Athlon II 620 Oced

kbaer89

Weaksauce
Joined
Dec 28, 2009
Messages
78
Does anyone have any experience overclocking the 620?

I have an AM2+ Mobo.
 
they should overclock decently.. theres a few reviews out there with overclock numbers.. you might want to take a look at those.. should give you an idea of where to set a goal at.. just remember it will be a pain in the ass overclocking it since you have to deal with the FSB, HT multiplier, and Ram multiplier..
 
I used this proc for a build for my mom. OC'd to 3.0ghz w/stock cooling no problem. Didn't wanna push any further since she doesn't care about OC'ing and we had stock cooling. Could probably hit 3.2ghz without incident.

This was using a 785G Biostar uATX board.
 
I have seen a few reviews including overclocking results and benchmarks. I believe TweakTown got it up to 3.9GHz; it generally outperformed a stock Phenom II X4 955 in most real world tests.
 
Easy overclock if you have ddr2-800- manually set the ram speed to 667, set your timings-
Raise fsb to 240 - Done! (~3.2ghz)
 
After about 10 hours of tweaking I'm at 3.25GHz. RAM will not go above 1333MHz, (Tom's review points out that this is to be expected, but I'm still disappointed after laying out for high-end RAM....) but the timings are good.

I wonder if the CPU is holding me back.

I've looked at a bunch! of reviews and forum sites and can't find a real "enthusiast" thread for my MB, so all voltages except RAM are on Auto. (The RAM is spec'd 1.85-1.95V so I have it set at 1.92V as that's the only value the MB supports that falls in the range.) I tried a number of iterations with modified voltages for everything including: CPU VDD, CPU-NB VDD, CPU, CPU-NB, NB & HT Link, but I was just guessing really, and whatever I did in my limited attempts didn't seem to make a difference from the Auto values.....

At this OC I can only run IntelBurnTest once as the temps go through the roof with the stock cooler/thermal paste. But it does pass each time. I've tested with Prime95 and OCCT as well, but they don't seem to be nearly as stressful as IBT.

751952120_FNNba-X3.jpg


751952103_hD6eU-X3.jpg


751952152_NR5Fd-X3.jpg


751952145_8SAFf-X3.jpg


751952157_yJjzf-X3.jpg


751952119_Jm9iF-X3.jpg
 
After about 10 hours of tweaking I'm at 3.25GHz.
I wonder if the CPU is holding me back.

At this OC I can only run IntelBurnTest once as the temps go through the roof with the stock cooler/thermal paste.
I would think that at stock voltage . . . the voltage is holding you back. And with any increase in voltage, the heatsink would be holding you back. Also, by most accounts 3.2GHz or so is the wall at stock settings.
 
id also say raising the NB voltage will help get the ram above ddr3 1333.. but on stock cooling i wouldnt really recommend it.. from my experience the NB voltage has a huge effect on cpu temps..
 
I would think that at stock voltage . . . the voltage is holding you back. And with any increase in voltage, the heatsink would be holding you back. Also, by most accounts 3.2GHz or so is the wall at stock settings.

Yeah, I tried;

-- All Auto, except RAM
-- All Auto, except RAM and CPU VCore at 1.425V (this was the value in software, higher in BIOS settings 1.43xV? which was reported even lower in the BIOS report (I think 1.3xV.))
-- And then a mix of various bumps up from Auto for everything I listed above. 15+ iterations. The BIOS on this MB shows "Not Recommended" values i.e., too high or too low -- in RED, so I tried to keep everything just below that level when I was pushing FSB to 270~280.)

The Auto value for CPU VCore is 1.384-1.392

Almost every crash happened within seconds of starting IBT/Prime95/OCCT, though a few boots made it to a minute or two. I read somewhere that a screen freeze is typically not enough CPU VCore, and BSOD's are something else. That seemed to be the case with my testing.

With my current settings it seems to be stable, as far as I can test it anyway. But it's been running for a couple days solid with some testing thrown in every now and then. The CPU-Z screen above shows a low CPU VCore as I have C&Q enabled. The last thing I did after settling on the current settings was to re-enable it, and it didn't affect anything at this level OC.

id also say raising the NB voltage will help get the ram above ddr3 1333.. but on stock cooling i wouldnt really recommend it.. from my experience the NB voltage has a huge effect on cpu temps..

I guess I need to shell out for a better cooler if I want to go any further. But with the crashes happening so quickly after starting testing at a given level, I'm not so sure.
 
Here the Tom's comment I aluded to above:
Tom's said:
The MSI 785GM-E65 couldn't boot our Mushkin XP3-12800 at 1,600 MT/s using relaxed 9-10-10-25 timings, even though this memory is sold as a 1,600 MHz 7-8-7-20 kit. Once again, we know that it's difficult to get AM3 CPUs working with memory speeds faster than 1,333 MHz, and we appreciate that MSI doesn't push this as a board that can handle 1,600 MHz memory. While we did find a small 1,600 MHz overclocking claim on the back of the box, the manual lists 1,333 MHz as the highest compatible memory frequency.

http://www.tomshardware.com/reviews/amd-785g-motherboard,2441-16.html

I did use relaxed timings (Auto=8-8-8-29) when trying for 1600MHz...

In almost all of my testing I did reduce the multpliers for HT-L/CPU-NB to keep them at or slightly below 2GHz, a couple times using the higher end of the range for voltages I went slightly above, like 2.1GHz..

Here's a good blog report for getting into the 3.7GHz range with this CPU. (He mentions his sample of the MB I have won't get past the 260 range, but I have seen other reviews where it has, and I'll have to dig them up and make a new thread.) http://www.techreaction.net/2009/09/25/athlon-ii-x4-620-ferrari-this-mustang-can-hang/
 
Last edited:
I guess I need to shell out for a better cooler if I want to go any further. But with the crashes happening so quickly after starting testing at a given level, I'm not so sure.
I think it's worth a shot. As I said, I have seen some pretty impressive numbers.
 
I think it's worth a shot. As I said, I have seen some pretty impressive numbers.

Good stuff, thanks for the link.

From the article linked above.... said:
The one thing that the Athlon II has in abundance is overclocking headroom. Our stock sample 2.6 GHz part was able to hit 3.9 GHz on air. And maybe with better cooling, more time and some fine tuning, overclocking to 4 GHz and beyond might be obtainable.

To us it seems that the L3 cache on Phenom II processors hampers the clock speeds quite a bit. We used the same motherboard that we overclocked the Phenom II with; it only managed a few megahertz extra FSB to increase in speed. But with the Athlon II, we managed to clock things right to the max, as we had to rely on FSB as the multipliers were locked.

AMD Athlon II is a cheap overclocking CPU that really isn't worth it at stock, the real gem of this CPU is the ability to overclock. If you're going to stick only to stock clock speeds go with the Phenom II, but for overclocking take the Athlon II and clock it through the roof.
 
Yeah, I tried;

-- All Auto, except RAM
-- All Auto, except RAM and CPU VCore at 1.425V (this was the value in software, higher in BIOS settings 1.43xV? which was reported even lower in the BIOS report (I think 1.3xV.))
-- And then a mix of various bumps up from Auto for everything I listed above. 15+ iterations. The BIOS on this MB shows "Not Recommended" values i.e., too high or too low -- in RED, so I tried to keep everything just below that level when I was pushing FSB to 270~280.)

The Auto value for CPU VCore is 1.384-1.392

Almost every crash happened within seconds of starting IBT/Prime95/OCCT, though a few boots made it to a minute or two. I read somewhere that a screen freeze is typically not enough CPU VCore, and BSOD's are something else. That seemed to be the case with my testing.

With my current settings it seems to be stable, as far as I can test it anyway. But it's been running for a couple days solid with some testing thrown in every now and then. The CPU-Z screen above shows a low CPU VCore as I have C&Q enabled. The last thing I did after settling on the current settings was to re-enable it, and it didn't affect anything at this level OC.



I guess I need to shell out for a better cooler if I want to go any further. But with the crashes happening so quickly after starting testing at a given level, I'm not so sure.


whats the BSOD error? if its machine exception error then that is cpu voltage.. thats the only thing that would cause that bsod error.. memory you will usually the IRQ not equal or what ever crap.. typically you will get the machine exception error bsod if you are really close to the stable voltage.. if you arent then it will usually just crash and restart..
 
With the right board these OC really good, my 620 is at 3.8GHz with 1.425 vcore. I can do 3.9 with 1.5v but I keep it at 3.8 24/7. I have a Gigabyte 790XT-UD4P (I love this board very stable) with 4GB of Supertalent DDR3 and a Coolermaster N520 CPU cooler. Anything over 1300 MHz for the memmory and it isn't stable, I think it's the ram because it's rated at 1.5v and the board doesn't go lower than 1.6v. With better suited ram I think I could possibly hit 4 GHz.
 
I got my 620 to 3.3ghz pretty easily. RAM is running at 667mhz. I ran a torture test for 6 hours with no errors.

I'm going to try to bump the fsb mhz up to 260 and try to get the CPU over 3.5ghz.
 
I got my 620 to 3.3ghz pretty easily. RAM is running at 667mhz. I ran a torture test for 6 hours with no errors.

I'm going to try to bump the fsb mhz up to 260 and try to get the CPU over 3.5ghz.

I'm going to install the new Noctua cooler and probably upgrade my PSU before I try again. I've found that my 620 at 3.25GHz feels about the same performance-wise as the E8400 in my sig.

Here's my target for the MSI board I'm using: :eek:
http://www.techreaction.net/2009/10/05/msi-785gm-e65-m-atx-unlocks-and-overclocks/

whats the BSOD error? if its machine exception error then that is cpu voltage.. thats the only thing that would cause that bsod error.. memory you will usually the IRQ not equal or what ever crap.. typically you will get the machine exception error bsod if you are really close to the stable voltage.. if you arent then it will usually just crash and restart..

Not sure, I'll make better notes next time. Thanks though.
 
Back
Top