Jump to content


Photo
- - - - -

WARNING to all Future AMD 64 3200+ Venice Shoppers!


  • Please log in to reply
27 replies to this topic

#1 Vulkanonistix

Vulkanonistix

    Member

  • Members
  • PipPip
  • 87 posts

Posted 29 December 2005 - 02:00 PM

I noticed very weird overclock for my Venice 3200. I worked hours and hours searching for the stable point but I could never get the CPU past 2.5 GHz even with 1.6 Vcore volts! So I am actually stuck but 100% stable at 2.5 GHz. However, whats interesting is I ran this CPU at this voltage throughout Prime95 stress testing for at least 1 hour 8 minutes. There were no errors even with OCCT regular testing for 30 minutes. I even ran 3dMark05 and 3dMark03 without any crashes. So the highest my CPU went up to is 2.5. What amazed me is the temperature when I did the prime testing. The temperature only went up to 39 degrees the highest according to my expert reading. When I used the probe I just ordered from Japan, I got 37.4 as the reading. The chipset stayed very low around 38 degrees. The motherboard temperature went up to 41 degrees because the CPU fan was dumping out a lot of heat at 5000 RPM. The memory voltage was increased only to 2.80 volts for stability. I didn't raise anything else.
My conclusion from this overclock is that NEVER EVER BUY AN AMD 64 3200+ VENICE THAT HAS THE FOLLOWING MODEL NUMBER AND THE DATE WHEN CREATED:

ADA3200DAA4BP

LBBLE 0518APEW


I have been searching in google for these numbers and the owners also had the worst overclock with the 0518 version. Any Venice core that has been created on 0517 or before are said to be nice overclockers with less voltage.
Second thought is I couldn't believe how much voltage this CPU can take. I even tried 1.72 once for couple of minutes to see if I could reach my clock higher. Oh yea, I haven't done anything wrong in the overclocking process. I read the "Definitive Guide to A64 Overclocking" on PDF and then I went to HardForums and read their guide. The DFI guide was better so I did first. But imagine with a stock cooler running at this voltage and keeping the temperature at or below 40 degrees load.
Do you think it is dangerous for a Venice core to take this amount of voltage due to its 90 nanometer processing as opposed to older core such as Clawhammer and Newcastle? It is because I plan to use this overclocked speed when I am doing CPU intensive applications that need a little more performance boost.
Here is the screenshot of my desktop by the way, desktop.

#2 CygnusX1

CygnusX1

    Member

  • Members
  • PipPip
  • 91 posts

Posted 29 December 2005 - 02:28 PM

Mhh my venice 3200 does 2540 @ 1.58vcore with Heatspreader removed. This was the max. But it is the 0517, the ones you proclaim to be good. But this was done with a passive watercooling, will try it tomorrow with the fans on, forgot to plug them in :x

#3 Vulkanonistix

Vulkanonistix

    Member

  • Members
  • PipPip
  • 87 posts

Posted 29 December 2005 - 05:58 PM

That is an interesting update. Thank you for notifying. Do others think running my cpu at this voltage is safe even though the temperature are low at load?

#4 J054

J054

    Member

  • Members
  • PipPip
  • 72 posts

Posted 29 December 2005 - 06:14 PM

I run mine at 1.58volts at 2.6 stable
I upped it to 1.7 and got 2.72 stable but felt funny about voltage, temps were fine though. It should be fine the only thing im worried about is long term effects, and if a cpu blows do all the components connected to it blow too?

#5 Vulkanonistix

Vulkanonistix

    Member

  • Members
  • PipPip
  • 87 posts

Posted 29 December 2005 - 07:12 PM

I don't think other components like the capacitors will blow up because DFI designed the motherboard to take high voltages to four main parts - LDT bus, CPU, chipset, and RAM. So I highly doubt other components will break unless you have 0 airflow in your case. If the motherboard temperatures are at or below 45 degrees, then it is perfectly fine for long term effects. Why else would DFI include so many voltage options and be proclaimed to be one of the best stable overclocking board? Besides, I never bump up voltages on other components unless if it becomes necessary like the RAM. But LDT and chipset voltage, I don't see how that will benefit me with stability.

#6 Thasp

Thasp

    Certified Geek

  • Members
  • PipPipPipPip
  • 1638 posts

Posted 29 December 2005 - 07:43 PM

I noticed very weird overclock for my Venice 3200. I worked hours and hours searching for the stable point but I could never get the CPU past 2.5 GHz even with 1.6 Vcore volts!


You say that as if your chip is horrible because it only overclocks 25%. My 3000+ requires 1.575v-1.625v(haven't done more than 1 hour rpime with 1.575v, did 9 hrs with 1.625v) to get to 2.5 GHz.

I know people with 3500+ CPUs that can't get over 2.5 Ghz without boatloads of voltage(into watercooling territory)

You're not entitled to any overclock, much less a 25% one. Granted, some chips do better than others, but the only speed you're 100% entitled to is what it says on the package.

#7 Pure_evil

Pure_evil

    New Member

  • Members
  • 36 posts

Posted 30 December 2005 - 05:25 AM

oink

my venice (3000+, don't know the numbers cause I still gotta remove my hs to put some AS5 on it) takes me all the way up to 2.85GHz with 1.47V but my ram (that 2.85 was tested with other ram) only allows to go to 2.7 @ 1.44V

I thought almost all venice's were great oc'ers, guess I was wrong.
Feel sorry for u ppl though :s

#8 hawkeyefan

hawkeyefan

    Member

  • Members
  • PipPip
  • 335 posts

Posted 30 December 2005 - 12:29 PM

in fairness to all 0518APEW and 0518BPEW, I have seen several of these do better than my 0517APAW. these chips remain good clockers on a hit-and-miss basis up through week 0530 or so. You can't make blanket statements about a week/stepping/batch# because every chip in the batch will get different results.

looks like you got the miss?:)

#9 EVOfreak

EVOfreak

    New Member

  • Members
  • 33 posts

Posted 30 December 2005 - 12:33 PM

my venice is very similar to yours

my chip is
ADA3200DAA4BP
LBBLE 0540APDW

right now i'm at 2.6 ghz @ 1.5 volts with stock cooling. Temp is 38c idle and 50 load. I haven't tried to push this to the max due to my stock cooling. I also hope i can push this atleast 2.8.

#10 CMoney85

CMoney85

    New Member

  • Members
  • 3 posts

Posted 30 December 2005 - 01:02 PM

my 3200 venice runs just fine its rock stable for 13hrs of prime at 2.5ghz at 1.392 volts :)

#11 KimTjik

KimTjik

    Total Nerd

  • Members
  • PipPipPipPipPip
  • 3207 posts

Posted 30 December 2005 - 01:20 PM

Vulkanonistix - don't you think there's a great variety between so called identical processors? And that sometimes it's not just decided by the CPU, but also by the "individuality" of the whole system?

I see differencies what concerns my own Venice 3200+ E6. If I don't push it over 1.55V it will stay stabile at about 2.60Ghz. If I let it go to about 1.65 I can reach 1.72-1.75 stabile on air. I feel lucky, even though some get even better results. When I bought it I didn't expect a miracle, since to be fair: Venice 3200+ is good value since it's quite copetitive with much more expensive CPU:s. Do I then get +20% or over, great, but I don't demand it.

Do you think it is dangerous for a Venice core to take this amount of voltage due to its 90 nanometer processing as opposed to older core such as Clawhammer and Newcastle? It is because I plan to use this overclocked speed when I am doing CPU intensive applications that need a little more performance boost.


If OC you can never expect anyone to give you an guarantee like: "that's safe or that's not safe". Since these CPU hasn't been out on the market for so long, we still haven't much more guidence than the estimations and recommendations already given at this forum. Nevertheless, I wouldn't advice anyone to use a OC system for sensitive applications.

#12 Vulkanonistix

Vulkanonistix

    Member

  • Members
  • PipPip
  • 87 posts

Posted 30 December 2005 - 04:17 PM

Thank you guys for your thoughts and experiences in oc'ing Venice chips. I guess I do have a crappy one. But to me, I am still happy with my stock clock after upgrading from 2.4 GHz Pentium 4 Northwood.