel but that's just my experience. What's your RAM and PCIE running at? 

That's an older cpu-z (4 years ago). It was done on Dry Ice purposely to achieve maximum clock speed and nothing more. However this FX-4350 still holds highest frequency for this Cpu.

[svmlegacy]

7 hours ago, apv507 said:

Out of curiosity what is the reason a little more voltage prevents these spikes?  

Because the system isn't 100% stable, certain calculations can get erroneous numbers. You'll notice it's a rare occurance (Both your and my averages were right where it should be), By raising the voltage, you will make the system a little bit more stable. If you open even viewer, you'll likely see a lot of WHEA events about correctable errors by the processor.

 

I happened to be raising the system's base clock at the time, whereas yours seems to be the core itself.

[svmlegacy]

7 hours ago, apv507 said:

(I know there's a bit of a debate on overclocking and adaptive voltage, but I just love saving power when it's not needed and I tend to leave my machine on 24/7 as I connect from my phone for various reasons throughout the day. Plus 21c to 25c Idle temp while drawing only .7 volts is just beautiful to me.)

That's a good choice. My dead 4790K had lived most of its life at 4.6 GHz, without EIST or C-states, 24/7 run time. IIRC, it had about a 5 year life.

[apv507]

On 4/4/2020 at 3:02 PM, svmlegacy said:

That's a good choice. My dead 4790K had lived most of its life at 4.6 GHz, without EIST or C-states, 24/7 run time. IIRC, it had about a 5 year life.

I just got this CPU for Christmas and I'd love to get anything 5-7 years out of it. It's a legend in my book and it loves up to my expectations! Glad to know even at 4.6 24/7 I can expect 5 years. Thanks!

[apv507]

On 4/4/2020 at 3:00 PM, svmlegacy said:

Because the system isn't 100% stable, certain calculations can get erroneous numbers. You'll notice it's a rare occurance (Both your and my averages were right where it should be), By raising the voltage, you will make the system a little bit more stable. If you open even viewer, you'll likely see a lot of WHEA events about correctable errors by the processor.

 

I happened to be raising the system's base clock at the time, whereas yours seems to be the core itself.

Makes sense, I'll keep an eye out for WHEA events as well. I have 5c of headroom before I hit 90c under sustained stress testing so I am comfortable bumping up voltage a touch. Thanks for the insight!

[apv507]

9 minutes ago, Benji said:

I actually think that there is some misreport happening. If your BLCK were actually that high, the entire system would be so instable it wouldn't boot. Since the BLCK is also responsible for the PCIe clock and what not, basically no device would work properly anymore. Most configurations fail to boot around ~105MHz BLCK and the highest should probably around 110 if you are really lucky. But almost 140MHz? That seems unrealistic.

I agree. It was set at 100.0 I know they fluctuate but 37% seems odd. Hasn't happened since.

[svmlegacy]

5 hours ago, apv507 said:

I just got this CPU for Christmas and I'd love to get anything 5-7 years out of it. It's a legend in my book and it loves up to my expectations! Glad to know even at 4.6 24/7 I can expect 5 years. Thanks!

Unless it's NIB, keep in mind, it is still ~ 5 years old as it is. Many people buy K-SKU's and don't overclock though (They are typically the fastest in a range even without OC), so you may be lucky to have a "grandma-driven" chip.

[apv507]

7 hours ago, svmlegacy said:

Unless it's NIB, keep in mind, it is still ~ 5 years old as it is. Many people buy K-SKU's and don't overclock though (They are typically the fastest in a range even without OC), so you may be lucky to have a "grandma-driven" chip.

Brand Spanking New, possibly one of the last Newegg had in stock. Looks like Korean on the box so I know stocks were low as it wasn't intended for the US market. Luckily my wife snagged it before they sold out. I've wanted this CPU since it came out and love that I finally got one! 

[svmlegacy]

4 hours ago, apv507 said:

Brand Spanking New, possibly one of the last Newegg had in stock. Looks like Korean on the box so I know stocks were low as it wasn't intended for the US market. Luckily my wife snagged it before they sold out. I've wanted this CPU since it came out and love that I finally got one! 

That's an absolute gem. What week was it made, if you don't mind? 2nd digit of the batch number, is the last digit of the year, and the next two numbers are the week, Should be on the box, and the lid of the CPU.

[apv507]

13 hours ago, svmlegacy said:

That's an absolute gem. What week was it made, if you don't mind? 2nd digit of the batch number, is the last digit of the year, and the next two numbers are the week, Should be on the box, and the lid of the CPU.

Batch# is X601A586 - January of 2016 if I'm understanding. Is there anything else you can tell me based on that number?

 

I found a site that says X is somewhere in the USA, which is unexpected because it clearly looks like Chinese.

[papajo]

I believe something is broken in your CPU and it may die eventually those numbers (especially the 10Ghz one) are just junk you didnt really achieve those something is amiss I would try not to keep an OC on that CPU unless you are prepared to upgrade soon. 

 

Or there might be something wrong with windows (which is best case scenario since it means your CPU will not die) try to install fresh windows on a new partition and see if you get similar numbers if the answer is yes then the news are bad. 

 

You can also try to verify that with  intel's XTU and CPU-Z because it might just be that for some reason hwinfo bugged out (even better case scenario) 

 

edit and upon researching a little bit about it, it turns out that hwinfo can bug out that way https://www.hwinfo.com/forum/threads/hwinfo-maximum-cpu-mhz-showing-crazy-numbers.1310/ so that's good news your CPU probably isnt dying and you dont need to format windows

[svmlegacy]

4 hours ago, apv507 said:

Batch# is X601A586 - January of 2016 if I'm understanding. Is there anything else you can tell me based on that number?

 

I found a site that says X is somewhere in the USA, which is unexpected because it clearly looks like Chinese.

The X means the CPU itself was assembled in USA. Intel typically packages in a different plant, so it could very easily have been packaged in another country. That's a relatively late 4790K, but probably not one of the latest. As far as I can tell, the last shipment date for the 4790K would have been 2017-07-14, for tray or boxed.

[svmlegacy]

36 minutes ago, papajo said:

hwinfo can bug out

Do note that the bug was resolved in 2014. This CPU isn't dying, however, just a little unstable at the overclock OP was at.

[papajo]

15 minutes ago, svmlegacy said:

Do note that the bug was resolved in 2014. This CPU isn't dying, however, just a little unstable at the overclock OP was at.

Well the discussion was in 2014 doesnt mean that a bug cant reappear or that its not a different bug the sure thing is that if that CPU run at 6GHz on air  or on AIO it would have gotten fried (there is a reason people use Liquid nitrogen a -195C substance to cool CPUs in order to get close to such clocks.. because otherwise the system wouldnt even boot or if all the system safety measures were disabled it would start to smoke in a matter of seconds, magic simply does not happen.) 

 

And if its not a bug on hwinfo side of things then its even worse since (incase its not a windows bug) means that the internal power and clock measurement circuitry of that CPU has gone bananas which will eventually lead to the CPU dying because it will do something it was not supposed to do and those circuits that are responsible for checking that stuff out are failing. 

[apv507]

5 hours ago, papajo said:

I believe something is broken in your CPU and it may die eventually those numbers (especially the 10Ghz one) are just junk you didnt really achieve those something is amiss I would try not to keep an OC on that CPU unless you are prepared to upgrade soon. 

 

Or there might be something wrong with windows (which is best case scenario since it means your CPU will not die) try to install fresh windows on a new partition and see if you get similar numbers if the answer is yes then the news are bad. 

 

You can also try to verify that with  intel's XTU and CPU-Z because it might just be that for some reason hwinfo bugged out (even better case scenario) 

 

edit and upon researching a little bit about it, it turns out that hwinfo can bug out that way https://www.hwinfo.com/forum/threads/hwinfo-maximum-cpu-mhz-showing-crazy-numbers.1310/ so that's good news your CPU probably isnt dying and you dont need to format windows

Every other monitoring and reporting program hasn't gone above 4.8 GHz. Even HwiNFO hasn't displayed that number before or since. Temps and stability remain good.

 

Out of curiosity what 10Ghz number are you talking about?