Revision E OC discussion thread

ashmedai said:
Oh yeah, and on the 3000+? Don't count on overclocking your memory much. Getting the CPU maxed out and stable at that high a HT is pretty much all it's going to do. I could get the memory up to about 245 for a little bit at a time, but mostly it doesn't want to go very far over 200 once you're in the neighborhood of 300 HT.
Click to expand...

Yeah well who really cares? :confused:

a $160 proc at 2.8ghz!!! I mean jeezus christ, that's faster in almost ALL applications than a stock FX-55. Essentially this means as an overclocker / dork we have the ability to purchase the performance of a $1000.00 proc at $160 or so.

I guess you can always want more though. Don't we all? ;)
 
HardKard said:
Apologies if posting this here disturbs someone . . . seemed the most logical place.

Just got a 3700+ San Diego. Will be putting it in an SN25P Shuttle box next week and doing a mild air overclock.

Was checking the OPN numbers and have noticed that mine are a bit off from other 3700+ SDs:

1st line: ADA3700DAA5BN
2nd line: CABGE0515SPMW

This last part four letters of the second line (SPMW) are what is curious; all other 3700SDs I've seen have "VPMW" in the last part of the second line.

It is a retail box, bought it from Monarch, the seals are intact, yes I read it right (looking *again* now to be sure), and, yes, I'm wearing pants.

Normal? Point of concern?

-hk

Off topic - what is the last, third line? Just a serial number, or are there nuggets of info contained in that one too?
Click to expand...


i hope this answers your question
First Digit is either R through to X for Monday to Sunday or A through to G for same. M means mixed assembly lot.
P merely means Penang
3rd digit is the sequential batch of cores used in assembly - so if they are using the first batch for that day its an A second batch is B and so forth. M means the cores used in the assembly came from different batches.
The existence of a W for the 4th letter means mixing of wafer lots in the core fabrication was prohibited. Its absence means the wafer lots could be mixed when making a batch of cores.
0450 etc is self explanatory and is the week of assembly.
Click to expand...

basically the code after the date is simply which day and batch it's from. :D
it seems like the stepping is most important. i have a CAGBE newark that does 2.8ghz on air without much effort.
 
. . . great, thanks for the beta !!

P for Penang . . . Penag, Malaysia . . . had to ponder that for a sec

Bottom line, simple day/style of production information.

Thanks, Eclipse. I'll throw up the OC on this once it goes down . . .

-hk
 
NICE!
"unknown mods"

hahaha, thanks for letting me know, it seems that there are quite a few people with neo2's that are having slight issues with venice/san diego.. might just be me imagining things though :p
 
(cf)Eclipse said:
NICE!
"unknown mods"

hahaha, thanks for letting me know, it seems that there are quite a few people with neo2's that are having slight issues with venice/san diego.. might just be me imagining things though :p
Click to expand...
pft msi haha....thats why DFI is DA BOMB...honestly dfi is now my new favorite mobo manufacturer. Ive never had such a rock solid board thats also a badass overclocker, well except for my old p4c800e-dlx it was the only board that could properly handle my prescott at 4ghz+...wow im rambling now
 
J-Mag said:
I guess you can always want more though. Don't we all? ;)
Click to expand...

Would have been nice if it had Pacifica so I could play Civ3 while I watch video...for some reason that's coming out after dual core, on dual core, only on the new socket. Doesn't matter that much, when I feel like it again I have a enough guts sitting around to build another (albeit much slower) system around and slave it. Then I can play more than just Civ3. Good game and all, but it's one of the very few that will play nice with more than one active monitor.

But yeah...$149 shipped for a 2.8GHz A64...kinda hard to beat the [price/power].
 
Viper87227 said:
Nice.... I didn't realise newarks were 1MB cache. What board is that in?
Click to expand...
read all about it.. here :D
the msi k8n neo tops out sooner, due to the extreme voltage limitation :(
 
peterhoang2002 said:
will 1.7 volts with 60c load kill a 3000+ venice? this is with a zalman 7000B Al-Cu cooler.
Click to expand...
holy crapolie you need to back down the friggin voltage or buy a watercooling system
 
peterhoang2002 said:
will 1.7 volts with 60c load kill a 3000+ venice? this is with a zalman 7000B Al-Cu cooler.
Click to expand...


Wow, what is your ambient? I ran my 3000+ at 1.75v 2.7ghz full load was around 54C with an ambient of 24C.
 
peterhoang2002 said:
will 1.7 volts with 60c load kill a 3000+ venice? this is with a zalman 7000B Al-Cu cooler.
Click to expand...

WTF?? Okay either Zalman's not as good as I thought they are, or Thermalright's better than I thought it was, or your thermal paste wasn't applied very well. Possible other explainations but I'm sticking with those unless something else pops up.

I'm just browsing (15ish tabs open) plus a buncha crap in the system tray none of which uses terribly much power. Smartguardian says 26C and 1.56V right now, this is with a XP-90 on a 2.7GHz Venice 3000+.
 
I think im going to modify my cooling system so maybe i can get a higher oc. Im going to remove the pumps from the exos resevoir and the install a good pump inline somewhere inside the pc case. does anyone have a good suggestion for a good pump that can run off the psu and has plenty of head to push through 1/4in tubing
 
Try here: www.dangerden.com

If memory serves they have a few 12V DC options, and lots of information on the pumps they sell. Home Depot is a great place to get 1/2 to 1/4 fittings for the pump.
 
is anyone else running their venice on a neo2? i can get into windows @ 2.8ghz on stock voltage, but its not stable and nothing i change in the bios voltage wise for the cpu seems to take affect(ie i set the cpu voltage to 1.5v +10% and cpuz and everest still show 1.4v)

ive got the latest bios out from msi, dated 4/12.

shes prime stable @ 240x11 1:1 and game/benchable at 250x11 1:1, but not prime stable. im sure if i could up the vcore, i could get it prime stable at 250x11 1:1.
 
That is odd... I have no idea, never not had bios voltage settings not work....
 
it happened on my k8n neo when i tried the newark in there. i could only change the % over vid, the actual VID setting wouldn't do anything. thus i was limited to ~1.48v actual :(
 
i went back and changed the % over VID value, that gave me a bump in teh voltage. still trying to fiddle with the % to get a happy medium in the voltage. ive got a stable 2750mhz at about 1.55v..could prolly do it on 1.525 or even 1.5, gonna see if i can hit one of those numbers. only bad news now is that while idle temps are a reasonable 40, load temp is bouncing on the 59-60c spot...touch warm for me. may have to go on and get an xp-90c or an xp-120.
 
Baldrik, mine are about 26-40 with an XP-90. Might not need the 90c or 120 to get significant improvement. Using that Gigabyte cooler from your sig?
 
yea, im using the gigabyte one. its not bad, but its not the best either ;) i didnt like the 60c temps so i brought her back down to stock voltages and dropped the fsb to 240...she primed all night long at 2640mhz on 1.4v(temps are 34c idle to 50c load). gonna work my way back up from 240 to 250 with the stock voltages, see where she fails...maybe i can hit 2.7ghz.
 
Yeah...50 or 60C on a Venice is just...something is REALLY wrong with the cooling if you hit that even under prolonged load. Don't suppose you've got 90F ambients with high humidity or something?
 
it was 80F in my apt all day, even with the ac running non-stop..im sure that didnt help. suppose i could do like we did in highschool and set an oscillating fan and a bowl of ice blowin in on the case =D
 
GodsMadClown said:
High humidity would actually enhance HS performance maginally.
Click to expand...

Doh. You're right, since the ambient temp is lower than whatever the CPU will be throwing out in the first place, it just means there's "more" of the lower ambient temp even if it's relatively higher than normal. Then again it also makes it harder to lower the ambient temp artificially, and it's still higher than is exactly good for the computer. Or the user.
 
mikelz85 said:
just an idea, but lets say CPU X does 2.6ghz at stock, but CPU Y only does 2.5
Is it still possible that with more juice CPU Y will scale higher in clock speed than X with more juice?

If the chip that goes the furthest with stock voltage is always the best chip
Click to expand...

it's a good idea, however.. stock voltage doesn't mean much. i've seen cpu's do 2.7ghz at stock voltage, just to top out at 2.8ghz with more voltage, whereas my own newark gets around 2.6ghz with stock volts, but ends up topping out at 2.9ghz

though, removing a variable somewhere would be very nice. i thought about everyone finding their max oc's at 1.4, 1.5 and 1.6v, however.. this would take a long time, and a lot of people wouldn't post, which is the opposite of what i'm trying to do at this point
 
I think stock speed is legitimate baseline info for the overclock thread but, no matter. I hesitate to push my voltages above 1.54 (or 1.57 that's part of the problem I don't know the real voltage). Even if I can get stable at 2.9g @ 1.65+ I'm not sure I want to risk running 24/7 at that voltage on a 90mm chip.
 
Here's my post in the official Rev E OC thread.

I'd like to see more info on ram settings on these OC's. I may be wrong, but I've had higher benchmarks,faster Super Pi times w/higher memory speeds @ a lower cpu speed, than a lower mem speed @ higher cpu speed.
Case in point, 2727 Mhz ram @ 192 was a full second slower in 1M super pi, and 700 bungholio marks less in 3dmark 2k1, than 2603 Mhz @ 236 ram speed.
I juiced vcore up to 1.425 and vdimm to 2.75 and currently testing cpu @ 2628, htt @ 292, ram @ 239 (166 divider) 1T 2-3-3-6 timings. Prime95 is chugging along nicely for 9 hrs, w/ folding@home in the background.....we'll see where how it goes by morning and push on.

For those of you with a neo2 platinum..... a few hints and observations:
5x htt just doesnt work (if your not OC'ing and using 200fsb)
If you have a SATA hard drive...plug in channels 3 or 4, otherwise your oc will go nowhere over 230-240 htt.
For me, using a 5 or 6 cpu multi limited my htt to 277.....set to 7x or above before adding more vcore.
I've been using the 1.8 bios without too many issues to be concerned about.
Vcore problems.....for awhile I couldn't post @ 1.42v....cleared cmos and voltage increases are taking now (hopefully this is an issue msi is working on)
Other than these few quirks, I'm quite pleased with the neo2 plat.

Lastly, I'm glad I held out til the venice core came out. These things are oc'ing like crazy, they're cheap...and very fast. I did notice that my mobile 2500+ @ 2.4Ghz was faster than the 3000+ @ 1.8 Ghz...but @ 2.6 i shaved 14 seconds off super pi 1M, and gained 5,000 pts in 3dmark2k1
 
SpoogeMonkey said:
I'd like to see more info on ram settings on these OC's. I may be wrong, but I've had higher benchmarks,faster Super Pi times w/higher memory speeds @ a lower cpu speed, than a lower mem speed @ higher cpu speed.
Click to expand...
superPI is one of those things that's extremely mhz and latency dependant ;)

awesome post otherwise, i like how you included the quirks you found :p ;)
 
Are there temp reporting issues with the Venice? I got my 3200+ a few days ago and at the same time installed a new XP90. When I booted up it was showing 55c idle. Needless to say, I was dropping a load in my pants. I pulled the board back out, pulled the XP90 off to check it and it pulled the CPU out bending some pins. :eek: and I dropped a bigger load in my pants. I was able to get it back in without breaking any pins fortunately. The Artic Silver was spread out nicely though so I don't think it was the HS installation.

I got the board back in and it was showing the same temps. I decided screw it, that couldn't be right and I bought a Venice to overclock (and I still had my 3000+ winny just in case). I got it stable at 2.77 GHz at 1.55v. Ran small FFT's in Prime 95 overnight for about 8 hours with load temps of about 80c. Prime 95 didn't error and the XP90 was only slightly warm. It could just be a problem with this Soltek board, but I was also reading through a Chaintech forum and others are having similar problems with the Venice. Chaintech said they are working on a new bios to help with the problem.
 
I was following eclipses guide to OC the amd64 and said If you didn't get over 250mhz HT something was wrong, well I got to 231 and it crashed as soon as I hit set, so umm what's wrong? lol

I set the HTT x2
and the cpu to x5

and if I understand correctly that would make a divider of 1:2

I'm running Corsair Value Select ram pc3200 with timings of 2.5-3-3-8


Oh I have the MSI k8n Neo2 with a 3000+ Venice
 
sata, move it to the ports closest to the cpu (3+4)

also, try 2.5x or 3x ht link multi, as there are sometimes issues with the lowest multi's :(
 
hadn't made the move to Sata yet, so I turned both of those off in BIOS, I'll try the 3x multi as I have read that 2.5 is quirky with the 64's or something to that affect
 
You know looking at cpu-z it shows my multiplier being x5.0 did I change something wrong :confused:
 
Ok so on the Clocks it has this

Core Speed 1030.9
Multiplier x 5.0
HTT 206.2Mhz

I changed the cpu multiplier as your guide said to x5

On the MSI Core Center it shows 211 x 5.0 @ 1057 mhz

And my Temperature running at 48*C

I double checked to make sure Cool and quiet was disabled

Oh I should update my bios too huh, hrmmm when i checked it said 1.3
 
i kinda fail to see the problem then, just follow through with the guide, once you get to the end, everything will fall into place.
 
well what I failed to do was go into DRAM and set the divider to 100mhz :rolleyes:
 
hehe, i understand now.. it's cool man ;)

it's like me doing some ram benching and trying to figure out why my unbuffered bandwidth is so low. turns out i forgot to enable CPC (aka 1T) :D :rolleyes: :p
 
You must log in or register to reply here.
Back
Top