A64 OCs (Air) Poll

[imported_Slider]

I finally figured out that my vnf4ultra doesn't actually apply the cpu voltage until I put it at 1.6. Now I'm at 2.4ghz. I think I'll keep it this way for a while.

 

[benrico]

My 3000 made it to 2.2 @1.52, 245 * 9, dfi nf4 ultra, xp 120. 0503.

 

[Pollock]

Well, I originally set mine to 9*270 for 2430MHz and've been too lazy to try any more since then. Corecell is reporting slightly above default voltages too (where I set it to much higher), which may be a good thing (well, it means my PSU sucks, which it does, but it might mean my chip is pretty good).

 

[thriemus]

Bear in mind that too much vcore is worse for not only stability but the life of the processor.

vcore increases should only be applied in small increments to ad stability, not to pust the MHz up to far, unless you plan to have the cpu cook itself within the warranty time to a new replacement

 

[MrControversial]

Okay, I just set my stuff up. I currently hit 2.4 Prime stable last night with 3xHTT, 240HTT, @ 1.56V. Idle temp is 32C and load is 33C. The memory was running at 160MHz and 2.7V. Currently, I'm priming it at 4xHTT and 250HTT.

If you're on a VNF4 Ultra I suggest you bump your Vcore up to 1.6 unless it won't change. Don't worry about it being too high because it's really 1.56V which is safe. You might also want to take your HT off of STARTUP and change it to 3 or 4. If you're using value RAM, divide it for a higher HTT.

 

[Duvie]

bump....

 

[Duvie]

bump...

 

[Duvie]

With 129 votes the curve on the graph speaks volumes.....I wish I could have correlated it with week numbers but that would have been a whole lot of threads....

I wonder if setting up like 4 polls with something like this...

week 35-39
week 40-44
week 45-49
week 50-04

has anyone seen week 5's yet for 2005???

 

[Duvie]

Hey bump for sentential.....look at the numbers buddy...still think 2%???

 

[sangyup81]

did he mean 2% at any voltage?

 

[Duvie]

As far as I know it was asked to him on several occasions in the thread he was participating in and he was not talking about default vcore alone.....He was meaning topping out on air cooling....

If he did mean that then he would be closer...He had ample time to make that point when ppl argued with him about the numbers....like I said I dont remember him ever going that way with his argument. He really just felt most were not getting to 2.6ghz period with the iwinchester 3000+ to 3500+

 

[DrMrLordX]

Originally posted by: Insomniak

Originally posted by: Duvie
Clock for the clock the sempron is a beast...just look past those damn pr rating BS....a 2.5ghz sempron appears to run close to a 2.5ghz A64 in cpu intensive apps...the only plavce it may lose and lose only slightly will be in apps that can take advanatge of 512 or even 1mb of L2 cache and the 3-5% dual channle can give in just the right apps....

Well it should be - all it is is Athlon XP renamed.

No, it isn't. He's referring to the Socket 754 Semprons. I'm running one right now. The 90 nm ones are based on the Palermo core, have no x86-64 functionality, lack FID and Cool n' Quiet in the 2600+ and 2800+ models, and have anywhere from 128k to 256k l2 cache as follows:

2600+, 3000+: 128k
2800+, new 3100+: 256k

There is a "new" 3100+ floating around based on Palermo, or so I have heard, but I have not seen one used/tested myself.

They apparently use SS and SOI, and have a stock vcore of 1.4v.

The 2800+ looks like a good overclocker, and a great one if you stick it in a DFI board.

 

[Acanthus]

Originally posted by: Duvie

Originally posted by: teutonicknight
IDLE: 30c
LOAD: 43c

3200 @ 3.5 GHz

No water cooling or vapo/prometia.....I SAID AIR!!!!

His temps would be far lower on either. Hes just flat out lieing, or its a typo and its 2.5ghz.

Edit: confirmed typo didnt read all the way down before posting.

 

[MTSteel]

A643000 0451 @ 2.25Gig Prime for 13 hours before I stopped it to try some higher clocks. Prime will not run for more than a minute at 260x9 with Vcore at 1.6. I was using HTTx3 and the 100 memory divider to really isolate the RAM while going over 250
The board has no troubles going over 300, so it appears to be my CPU. Such is my luck

 

[edge929]

A64 3000+ 0504 @ 2.25 250x9 @ 1.55VCore

FAILS Prime in less than a minute, RAM @ 100 to isolate CPU max, HTTx3. My RAM is Corsair C254400 which can run 1:1 all day long even if my chip COULD do 2.6

Another $hitty chip.

EDIT: After finding some forums on faulty VCore settings on the Neo2 I was able to hit 2.475 @ 1.65VCore stable according to everything BUT Prime. Still fails within a minute. Heard a few rumors that others with the C254400 have the same problem with Prime. Need to dig more.

 

[imported_Kensai]

Mmmbop. Gogo. Nice percentage to over 2.6GHz.

 

[CP5670]

3200, week 15, 2004
stable at 2250mhz (250x9), 5/6 mem divider, 3x LDT, 1.425V; doesn't go much further
uses XP90 with panaflo 92 H1BX

Preproduction model, seems to only support 800mhz hypertransport. It's a crap overclocker, but I got it free so I'm happy anyway.

 

[Dadofamunky]

Well, I just "sidegraded" to a 3000+ Winnie from a 3500+ Newcastle, and I'm running much cooler - 10C cooler. Sig has my ratings. The benchies are pretty much exactly the same. The funny thing is, I got the exact same ocverclocking result of 2.52 GHz on air. Only difference is the temps and I got a 700 MHz boost in clock speed from the Winnie, which is an 0450 CBBHD. I had only gotten about 320 MHz of OC from the Newcastle, which is now going to be sold while it's still worth something.

 

[sciwizam]

A Newcastle 3400+ (2.4Ghz) at 2.6Ghz (217 * 12). Taking things slow, as I'm new to overclocking. Temps are ~37-38C idle and max out at 58C under full load. The system is pretty stable as I run DC projects and haven't seen any isssues.

 

[fstime]

Cant wait to get some vapo chill on the new FX-57. Should be great with VX. Hopefully it doesnt have the coldbug like winchesters......