Sign in with
Sign up | Sign in
Your question

First OC and looking for feedback

Tags:
  • CPUs
  • Overclocking
Last response: in Overclocking
Share
August 3, 2009 1:26:58 AM

I have read what seems like every thread here and elsewhere on the Internet regarding overclocking. After what I felt was at least a comfortable working knowledge of what I was getting into, I attempted my first OC. I would appreciate it very much if the "experts" here could critique my OC and give suggestions on how to improve upon it. Thank you in advance and be nice, I'm an OC virgin :D 

System Specs:

Gigabyte GA-X48-DS5
Intel Q9550 E0 stepping
Zalman CNPS9500
EVGA GeForce 9800 GTX+
8GB G.Skill PC2-9600
Antec 900 + 650w PSU

Here are the BIOS specs:

Robust Graphics Booster [Auto]
CPU Clock Ratio [8X]
Fine CPU Clock Ratio [+0.5]
CPU Frequency 3.23GHz (380x8.5)
CPU Host Clock Control [Enabled]
CPU Host Frequency (Mhz) [380]
PCI Express Frequency (Mhz) [100Mhz]
CIA2 [Disabled]
Performance Enhance [Standard]
System Memory Multiplier [Auto]
Memory Frequency (Mhz) [1066 1216]
DRAM Timing Schedule [Manual]

Standard Timing Control
-------------------------------
CAS Latency Time [5]
DRAM RAS# to CAS# Delay [5]
DRAM RAS# Precharge [5]
Precharge Delay (tRAS) [15]

Advanced Timing Control
-------------------------------
ACT to ACT Delay (tRRD) [4][Auto]
Rank Write to Read Delay [4][Auto]
Write to Precharge Delay [8][Auto]
Refresh to ACT Delay [68][Auto]
Read to Precharge Delay [4][Auto]
Static tRead Value [6][Auto]
Static tRead Phase Adjust [2][Auto]
Command Rate (CMD) [2][Auto]

Clock Driving of Skew Control
-------------------------------------
CPU/PCIEX Clock Driving Control [800mV]
CPU Clock Skew Control [Normal]
(G)MCH Clock Skew Control [Normal]
System Voltage Control [Manual]
DDR2 OverVoltage Control [Normal]
PCI-E OverVoltage Control [Normal]
FSB OverVoltage Control [Normal]
(G)MCH OverVoltage Control [Normal]
Loadline Calibration [Auto]
CPU Voltage Control [1.25000V]

Here's a screenshot of Prime95 in action from this afternoon. It ran for 5 hours without any hiccups and core temps never went over 58*C, which, as I understand it, is actually very low for 100% load for that long. Click for a larger pic.



Here's what I don't understand, even though BIOS says the CPU is running at 3.23 GHz (as well as Windows), CPU-Z only shows a multiplier of 8.0 (instead of 8.5 like BIOS) and hence a CPU running at only 3.04 GHz. Also, CPU-Z, as well as CPUID Hardware Monitor only show vcore at 1.2x during 100% load, despite having it set at 1.25000V in BIOS. As I understand it, this is vdrop, correct? The thing is though, it seems like no matter how high I set the BIOS, the vcore reading under full load only goes up to 1.2x. Of course, I didn't exceed the Intel rating of 1.3625v in the BIOS, so I don't know how that affects what CPU-Z and CPUID reflect.

Thanks again for taking the time to read and (hopefully) provide constructive criticism!

More about : feedback

a b à CPUs
a b K Overclocking
August 3, 2009 4:47:08 AM

FURTHER!
m
0
l
August 3, 2009 9:47:19 AM

Yes, I know it can go further, but as I am new at this, I was hoping for some pointers about exactly how to go further. I'm concerned that though I am adjusting the vcore in the BIOS, it's not being reflected in CPU-Z. I'm also concerned about the multiplier not being correct in CPU-Z. Should I not worry about either one of them? Can I take the vcore in CPU-Z to be fact and that even if I adjust vcore to above Intel's limit in BIOS, as long as CPU-Z is showing a vcore below Intel's limit that it's OK?
m
0
l
Related resources
a b à CPUs
August 3, 2009 11:46:34 AM

I agree with overshocked, you could take it further. The voltage setting on the mobo is the maximum value that your motherboard is going to supply the cpu. I've read somewhere that it is actually is impossible to supply an exact voltage to the cpu every single time (though I couldn't remember the full reason), so the setting is just the maximum value.

The upper limits of overclocking I think is either when you're at 1.4v core (I forgot if this is the maximum voltage for a C2Q) or when you're reaching 65~70C at full load.
m
0
l
August 3, 2009 1:00:29 PM

I am aware that the vcore set in the BIOS is the maximum, but why does it never seem to equate to what the core is actually getting? As of right now, it is under-volted from the stock settings, despite the CPU being OC'd. Stock vcore is 1.2625 I believe. I think Intel says max vcore is 1.3625?? If I lower vcore any further from where it is now in the BIOS, I was getting rounding errors in Prime95 with the OC because, according to CPU-Z, vcore was below 1.2 at 100% load. Is it safe to assume that vcore in CPU-Z is what the core is actually seeing? If so, then I will raise vcore in the BIOS to reflect what I really want it at in CPU-Z.
m
0
l
a b à CPUs
August 3, 2009 1:32:55 PM

From what I read here: http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx...

CPU-Z might actually be reading the actual voltage being supplied. The article explains the vDroop and vOffset which shows that it would be impossible for the processor to run at the actual voltage you specified.

Basically if you try to up the voltage to make it show 1.2625, your chip might hit that 1.3625 limit or way beyond when switching from heavy to light load. I think Intel's spec might indicate the max 1.3625 in the BIOS and not what CPU-Z reads, as it would be nearly impossible for the user to check at all times what the maximum voltage is being supplied.

If you tried lowering the voltage and ended up being unstable then it means that your cpu just can't run properly at any lower voltage, your only way is up.


*Also voltage doesn't only increase heat and power consumption, any increase in frequency also increases those two values. Read up on other OC articles and you'll notice that some OC with stock voltages and the heat and consumption also increases.
m
0
l
August 4, 2009 12:49:05 AM

Tried a couple combos tonight and the bottom line is, it's stable at 3.23GHz with vcore set at 1.25v in the BIOS, which is being reported as between 1.22v-1.23v in CPU-Z and CPUID. I raised the vcore in the BIOS to factory spec (1.2625v) and CPU-Z and CPUID still showed between 1.22v-1.23v. Raising the FSB anywhere beyond 380Mhz (where it's set now) caused errors with Prime95. So, anyone have any final thoughts before I call it quits and be thankful that it's stable at 3.23GHz?
m
0
l
August 4, 2009 1:09:25 AM

bump up the northbridge voltage to get your fsb up higher...you should be able to get over 400fsb on that chipset.
m
0
l
August 4, 2009 2:50:41 AM

That's the G(MCH) OverVoltage control, correct? Any suggestions for how much I should bump it up?
m
0
l
a b à CPUs
August 4, 2009 9:25:39 AM


According to this guide: http://www.tomshardware.com/forum/240001-29-howto-overc...

Quote:
NBCore – Start off low, 1.33 or 1.37 and see if you need more. Also, a little bit can go a long way. My system is unstable @ 1.330V here but stable @ 1.370V which is a difference of only 40 mV (0.04V).


Unfortunately I haven't tried to modify the voltage on my NB yet, so I couldn't really advise you much anymore. Try reading the guide, start with the 2nd post since the 1st post contains mostly introduction.
m
0
l
!