11.5 multi with 3200+ winchester

[ArcTan]

This seems like something weird that I haven't found a valid explanation for.
Note - This isn't my cpu


Motherboard used is a Gigabyte K8NSC-939
doesn't use the nf3 ultra but nf3 250gb chipset apparently
Maybe something to do with it defaulting to 11.5?

CPU-Z validation:
http://valid.x86-secret.com/show_oc?id=10591

[D_o_S]

Wow, that is interesting. Well there are two possibilities:

1) I don't know the explanation, but it is that way some how
2) You found a security bug in CPU Z.

[sandybeach]

gigabyte boards are know to cycle via cpu codes freq. I think Its jsut a bug is all

check out some of the top 3dmarkers with gigbayte boards you will see a 2600mhz cpu showing when they where really running a 3300mhz cpu, lol

[ozzimark]

it's defaulting to 11.5? did whoever has it try clearing the bios?


edit: wait. i just realized... i'm slow. haha, a 3200+ can't use an 11.5x multi

[ArcTan]

Wow, that is interesting. Well there are two possibilities:

1) I don't know the explanation, but it is that way some how
2) You found a security bug in CPU Z.

It's weird but apparently the multi is locked on his board

[Silicon Effect]

I want a decent OC'iong mobo that will do that :P

(give extra multis)

[ArcTan]

I want a decent OC'iong mobo that will do that :P

(give extra multis)

Hey SE
don't we all
I got just this from ocau and thought it was weird

[ozzimark]

well if we can figure out how amd locks their cpu's.. we can figure out why it's doing this. was it ever established that there's a small rom in the core that stores the needed information for the bios?

[Crankybugga]

hehehe weird indeed...somebody send me there spare 3500 to test in this mobo, maybe it will have a 13 multi.

[$0m#0n#]

rofl
@ArcTan can u pls try the limit of the CPU @11.5 & @10 mutli 2 c if it stop @the same clocks

[rzonealley]

totally weird but i feel i'm lucky if that clock is real and those proc are mine.
11.5 multi on 3200+ can the proc be OC then?

[M.Beier]

nice mate nice - Eventhough I perfer high FSB

But... If mem is the bottleneck in a OC, that CPU ROCKS

Best regards
12k

[Silicon Effect]

Hey SE
don't we all
I got just this from ocau and thought it was weird

HEHE yeah I seen it there too, :P OCAU/Pub is where i hang out mostlly ^_^

[perry_78]

Weird, weird indeed.

I put my money on CPU-Z bug in relation to Gigabyte motherboard.

[don_vercetti]

get your mate (or whoever owns that board) to run CPUID and any other freq. identifying stuff, see what speed they say it is. Also do some benches (e.g. sandra) to see what speeds it gets, if it's the same as a 2.3ghz Athlon64. If it is, lucky bugger.

[trance565]

does it saythe same speeds in the bios? or does the bios clock say different?

[SLaY3r07]

Yep this is very weird.

[Crankybugga]

Ok fellas...this is my chip, and from what i can gather it is in fact running at 2.3Ghz. A short story...I had this chip running on a AV8 for a few weeks playing around with some Gold VX4000 and 2x512 Bh5(KHX) combined with a booster and it was your standard 05/05 3200 with a 10x multi. One morning I slotted the BH5s in without resetting the cmos and from the time I started the PC I had no vga signal, refitted VX and still no vga. Ordered K8ns-939 and got a K8NSC-939..so I slotted the cpu in with the VX and loaded windows and stuff and was thinking this is pretty quick (a lot snappier than the AV8) installed cpuz and found the multi at 11.5.
Im not sure if the cpu is in fact unlocked or the chipset is to blame. Ive run loads of mem and cpu benchs.
The old SiSoft reports
Mem Bandwidth
11.5x200 3647/3711
11x5x226 (2.6Ghz) 4244/4257

Cpu Arith
11.5x200 6256/4656
11.5x226 6808/5193
note this is on SSoft Sandra 2002

[IMG][/IMG]



At 11.5x 226-2.6Ghz I get a 32 sec Pi . I guess im going to have to wait to RMA AV8 to find out if cpu or mobo is to blame..

Ive tried almost all programs to try and adjust the multiplier, there is very limited adjustments on this budget mobo, no divider and no multi. There is only one bios for this mobo. Its been sugested the A64 instructions are missing..not sure/dont care..also it runs out of sync due to half multi..RAM only at 192 for a HTT of 200.

[Crankybugga]

Weird, weird indeed.

I put my money on CPU-Z bug in relation to Gigabyte motherboard.

How much?

[i found nemo]

is it a counterfit chip, cuz this is odd lol and that would explain the missing a64 instructions

[Crankybugga]

is it a counterfit chip, cuz this is odd lol and that would explain the missing a64 instructions

lol I can imagine a shed full of crims making counterfiet cpus

[uwackme]

Maybe something in your system happened to "burn" "zap" etc something in the multi bridges that are there somewhere but we dont know how to FK with them yet.

You may have hit on a rosetta-stone of A64 unlocking. you'd wanna analyze what bridge or code difference there is between x10 and x11.5. Is it like ONE bit difference in the code? Suggesting one bit of a default bridge got zapped and opened? to change the code to 11.5 instead of 10.

Ive been wondering how come noone has figured out how to UNLOCK these friggin chips yet.

Im suspecting its a bridge setup but under the IHS and under the soldermask on the substrate. If we knew WHERE it was, we'd all be digging in with exacto knives ;-)

Assume they speed sort the chips BEFORE the IHS goes on. make the tweak to the lock, and slap an IHS on. Then SHIP IT.

[Crankybugga]

That sounds much along the lines ive been thinking. I wish I had another 3200 or 3500 to test on this mobo then id know either way. I wonder should I take the IHS off and have a gander...is there any pics of a A64 without IHS on the web?

[Mats]

I wonder should I take the IHS off and have a gander...is there any pics of a A64 without IHS on the web?

I bet testing another mobo and CPU would be more intersting...
Don't you know somebody with a S939 machine?
Have you tried CrystalCPUID for changing multiplier?

[DeNs]

I'm just gonna be a skeptic and say it's a bug. Strange though. I've seen MSI's corecenter playing funny buggers with what my CPU frequency my 3200+ is being reported as, as well as ThrottleWatch on my P4 when minimised fluctuates randomly when showing the CPU frequency.

I think i'll just put it down to being a bug.

---dens