Yet Another Overclocking Thread. 16 replies

  • 1
  • 2

Please wait...

rob.

I am the Walrus

50 XP

24th October 2004

0 Uploads

5,580 Posts

0 Threads

#1 12 years ago

Well, I have been fiddling around, and I upped my CPU from 2.5 to 2.6.

Now if anyone remembers my old thread, I had serious issues with 2.6 being stable at the stock vcore, so I increased the voltage to 1.5V.

Now what also happened before, is my HT Frequency would drop, and my Memory would go up.

So what I did now, was increase the HT multiplier from 4x, to 5x... and I got this result. (Look at nvmonitor 'speeds' section)

http://img146.imageshack.us/img146/6043/woahnelly2qz.jpg

CPU is running a good 2 degrees warmer, but I'm not worried about that.

I just want to know if having the HT set to 5x is safe to do?

Edit - I ran a Prime95 Torture test with the 'Blend' setting, and it failed practacly 5 seconds in, displaying this. http://img93.imageshack.us/img93/7483/untitled19qo.jpg Hmmm,

Edit 2 - I ran Prime 95 Torture test with the 'Small FFT's' setting, it lasted around 6 minutes, and gave me the same error. http://img233.imageshack.us/img233/9486/1112sd.jpg I think it is a memory related problem. Any help would be very much appreciated.




rob.

I am the Walrus

50 XP

24th October 2004

0 Uploads

5,580 Posts

0 Threads

#2 12 years ago

Update, I ran S&M, Max Intensive test.

It passed all the CPU Tests which are,

Cache Level 1, and 2. CPU (Integer) CPU (FPU) Power Supply,

But the absolute second it moved onto the Memory Test, I got a shite load of errors. Programs started crashing, etc. I got a screenshot, but unfortunatly, when I hit Ok to quit the program, I recieved a BSOD. One I haven't seem before,

It said "STOP d0000144 Unknown Hard Error Unknown Hard Error Beginning Dump of physical memory Dumping physical memory to disk : 28*"

* 28 being the number it was up to.

It seems to be a memory problem.

But I don't know if it is my Ram Dimms, or the memory controller on my CPU.




MG42Maniac

A man of dubious moral fibre

50 XP

28th May 2003

0 Uploads

3,932 Posts

0 Threads

#3 12 years ago

I think something in your system is holding you back big time. I have a DFI Lanparty Ultra-D, 2GB G.Skill @ 2-3-2-5, stepping for my 3700+ is KACAE 0552FPBW and mine has hit 2.6GHz at 1.45V with an Artic Cooler 64 Pro. It would not let me boot into windows @ 2.8GHz so I might have reached my chips limit or it needs more volts. I got some errors in Prime 95 at 1.375 volts so I upped to volts too 1.45 volts and its passed the Prime 95 12 hour test.

This is the error I got on Prime 95 before I upped the volts This is my Current Settings CPUZ

If your on stock cooling I would not venture above 1.5v so it could be that you chip needs alot of volts to get too 2.6GHz stable, far more than you would want to put through it.

If your failing Prime 95 after 5 seconds you have an extremely unstable OC. I don't know the OC potential of a Asus mobo but I know my DFI can do 300 HTT without breaking a sweat so check to make sure your mobo can run comfortably at that HTT.

Also Geil Value is not the best RAM to be using if your going to OC so it might be your RAM is maxed out.




rob.

I am the Walrus

50 XP

24th October 2004

0 Uploads

5,580 Posts

0 Threads

#4 12 years ago

Yeah, the thing is it is fine at 2.5 at stock, or close to stock voltages. But the second I put it to 2.6, the shit hits the fan.

I had the vcore at 1.5v and at 2.6GHz, it was fine until I ran Prime95, gave me that error straight away. I was idling around 38 Degrees Centrigrade on 1.5v, this is with stock cooling.

I ran S&M, and the second it started testing my memory, it goofed and the computer crashed with BSOD. It passed all of the CPU tests though.

Now I hear the A8N is a pretty good overclocker, and C38368 himself uses an A8N.

You mentioned 300 HTT, Hyper-Transport Clock?

http://img146.imageshack.us/img146/6043/woahnelly2qz.jpg

Look at nvMonitor at the little bar graph, and tell me what is not ordinary, because I think those numbers are strangly high.

So here are my questions,

I need to fine out if it is my Ram, or my CPU's memory controller acting up.

Also, what would you recommend a great HSF kit for my CPU? Been looking at the thermalright XP-120, but the heatsink alone is $100 AUD.

Last but not least. If it is my shoddy Ram, which I hope it is because I was planning on buying new 2GB Dimm kit. What Ram would you recommend, for overclocking and gaming in mind aswell as practacality?

I like the OCZ Ram, but haven't found any suppliers in Australia that have any yet. Unless I order online.




MG42Maniac

A man of dubious moral fibre

50 XP

28th May 2003

0 Uploads

3,932 Posts

0 Threads

#5 12 years ago

Kommandant RoβYeah, the thing is it is fine at 2.5 at stock, or close to stock voltages. But the second I put it to 2.6, the shit hits the fan.

I had the vcore at 1.5v and at 2.6GHz, it was fine until I ran Prime95, gave me that error straight away. I was idling around 38 Degrees Centrigrade on 1.5v, this is with stock cooling.

I ran S&M, and the second it started testing my memory, it goofed and the computer crashed with BSOD. It passed all of the CPU tests though.

Now I hear the A8N is a pretty good overclocker, and C38368 himself uses an A8N.

You mentioned 300 HTT, Hyper-Transport Clock?

http://img146.imageshack.us/img146/6043/woahnelly2qz.jpg

Look at nvMonitor at the little bar graph, and tell me what is not ordinary, because I think those numbers are strangly high.

So here are my questions,

I need to fine out if it is my Ram, or my CPU's memory controller acting up.

Also, what would you recommend a great HSF kit for my CPU? Been looking at the thermalright XP-120, but the heatsink alone is $100 AUD.

Last but not least. If it is my shoddy Ram, which I hope it is because I was planning on buying new 2GB Dimm kit. What Ram would you recommend, for overclocking and gaming in mind aswell as practacality?

I like the OCZ Ram, but haven't found any suppliers in Australia that have If your running at 240 Hany yet. Unless I order online.

If your running at 240 HTT did you drop the HTT multiplier down to 4x? I know alot of boards dont like to run over 1000 HTT.

I also get that red bar in nvMonitor, but I don't really care as I never use that software and i've had no problems with my overclock.

Too see if its your RAM acting up you could run this simple little test

As far as RAM goes i'm using some G.Skill ZX which cost me £155 / $277 / 372 AUD. RAM is very good and will do 250 ush with slack timing though I run a divider to give better performance. In terms of coolers I'd look at the Artic Coole 4 Pro, its a very good HSF and keeps my CPU very cool and wasn't that expensive.




C38368

...burning angel wings to dust

50 XP

14th February 2004

0 Uploads

5,013 Posts

0 Threads

#6 12 years ago
MG42ManiacIf your running at 240 HTT did you drop the HTT multiplier down to 4x? I know alot of boards dont like to run over 1000 HTT.

QFE.




MG42Maniac

A man of dubious moral fibre

50 XP

28th May 2003

0 Uploads

3,932 Posts

0 Threads

#7 12 years ago

Something is definetly holding you back. I rasied my voltage to 1.475 and my HTT too 255 and I got this: 28ghz9qr.png

I've primed it for 2 hours for a quick stability test and ran 3DMark 01, 03, 05 & 06 all with improved scores and no errors.




rob.

I am the Walrus

50 XP

24th October 2004

0 Uploads

5,580 Posts

0 Threads

#8 12 years ago

Ah. I am such a dimwit.

It's obvious that my PC is having a fit due to my HT being what, 1200MHz?

1000MHz is pretty much as high as it will go, being stable.

Thanks for the help guys.

Quick Question MG42. It says you are running a 1.440 vcore at 2.8GHz?

That is insane, I thought you would have to have like 1.5v pumping into it, to run at that speed.

My in-experience with Overclocking is showing. But I am learning. ;)

Still going ahead with new ram soon, and cooling.

Thanks again.




Oblivious

I tawt I taw a puddy tat...

50 XP

30th December 2002

0 Uploads

2,806 Posts

0 Threads

#9 12 years ago

I've yet to really push my X2 to it's limits, but I intend to play around some more with it once I get a better cooler. The stock one is actually fairly beefy, but I still don't trust it. With my current OC, It idles at about 42 and under load reaches 49-50

This is probably apples and oranges here, but anyways....

I didn't have to raise the voltage too much to increase from 2.20ghz to 2.64ghz (240fsb x11). Voltage was 1.35, now it's 1.40. I think my htt multiplier is set to 4x so that's running at 960mhz. I hear on this board it can be stable at up to 1200mhz, but there's no measurable benefit to having it that high.

It seems stable as I ran Prime95 overnight, and everything passed. I also ran 3DMark05 and 06 with no problems.

I'm a complete noob to overclocking, so when my pc wouldn't even post or go into the bios after I'd made a few changes, I figured I'd fried something. I was relieved after clearing the cmos jumper that all was well. There's alot of variables involved, and I'm trying to take it slow for now.

14v5sn.jpg




rob.

I am the Walrus

50 XP

24th October 2004

0 Uploads

5,580 Posts

0 Threads

#10 12 years ago

Heh.... Running pretty much identical speeds to you Oblivious. Despite lack of second core for me. :lookaround:

http://img161.imageshack.us/img161/738/untitled19cz1.jpg

I tried 255MHz HTT, which is enormous. Computer failed to even get up to the Windows startup splash screen. So I moved it down to 240MHz HTT at 1.5Vcore,

Blue screened at the logon screen. But it down to 1.475v, running fine. Will run some tests tomorrow morning, it's late.

So, something is wrong if MG42 can run 2.8 at near stock voltages, and I can't at 1.5. :uhm:




  • 1
  • 2