2003 week last unlocked Barton was manufactured

[Fleck]

I don't even really care about having a locked processor... I'm just happy at least I can play with the front side bus, and who knows, maybe I'll be lucky enough to push mine over 400.

 

[masitti]

You guys don't get POST when you change the multi? That's weird... I could post with whatever I changed the multi (even though it didn't do anything). Am I missing something here?

 

[Fleck]

Maybe you're missing a crappy motherboard.

 

[masitti]

Maybe you're missing a crappy motherboard.

Under my impression the NF7 is a good board, maybe it's just me though. Do I have to enable something on the mobo to change the multi, or is the BIOS option alone good enough?

 

[mw521]

Got this Barton today, xp2800, unlocked even. Im runnin an nf7-s version 2 board. Waitin on my thermalright to get here, this is with stock sink on it though. Time for some fun.

 

[Vaguerant]

With my locked 2600+ Barton any lower multi than 11.5 results in a no post, and anything higher just posts at 11.5

 

[Veeb0rg]

UPDATE: PQZEA 0332MPMW

Locked. I tried the 9, 9.5, 10 and 11.5 multipliers. All of them resulted in no post. I thought it was suspicious that such a new stepping had such an old date code, particularly when this was a fresh batch just shipped in to the computer store, with late date codes (0342, 0347), I thought it was strange that such an old one was in there. Maybe AMD threw that date code on to try and trick us

Anyways, yah, PQZEA 0332MPMW Locked.

Edit: It's running 200x11 @ 1.75v stable though, which isn't too bad. It's fast enough for my friend's computer anyways.

well then I guess i'll be hoping that my dads barton is unlocked *he got it like 6 months ago* and i'll just swap with him.

 

[Audioaficionado]

well then I guess i'll be hoping that my dads barton is unlocked *he got it like 6 months ago* and i'll just swap with him.

One of the newer locked Bartons may still out perform the older unlocked one.

 

[tom10167]

^^True that, some locked Bartons are running at higher FSBS than mine, with a lower multiplier!

 

[tom10167]

^^True that, some locked Bartons are running at higher FSBS than mine, with a lower multiplier!

 

[Veeb0rg]

whoa.. new years musta been hard on me.. i'm seein double..

I'll run the chip and see how it overclocks.. if it doesn't do it well then i'll swap em..

glad i sprung for the lvl2 muskin memory sticks..

 

[Fleck]

I have a question for you guys... what about turning down the memory frequency to 83% or 80% of the processor frequency on the EPoX 8RDA3+ to get higher overclocks on the CPU if the memory is too weak to go much beyond 200x2? Why isn't this more popular? Would this bump down to async and not function too well? Is it just something people avoid doing? I figure if I'm looking to go from 2.3GHz to 2.5GHz by taking my processor bus to 220 I wouldn't mind keeping the RAM at 183MHz...

 

[The Coolest]

I've been thinking of this as well. Also using cheap PC3200 memory made by samsung for example, I know my friend's rig runs just fine at 210MHz FSB on 1:1 ratio, with really relaxed timings, it should still perform better than lower memory speed w/ tighter timings.
One more thing, theoretically you should be only limited to the lower RAM b/w but the loss of performance shouldn't be more than that.

 

[Audioaficionado]

I've been thinking of this as well. Also using cheap PC3200 memory made by samsung for example, I know my friend's rig runs just fine at 210MHz FSB on 1:1 ratio, with really relaxed timings, it should still perform better than lower memory speed w/ tighter timings.
One more thing, theoretically you should be only limited to the lower RAM b/w but the loss of performance shouldn't be more than that.

Running asynchronously might make higher fsb possible if you have run of the mill RAM but ideally you should evenutaly get better RAM and then run 1:1 again.

My 2¢

 

[c627627]

Regardless of benchmarks, 100% in-sync will do better in real-world performance.

I did tests on that a long time ago here:
http://www.ocforums.com/vb/vb/showthread.php?s=&threadid=215500&perpage=50&pagenumber=2

 

[The Coolest]

yes, you lost some performance, which wasn't THAT much anyway, but think of the CPU speed that you will gain. That will pay off better than running 1:1 and having much lower CPU speed.

 

[c627627]

Graphic67 wrote on 01-01-04 01:46 PM:
Since both nForce and KTxxx chipsets handle memory functions away from the cpu, I do not think that there is much to be gained by forcing a system to run in-synch. One possible advantage to synchronous fsb and memory is that the northbridges may not REALLY be as stable as they should be when running in that mode. I expect that trying to have portions of the same chip operate at different speeds while being able to maintain a stable and coherent data link is unnecessarily problematic.

Since every board - cpu - ram combination will behave differently under stressful conditions, I have to say that it is best to run synchronous timings in most cases, but that on a case-by-case basis improvements can be gained in OVERALL system performance by allowing either the cpu (fsb) or ram to clock to its full potential.

Short answer: If you gain performance and do not loose stability, asynchronous is alright.

 

[mR. STuPiD]

The Coolest, you have to run 1.75 vcore just to reach 3200+ speed? My AQZEA 0341 runs at 3200+ at 1.65, the default voltage. I might push it up a bit because right now I'm not prime stable, only error free for 3 hours, but that might be because of my ram or mobo. Anywho, my point is that 1.75 is high for just 3200+ speeds.

 

[kayl]

man those new bartons seem to run around 3-5oc cooler
well that is comparing my week 0330 to the 0343

 

[Fleck]

I've figured out a way to put that memory speed % feature on the EPOX board to use. I only have Geil Value Series PC3200 memory. I've heard this RAM will do 210 or 220 for some people, but it's not likely. What I can do is set the memory to 80% the whole time I'm testing my CPU's top speed so that no matter what I'm setting the CPU's FSB to, the Geil won't be tested at above 200MHz, not even if it were possible to get the CPU to 11.5*250 (LOL riiight). Then I guess if the difference between the Geil RAM and better RAM is 200MHz or more of CPU speed when synced 100%, I'll defenitely be getting some PC3500 later on.