Holy TOLEDO! - Here's my x2 4600+

**[XC] moddolicous** · 05-29-2005, 04:23 PM

wow, tese chips are amazing. Cant wait for OPB to get this and reach some insane speeds. He could probably break 30k in sandra arithmetic.

**s7e9h3n** · 05-29-2005, 04:29 PM

Originally Posted by eva2000

so lovely... just noticed smartguardian says 12.41v on 12v rail

How high you set your psu's 12V rail via multi-meter? as on my DFI NF4 SLI-D, 12.09v in smartguardian is already 12.35-12.41v via multi-meter! You running 12.6-12.8v ?

what psu ya using ?

You know what, I really don't remember what it measured @ the molex, but I know for sure that there wasn't as much of a discrepancy as with your setup. My psu? Hehehe....it's a modded Zippy 700W that's capable of giving me 12.9+V on the 12v with maybe at most a .02V drop while OC'd

:

Ok yesterday was my day to play with the 11x multiplier. This cpu's OK, I guess

:

**s7e9h3n** · 05-29-2005, 04:41 PM

So this was basically all I've had time to do and most likely I'm not going to do too much more testing with somebody else's cpu

. There was, however, a couple of issues with the cpu and compatibility with a couple of progs. Clockgen didn't allow the changing of Vcore in windows. It could have made OC this cpu quite a horrible experience if it wasn't for the fact that it ran solid at whatever I set it at. Also, for some reason, it had the pcie speed clocked at ~88Mhz sporatically for no apparent reason. Another program which I had problems with was Sisoft's Sandra2005. Most of the issues came from the Memory Bandwidth Benchmark. I can only recall completing the benchmark only 2 times. Every other time I ran it, there was either an error in windows, a complete lockup, or a bsod. I've never seen this happen before where I could complete a spi1m run, but couldn't complete Sandra's Memory Bandwidth Bench. I think sisoft may need an update for the program to function correctly with the x2 (or maybe I just haven't dl'd it yet

) Here's a screenie - check out cpuz's speed and what sandra sees it at:

**sr4470** · 05-29-2005, 04:47 PM

Voltage changing in clockgen doesnt work for me either, so i dont think thats dc specific

**s7e9h3n** · 05-29-2005, 04:54 PM

Originally Posted by moddolicous

wow, tese chips are amazing. Cant wait for OPB to get this and reach some insane speeds. He could probably break 30k in sandra arithmetic.

I wouldn't doubt it. This cpu's got a lot of headroom left and OPB's someone who knows how to sqeeze every last mhz outta a cpu. Here's just a glimpse of the possibilities with this chip:

1.488Vcore! I tried bumping the vcore to complete spi1m, but only got through 17 iterations. I'd love to mess with it a little more, but I'm gonna end up taking it out soon and packing it up for opb. I'm sure we're gonna see some wild results from him

**eva2000** · 05-29-2005, 04:56 PM

Originally Posted by s7e9h3n

You know what, I really don't remember what it measured @ the molex, but I know for sure that there wasn't as much of a discrepancy as with your setup. My psu? Hehehe....it's a modded Zippy 700W that's capable of giving me 12.9+V on the 12v with maybe at most a .02V drop while OC'd

:
http://img255.echo.cx/img255/5131/zippy12v2mz.jpg

care to measure it right now ?

really interested to know what the diff between your smartguardian reported 12v rail and multi-meter

**s7e9h3n** · 05-29-2005, 04:57 PM

Originally Posted by sr4470

Voltage changing in clockgen doesnt work for me either, so i dont think thats dc specific

Is it unavailable? Or just simply doesn't work? Here what it looks like:

**s7e9h3n** · 05-29-2005, 04:59 PM

Originally Posted by eva2000

care to measure it right now ?

really interested to know what the diff between your smartguardian reported 12v rail and multi-meter

I would have done that for you earlier, but right now I'm using a backup psu (antec neo) since I took my zippy out to redo the paint job

I'll re-install it later and let you know. Sorry....

**eva2000** · 05-29-2005, 05:00 PM

try with VID = AUTO then use clockgen ?

no probs about psu but let me know.. seems most high clocking venice/sandy users have very high 12v rails set

**s7e9h3n** · 05-29-2005, 05:05 PM

Originally Posted by eva2000

try with VID = AUTO then use clockgen ?

Of course I've tried that!

along with every other bios setting available.

**eva2000** · 05-29-2005, 05:07 PM

maybe try different bios ?

**s7e9h3n** · 05-29-2005, 05:14 PM

Originally Posted by eva2000

maybe try different bios ?

That could work, but then again, clockgen was working fine with my 3200+ Venice adn same bios. BTW it's 510-3 don't think I could find anything more current

**althes** · 05-29-2005, 05:24 PM

Cant wait for opb to get his hands on that sucker

**eva2000** · 05-29-2005, 05:30 PM

Originally Posted by s7e9h3n

That could work, but then again, clockgen was working fine with my 3200+ Venice adn same bios. BTW it's 510-3 don't think I could find anything more current

With the cpus you get access to, you should join DFI's bios beta testing team and get the latest bioses to test with heh

**sr4470** · 05-29-2005, 05:32 PM

Originally Posted by s7e9h3n

Is it unavailable? Or just simply doesn't work? Here what it looks like:

Shows up, but has no effect when i select it

Edit: i'm still on 414-3 bios, if thats any help..

**s7e9h3n** · 05-29-2005, 05:39 PM

Originally Posted by sr4470

Shows up, but has no effect when i select it

Edit: i'm still on 414-3 bios, if thats any help..

Ok, I didn't think that it was the same prob

You should take eva's advice on this one.....your vid isn't set to auto in bios......

**s7e9h3n** · 05-29-2005, 05:40 PM

Originally Posted by althes

Cant wait for opb to get his hands on that sucker

Sorry I wan't good enough for you

**s7e9h3n** · 05-29-2005, 05:41 PM

Originally Posted by eva2000

With the cpus you get access to, you should join DFI's bios beta testing team and get the latest bioses to test with heh

I dunno about that. Weird bsod's and reboots aren't exactly what I strive for when testing

**sr4470** · 05-29-2005, 05:47 PM

Originally Posted by s7e9h3n

I dunno about that. Weird bsod's and reboots aren't exactly what I strive for when testing

The 1-25 bios was terrible, cant believe my mobo shipped with that in april

Edit: anyway back on topic, when are we likely to see these dc in significant numbers?

**s7e9h3n** · 05-29-2005, 05:59 PM

Originally Posted by sr4470

The 1-25 bios was terrible, cant believe my mobo shipped with that in april

Edit: anyway back on topic, when are we likely to see these dc in significant numbers?

Of course, they're already around, just not available to the general public. Only OEMS ATM, but expect to see them soon....

**sr4470** · 05-29-2005, 06:13 PM

Gonna have to wait a while for the 4200\4400 to come down to a reasonable price, however i hear it will be between 4000+ and FX prices in the UK (Â£410?)

**~~Onepagebook~~** · 05-29-2005, 07:35 PM

Originally Posted by s7e9h3n

I wouldn't doubt it. This cpu's got a lot of headroom left and OPB's someone who knows how to sqeeze every last mhz outta a cpu. Here's just a glimpse of the possibilities with this chip:
....
1.488Vcore! I tried bumping the vcore to complete spi1m, but only got through 17 iterations. I'd love to mess with it a little more, but I'm gonna end up taking it out soon and packing it up for opb. I'm sure we're gonna see some wild results from him

ha, ya, I will do my best to get her fly, thanks buddy, you did a great job.

..I am out of twon, will back home tomorrow.

**s7e9h3n** · 05-30-2005, 03:13 PM

OK, it's a foregone conclusion.....remember you heard it here first

OPB will be at the top of PCMARK04 orb in less than a week . Here' my final bench on this cpu with no tweaks or anything of that nature. Just my plain old 6600gt @ 561/1160 or something like that.(I have no idea why the grammar check doesn't compete since I know nothing about this benchmark). This is a comparison to Sampsa and Kamu's #1 score in PCMark2004 done with their x2 4800 on LN2 and an x850xtpe I believe (the 4600 is the one on top):
Oh yeah, and I'm running the cpu @ 3190 - 11x290 vs their 4800 @ 3223. This cpu is barely breaking a sweat at this speed.....