Venice OCing thread!

[Avalon]

Originally posted by: Duvie
Have you tried a 100 divider?? What divider are you running??? I have noticed in my recent testing that 200 divider works fine up to 2.44ghz (w/ default 1.4v) but withn that range I cannot hit 2.5ghz with 1.5v.....YOu better start steooing down the divider as you climb and dont be afreaid to be running 133 even if your ram wont be doing 400ddr at those HTT levels....

It was holding BJC back and I have seen it many times....forget what your ram says it will do....what vdimm have you been applying so far.....That wall is likely a memory issue....Every time I have seen someone have one like this they eventually got past was due to memory.....

Duvie, I am running a 133 divider, which puts my memory no higher than 200mhz
I actually don't even know the max of my UTT since this is the first time I've had a board that could supply it more than 3.1v to them.

TrueWisdom: I'll try running at 2.5-4-4-8, but I don't think UTT can do CAS 3. My plans for the night just got canceled, so I'm tossing my Zalman on right now. I did try 2T Command, but it didn't help.

*Edit*
I installed my Zalman, and noticed it was only spinning at 1300RPM as opposed to 2600RPM. Sure enough, I checked in my BIOS and it had all the fans thermally set at 50% speed while my CPU/PSU/NB respectively were under 50C. All of them are easily under 50C. I shut that off, so now they're all running at max. That, and with the Zalman on, I'll see what happens now.

 

[habs01]

Update - 2610 Mhz prime stable 6h34

-=General Info=-
CPU Model Number: ADA3000DAA4BP
CPU Code: LBBLE 0516DPJW

-=System Info=-
Motherboard: DFI NF4 Ultra-D
BIOS Revision: 310
Memory: 2x512 Geil PC3200
PSU: OCZ Powerstream 420W
Cooling Type, Brand, Model: Air, Zalman 7700cu

-=Stable OC Info=-
Max Stable OC: 2610 mhz
Prime95 SFFTs: 6:34 H:M
Prime95 LFFTs: N/A
HTT and Multiplier Used: 290 * 9
Actual VCore: 1.45V
CPU Temp in BIOS (Idle): haven't cheched
CPU Temp in OS (Idle):34C
CPU Temp Under Load: 47C
+12V Reading (Idle): 11.9V
+12V Reading (Load):11.9V
Memory: 7/10 divider

9x300 fails Prime95 within seconds though. I might try harder.

 

[Gusdawg]

My first post ever.... No more Lurking

Got my Venice 3200+ overclocking potential explored at default voltage and at 1.45 VCORE. (Bought from Monarch)

NOTE: I define "stable" as passing 3DMark05, Sandra, and 6+ hours of Prime,
I need more time to get 12 hrs in Prime....

Here max "stable" so far...
PU/Mobo FSBxMult=Speed @Vcore Mem speed (divider) latency Vdimm

3200+/VNF4ULTRA 306x9=2754 @1.45V mem 197 (2/3) 2.5-3-3-7-1T @2.60V
Passes 3DMark05, SANDRA for 10+ Hours, PRIME for 5 Hours so far (running)

3200+/VNF4ULTRA 302x9=2718 @1.40V mem 193 (2/3) 2.5-3-3-7-1T @2.60V
Passes 3DMark05, SANDRA for 4+ Hours, PRIME for 6+ Hours

-=General Info=-
CPU Model Number: ADA3200DAA4BP
CPU Code: LBBLE 0515APAW

-=System Info=-
Motherboard: CHAINTECH VNF4/ULTRA
BIOS Revision: 3.0 (3/15/05)
Memory: 2x512 Corsair Value Select PC3200 DDR400 2.5-3-3-8 1T
PSU: 420 Watt Turbolink LC-A420ATX, stock with ASPIRE X-Dreamer 2 case
Cooling Type, Brand, Model: Air, AMD Retail CPU HSF

-=Default Vcore=-
Max 6+ Hour Prime Stable: 2718 MHz (38C Under load)
HTT and Multiplier Used: 302 * 9

-=Stable OC Info=-
Max Stable OC: 2754 MHz (41C Under load)
Prime95: 6+ Hours so far
HTT and Multiplier Used: 306 * 9
Actual VCore: 1.45V (I set 1.45 cpu-z shows 1.44)
CPU Temp in BIOS (Idle): N/A
CPU Temp in OS (Idle): 32C
CPU Temp Under Load: 41C

-=Maximum Frequencies before failing:=-
FSB 351MHz @ 1.6V Vcore (boots)
CPU 2820+MHz @ 1.6V Vcore (did not find the ceiling yet, but 54C under load)
MEM 225MHz @ 2.9V DDR

I don't like to run 1.6 Vcore, CPU idles at 34C but Max temp under Prime is 54C when testing 2817 (9x313 and Ram @201.2 MHz)

Im plan on going with 306x9=2754 @1.45V mem 197 (2/3) and tighten the memory timings, or going with 270x10=2700 @1.45V mem 225 (5/6) with relaxed timings depending on a few more benchmarks...... (If only this Chaintech board had more miltipliers, dividers, and working Vcores!)

I'm highly pleased with my Venice, Chaintech board, and Corsair memory so far!

Thanks!

 

[user1234]

Originally posted by: Avalon

Originally posted by: Duvie
Have you tried a 100 divider?? What divider are you running??? I have noticed in my recent testing that 200 divider works fine up to 2.44ghz (w/ default 1.4v) but withn that range I cannot hit 2.5ghz with 1.5v.....YOu better start steooing down the divider as you climb and dont be afreaid to be running 133 even if your ram wont be doing 400ddr at those HTT levels....

It was holding BJC back and I have seen it many times....forget what your ram says it will do....what vdimm have you been applying so far.....That wall is likely a memory issue....Every time I have seen someone have one like this they eventually got past was due to memory.....

Duvie, I am running a 133 divider, which puts my memory no higher than 200mhz
I actually don't even know the max of my UTT since this is the first time I've had a board that could supply it more than 3.1v to them.

TrueWisdom: I'll try running at 2.5-4-4-8, but I don't think UTT can do CAS 3. My plans for the night just got canceled, so I'm tossing my Zalman on right now. I did try 2T Command, but it didn't help.

*Edit*
I installed my Zalman, and noticed it was only spinning at 1300RPM as opposed to 2600RPM. Sure enough, I checked in my BIOS and it had all the fans thermally set at 50% speed while my CPU/PSU/NB respectively were under 50C. All of them are easily under 50C. I shut that off, so now they're all running at max. That, and with the Zalman on, I'll see what happens now.

any updates ?

 

[d33pblue]

Alright, I dont have all of the info below, but I'll throw my hat into the ring.


-=General Info=-
CPU Model Number: ADA3000DAA4BP
CPU Code: LBBLE 0516CPW
CPU Serial Number:

-=System Info=-
Motherboard: DFI NF4 Ultra-D
Memory: 4x Corsiar VR @ 380mhz 3-4-4-8 2T 7/10 divider
PS: 400W Seasonic Supersilencer

-=Stable OC Info=-
Max Stable OC: 2450mhz @ 1.55v
Prime95 LFFTs: XX:XX Time
Prime95 SFFTs: XX:XX Time
CPU Temp Under Load: 38C

I get hard disk errors on boot if I try to put more than 1.55v through the processor. That, and several other signs tell me that my PSU is limiting my overclock. I dont feel like upgrading it right now though, so I think I'll just be happy with 2.45Ghz. I'm currently prime95ing at 2.45Ghz and its been going for three hours without a hiccup.

 

[birdpup]

d33pblue It may be possible to tighten your memory timings to a 5/6 (333/400) divider. I mention this because my memory runs at a 2/3 (266/400) at 290x9 and 5/6 (333/400) at 267x9. You seem to be in the 272 MHz range and may be able to achieve faster memory timings somewhere around the 267-272 memory range. I just noticed the 4x. Maybe it is not possible to tighten the memory timings with 4 dual sided 512MB sticks.

Gusdawg Very nice overclock. Especially with the stock heatsink. Do you have a 3000+ with a x9 multiplier or a 3200+ with a x10 multiplier? If you do have the 3200+, is there a reason you are using the x9 multiplier?

 

[d33pblue]

Originally posted by: birdpup
d33pblue It may be possible to tighten your memory timings to a 5/6 (333/400) divider. I mention this because my memory runs at a 2/3 (266/400) at 290x9 and 5/6 (333/400) at 267x9. You seem to be in the 272 MHz range and may be able to achieve faster memory timings somewhere around the 267-272 memory range. I just noticed the 4x. Maybe it is not possible to tighten the memory timings with 4 dual sided 512MB sticks.

I tried running 2.5-4-4-8 and the system wouldnt boot. So it appears I'm stuck at CL3. And just to clarify, I'm running 4 512MB sticks of Corsair Value select. They work fine at 400mhz, but seem to fall on their face around 405Mhz.

 

[Gusdawg]

Originally posted by: birdpup

Gusdawg Very nice overclock. Especially with the stock heatsink. Do you have a 3000+ with a x9 multiplier or a 3200+ with a x10 multiplier? If you do have the 3200+, is there a reason you are using the x9 multiplier?

I have a 3200+ I may run at 10x270 (2700) instead of 9x306 (2754) if the benchmarks look better that way. I can attain a higher CPU frequency by using the 9 multiplier. If I run much higher than 270 FSB with the 2/3 multiplier, my Ram starts to fail. I am limited to 2700 with the 10x because the memory tops out at 225 max stable and the only other divider option would give me low memory speeds. The Chaintech bios is currently limited to 1, 5/6, 2/3, and 1/2 dividers, no "working" voltage settings between 1.45 and 1.6, and integer multipliers only. I still recommend the VNF4 Ultra due to its cheap price and good overclocking regardless of these limitations. Also, I did get a Prime95 failure after 7Hr10min at 2754 (rounding error). I am pondering dropping the CPU clock another 9MHz and trying again, or is 7 Hrs good enough?

In my exploration of the best working combination, I noticed Prime95, when running a blend or long fft test, is the most overclock sensitive, followed by 3DMark05, followed by Prime95 connected to the server, and finally followed by Sissoft Sandra Lite 2005 and Aquamark (Sandra runs fine at 2790 1.4V but Prime95 locks instantly). Anyone else see this relationship?

 

[user1234]

Originally posted by: Gusdawg

Originally posted by: birdpup

Gusdawg Very nice overclock. Especially with the stock heatsink. Do you have a 3000+ with a x9 multiplier or a 3200+ with a x10 multiplier? If you do have the 3200+, is there a reason you are using the x9 multiplier?

I have a 3200+ I may run at 10x270 (2700) instead of 9x306 (2754) if the benchmarks look better that way. I can attain a higher CPU frequency by using the 9 multiplier. If I run much higher than 270 FSB with the 2/3 multiplier, my Ram starts to fail. I am limited to 2700 with the 10x because the memory tops out at 225 max stable and the only other divider option would give me low memory speeds. The Chaintech bios is currently limited to 1, 5/6, 2/3, and 1/2 dividers, no "working" voltage settings between 1.45 and 1.6, and integer multipliers only. I still recommend the VNF4 Ultra due to its cheap price and good overclocking regardless of these limitations. Also, I did get a Prime95 failure after 7Hr10min at 2754 (rounding error). I am pondering dropping the CPU clock another 9MHz and trying again, or is 7 Hrs good enough?

In my exploration of the best working combination, I noticed Prime95, when running a blend or long fft test, is the most overclock sensitive, followed by 3DMark05, followed by Prime95 connected to the server, and finally followed by Sissoft Sandra Lite 2005 and Aquamark (Sandra runs fine at 2790 1.4V but Prime95 locks instantly). Anyone else see this relationship?

I couldn't say as usually I run all the above together for 12 hours at least and only certify the overclocked speed if it passes without any glitches. Then and only then it is dependable enough for my taste.

 

[birdpup]

A prime95 failure shows system instability. The operating system can work around this and games do not care, but your system is still unstable. One of the Readme files in the Prime95 installation contain a line stating that the Prime95 group does not want someone submitting results if their system is unstable in even the smallest way. It is up to you but I would not accept it. I am currently on my fourth day of torture testing without errors in Prime95 and I am hesitant to stop it but I need to get some work accomplished. You can easily lower your speed by a small amount and still possess an excellent overclock.

There is the question of why the error was received. Is the system just running too fast? Is the memory overheating and requiring a fan rigged in the case? Is the power supply +12V line fluctuating? Are temperatures too high, possibly requiring a new heatsink?

 

[ToeJam13]

UPDATE:
New tests, new times, new results...

-=General Info=-
CPU Model Number: ADA3800BPBOX
CPU Code: CBBLE 0513CPBW
Stock Speed: 2400MHz
Stock HTT and Multiplier: 200MHz * 12

-=System Info=-
Motherboard: MSI K8N NEO4-F
BIOS Revision: 1.2
Memory: 2*512MB Patriot (PC3200, 2-3-2-8-1)
PS: Silverstone SST-ST460 (460W, Active PFC)
Cooling Type: AMD Retail CPU HSF

-=Default vCore=-
Speed: 2652MHz
HTT and Multiplier Used: 221 * 12
Memory: 2.5-3-2-8-1 [CPU/12]
SuperPI (2M): 75s
MemTest86: 2589MB/s

-=Stable OC Info=-
Speed: 2724MHz
HTT and Multiplier Used: 227 * 12
Memory: 2.5-3-2-8-1 [CPU/12]
SuperPI (2M): 74s
MemTest86: 2648MB/s
Actual VCore: 1.44-1.49V (fluctuates)
CPU Temp in BIOS (Idle): 43C
CPU Temp in OS (Idle): 33C
CPU Temp Under Load: 50C
+12V Reading (Idle): 11.63
+12V Reading (Load): 11.58

INFO:
MemTest86+ shows memory stable up to 232MHz on this system.
Changing my memory divider from CPU/12 to CPU/15 does nothing to increase stability
Have tried vCore as high as 1.62V with no effect
All times deemed stable after 30 minutes on Prime95 LFFT torture test

 

[bjc112]

Originally posted by: Avalon

Originally posted by: Duvie
Have you tried a 100 divider?? What divider are you running??? I have noticed in my recent testing that 200 divider works fine up to 2.44ghz (w/ default 1.4v) but withn that range I cannot hit 2.5ghz with 1.5v.....YOu better start steooing down the divider as you climb and dont be afreaid to be running 133 even if your ram wont be doing 400ddr at those HTT levels....

It was holding BJC back and I have seen it many times....forget what your ram says it will do....what vdimm have you been applying so far.....That wall is likely a memory issue....Every time I have seen someone have one like this they eventually got past was due to memory.....

Duvie, I am running a 133 divider, which puts my memory no higher than 200mhz
I actually don't even know the max of my UTT since this is the first time I've had a board that could supply it more than 3.1v to them.

My Hyper X 3200 would run 215 2/3/3/7 in my Epox 8rda ( nforce 2 ) 1:1

I cannot run anything close to that on my A64..


I can run stock 9x200 but need to jump all the way down to almost DDR 300 to acheive my stable 295x9!!

Drop to the 100 divider, just see what happens.

 

[BlingBlingArsch]

Q: i heard that the new memory controller of the Venice is able to use the AMD dual-channel feature even when equipped with unsymmetric Modules, eg with one 512mb and one 256mb stick of Ram. What u think?

 

[user1234]

It seems like Avalon has disappeared !!!! maybe the disappointment got him down, and he took the only way he knows out of the situation......RIP

 

[Gusdawg]

Update on my system...

-=General Info=-
CPU Model Number: ADA3200DAA4BP
CPU Code: LBBLE 0515APAW

-=System Info=-
Motherboard: CHAINTECH VNF4 ULTRA
BIOS Revision: 3.0 (3/15/05)
Memory: 2x512 Corsair Value Select PC3200 DDR400 2.5-3-3-8 1T
PSU: 420 Watt Turbolink LC-A420ATX, stock with ASPIRE X-Dreamer 2 case
Cooling Type, Brand, Model: Air, AMD Retail CPU HSF

-=Stable OC Info=-
Max Stable OC: 2736 MHz (42C Max Under load)
Prime95 Blend Test: Manually Halted after 28 Hours and 10 minutes
HTT and Multiplier Used: 304 * 9
Actual VCore: 1.45V (I set 1.45 cpu-z shows 1.44)
CPU Temp in BIOS (Idle): N/A
CPU Temp in OS (Idle): 30C
CPU Temp Under Load: 42C

Even though 2736 is rock solid, I plan on running 2754 since everything but Prime95 Torturetests works just fine.

The powersupply I have does undervolt quite a bit (11.52 on the 12 volt line, 2.82 on the 2.9 for ram!), The ram cant handle prime95 more than six hours over 222.5MHz at 2.9v. CPU temp isnt bad, The AMD retail HSF does the job nicely... can use better case cooling though. Under load the CPU reaches the case temperature and stabilizes (42C). I am confident I could get more out of this CPU with additional voltage options on the motherboard. I need faster ram to really take advantage of the 10x multiplier. My next task is to find the fastest working memory settings since I have to settle for 196 MHz ram speed if I want max cpu speed.

 

[bjc112]

Originally posted by: Gusdawg
Update on my system...

-=General Info=-
CPU Model Number: ADA3200DAA4BP
CPU Code: LBBLE 0515APAW

-=System Info=-
Motherboard: CHAINTECH VNF4 ULTRA
BIOS Revision: 3.0 (3/15/05)
Memory: 2x512 Corsair Value Select PC3200 DDR400 2.5-3-3-8 1T
PSU: 420 Watt Turbolink LC-A420ATX, stock with ASPIRE X-Dreamer 2 case
Cooling Type, Brand, Model: Air, AMD Retail CPU HSF

-=Stable OC Info=-
Max Stable OC: 2736 MHz (42C Max Under load)
Prime95 Blend Test: Manually Halted after 28 Hours and 10 minutes
HTT and Multiplier Used: 304 * 9
Actual VCore: 1.45V (I set 1.45 cpu-z shows 1.44)
CPU Temp in BIOS (Idle): N/A
CPU Temp in OS (Idle): 30C
CPU Temp Under Load: 42C

Even though 2736 is rock solid, I plan on running 2754 since everything but Prime95 Torturetests works just fine.

The powersupply I have does undervolt quite a bit (11.52 on the 12 volt line, 2.82 on the 2.9 for ram!), The ram cant handle prime95 more than six hours over 222.5MHz at 2.9v. CPU temp isnt bad, The AMD retail HSF does the job nicely... can use better case cooling though. Under load the CPU reaches the case temperature and stabilizes (42C). I am confident I could get more out of this CPU with additional voltage options on the motherboard. I need faster ram to really take advantage of the 10x multiplier. My next task is to find the fastest working memory settings since I have to settle for 196 MHz ram speed if I want max cpu speed.

What is your max OC with a little more voltage? bump it up to 1.5v and 1.55v.. You could easily have 2800mhz stable!

 

[Gusdawg]

Originally posted by: bjc112

What is your max OC with a little more voltage? bump it up to 1.5v and 1.55v.. You could easily have 2800mhz stable!

I cant change my Vcore to anything between 1.45 and 1.6 courtesy of a bug in the current Chaintech VNF4 Bios. If I could get access to 1.525 volts, a stable 311x9 (2799) would most likely be possible. I can run 311x9 with 1.6 volts now and have 54C for cpu temperature under load, but that load scares me a little at this point. I would likely have to live with max cpu temperatures in the upper 40s Celcius (If running 1.525), which cant be that bad since my 1600+ Athlon XP idles at 50C.

Edit: New Chaintech VNF4 Ultra BIOS coming in early June which will fix the voltage issue! - Source Chaintech Tech Support

 

[harpy82]

hav u looked at other BIOS that might help out ?

 

[Concillian]

-=General Info=-
CPU Model Number: ADA3000DAA4BP
CPU Code: LBBLE ???? ????

-=System Info=-
Motherboard: DFI LanParty UT NF4-D
Memory: Patriot 1GBx2 2-3-2-6 1T @197MHz (2560/13)
PS:Seasonic S12 430W

-=Stable OC Info=-
Max Stable OC: 2560MHz @ 1.57v (haven't gone higher yet)
Other stable OCs: 2500 @ 1.5v

-=Default Vcore=-
2420 MHz

I will edit this post with new info as it becomes available to me. Stable OC by my definition is 12+ hours of Prime95

 

[bjc112]

updating my

LBBLE 0515

http://pics.bbzzdd.com/users/bjc112/2800primestable.JPG

 

[nothingbutrich]

Well I have been lurking on these baords for a few years and decided it was time i joined.

I have a winnie 3200 and while trying to remove it from my heatsink since it was stuck i broke a pin of so now i have 938 pins. The funny thing is that my cpu still works well somewhat. MY dual channel is disabled it wont work. Also it seems like i can only use ram in no dual channel slots and only memeory that is 256 of ram or less. I always wanted an excuse to get a venice core and well thats one. Dont get me wrong i didnt do it on purpose or did I ^ ^ ? I decided to order aa 3200 venice from new egg. my 3200 from monarch was a bit disapointing. I hope i can hit at least 2-4 2-5 stable. I know i have a crappy board but hope it goes well.

 

[bjc112]

Originally posted by: blackwing
Well I have been lurking on these baords for a few years and decided it was time i joined.

I have a winnie 3200 and while trying to remove it from my heatsink since it was stuck i broke a pin of so now i have 938 pins. The funny thing is that my cpu still works well somewhat. MY dual channel is disabled it wont work. Also it seems like i can only use ram in no dual channel slots and only memeory that is 256 of ram or less. I always wanted an excuse to get a venice core and well thats one. Dont get me wrong i didnt do it on purpose or did I ^ ^ ? I decided to order aa 3200 venice from new egg. my 3200 from monarch was a bit disapointing. I hope i can hit at least 2-4 2-5 stable. I know i have a crappy board but hope it goes well.

That's a bummer dude, did it come off when ou removed the HSF?

You gotta TWIST!!!



It pulled one of my CPU's out of the socket, but didn't cause any damage at all!

 

[zShowtimez]

CBBLE 1.63v 2.85Ghz

http://users.adelphia.net/~foninferno/comp.jpg

 

[pnb263]

-=General Info=-
CPU Model Number: ada3000daa4bp
CPU Code: LBBLE 0516gpbw

-=System Info=-
Motherboard: Epox 9NDA3+
Memory: 2x512 OCZ PC3200 DDR400 3-4-3-5-2T
PSU: Antec True430 in 1000AMG
Cooling Type, Brand, Model: Air, Thermalright, X-120

-=Stable OC Info=-
Max Stable OC: 2700 (300X9) (so far, hoping for more)
HTT and Multiplier Used: 300 * 9
Actual VCore: 1.432V
CPU Temp in BIOS (Idle): 28C
CPU Temp in OS (Idle): 28C
CPU Temp Under Load: 36C

I just finished building the system and installed windows about 4 hours ago. I have been running Prime 95 since then with no lockups or errors. I am hoping to hit 3GHz on air with this chip, and looking at the voltage/temps thus far, I am happy with it. System temp and CPU temp are just about the same (CPU a couple degrees higher).

I have yet to do a bios update, and have not fully installed all my applications on this machine. I also have not tweaked the BIOS for tighter timings on the RAM. Right now I am letting it burn in to set the thermal paste. Hopefully within a week, ill see 3GHz!!! wish me luck

 

[BlingBlingArsch]

Can somebody already make save statements, if the Venice is performing better or has more oc-potential than the Winchester?