Post your DFI Lanparty UT nF3 250Gb motherboard results

it's an awesome board!...


that's just sitting on my floor in the box :(


waiting on my 3000+ to show up monday.
 
Well, unfortunately I'm not all that lucky compared to the rest of the guys here.

I know I have a lot of testing still to do, but the preliminary results do not look good.

First of all, I know my ram isn't all that great. Have some cheaper OCZ PC3200, and it will do 225 Mhz @ 2.5:3:3:6 timings, which isn't all that bad. Had to bring VDim up to 1.7V to achieve that.

But the part I'm worried about is my processor. It's a 3000+ newcastle (CO Stepping). Originally it was fine when I brought it up with the memory to 225. I dropped the multiplier down to 5:4 and kept going with the FSB. I had to up the voltage to 1.525V to get it past 230 HT (still no increase in temps). Anything past 238 freezes it, and prime95 fails at anything past 235, even though the temps held around 55c (as reported by the smartguardian tool). I'm not too happy with increasing the voltage any higher without additional cooling, but with those temps, I'd hope that it could handle a little more than I'm getting.

Any help would be appreciated. (specs are below).

A64 3000+ Newcastle (CO stepping) @ 2.35GHz (235HT)
DFI Lanparty UT nF3 250Gb
2X512MB OCZ PC3200 (5:4 ratio @ 195Mhz 2.5:3:3:6)
ATi 8500 (soon to be X800XT PE)

Stock AMD fan, + 2 case fans. Temps are 43c idle, 55c load (as reported by smartguardian, which apparantly this board reads a bit high, but I have no way of telling).
CPU V = 1.525
VDIMM = 2.7
 
intercollector said:
Well, unfortunately I'm not all that lucky compared to the rest of the guys here.

I know I have a lot of testing still to do, but the preliminary results do not look good.

First of all, I know my ram isn't all that great. Have some cheaper OCZ PC3200, and it will do 225 Mhz @ 2.5:3:3:6 timings, which isn't all that bad. Had to bring VDim up to 1.7V to achieve that.

But the part I'm worried about is my processor. It's a 3000+ newcastle (CO Stepping). Originally it was fine when I brought it up with the memory to 225. I dropped the multiplier down to 5:4 and kept going with the FSB. I had to up the voltage to 1.525V to get it past 230 HT (still no increase in temps). Anything past 238 freezes it, and prime95 fails at anything past 235, even though the temps held around 55c (as reported by the smartguardian tool). I'm not too happy with increasing the voltage any higher without additional cooling, but with those temps, I'd hope that it could handle a little more than I'm getting.

Any help would be appreciated. (specs are below).

A64 3000+ Newcastle (CO stepping) @ 2.35GHz (235HT)
DFI Lanparty UT nF3 250Gb
2X512MB OCZ PC3200 (5:4 ratio @ 195Mhz 2.5:3:3:6)
ATi 8500 (soon to be X800XT PE)

Stock AMD fan, + 2 case fans. Temps are 43c idle, 55c load (as reported by smartguardian, which apparantly this board reads a bit high, but I have no way of telling).
CPU V = 1.525
VDIMM = 2.7
Click to expand...

decrease your processor stepping. I have a 3200+ newcastle core with default stepping of 11 x 200 mhz HTT = 2200 mhz or 2.2ghz. By lowering the stepping to 8 I can increase my HTT to 300 easily and put my memory at a 3:4 ratio so it tops out at 250.

Also, decrease your LDT from 4 to 3. This will give you the headroom you need to go higher (even if you don't want to lower your stepping on the proc).

Give it a try, these boards like being pushed.
 
he means lower your multiplier. You cant decrease your "stepping" ... lmfao.

this is the funniest noob mistake ive seen in a while.
 
sadvocate said:
decrease your processor stepping. I have a 3200+ newcastle core with default stepping of 11 x 200 mhz HTT = 2200 mhz or 2.2ghz. By lowering the stepping to 8 I can increase my HTT to 300 easily and put my memory at a 3:4 ratio so it tops out at 250.

Also, decrease your LDT from 4 to 3. This will give you the headroom you need to go higher (even if you don't want to lower your stepping on the proc).

Give it a try, these boards like being pushed.
Click to expand...

Whats the point of running with a higher HTT with a lower multiplier and a 3:4 ratio on the ram? Is there any noticable increase from having a 9 X 260HTT = 2340MHz with a 9:10 ratio compared to a 10 X 235HTT =2340Mhz with a 1:1 ratio?
 
intercollector said:
Whats the point of running with a higher HTT with a lower multiplier and a 3:4 ratio on the ram? Is there any noticable increase from having a 9 X 260HTT = 2340MHz with a 9:10 ratio compared to a 10 X 235HTT =2340Mhz with a 1:1 ratio?
Click to expand...
Don't think so.
 
orkan said:
he means lower your multiplier. You cant decrease your "stepping" ... lmfao.

this is the funniest noob mistake ive seen in a while.
Click to expand...

/salutes grammar nazi

where should I submit my formal apology? I've got a few ideas of where you could shove it. :p

as far as noticeable increases go by lowering your multiplier, test it yourself. run 3dmark2001 and see if you can't see a difference spelled out in numbers?

also, take a look at the article at anandtech about how amazing the overclockability on the board was. notice what steps they used to overclock. they used a 2.0 ghz clawhammer, which has 10 x 200, yet they overclocked it on an 8 multiplier and saw pretty spectacular results. Check it out here.

edit: to answer the question about why lower the mem ratio, I've just noticed a bit better stability OC'ing when I put the ram at a ratio rather than 1:1 at a higher multiplier.
 
grammar... ?

has nothing to do with grammar. has to do with STEPPING of a cpu being a WHOLE hell of a lot different than the MULT! Some people that read these forums are here to get accurate information. ... not go trying to change their CPU stepping in bios.
 
orkan said:
grammar... ?

has nothing to do with grammar. has to do with STEPPING of a cpu being a WHOLE hell of a lot different than the MULT! Some people that read these forums are here to get accurate information. ... not go trying to change their CPU stepping in bios.
Click to expand...

the context of my message explained the difference, the term stepping can be used to refer to increasing the fsb speed a step at a time. I admit I used the wrong word for multiplier, but the rest of my post more than made up for the mistaken word in giving the proper meaning by context. So, yea, you were caught up in the grammar of it, because it's quite impossible to change the physical revision of a chip and I doubt anyone took my instructions for that.

Heaven forbid anyone make a mistake on a word, but thankfully we have helpful people like you to distract us from the topic and point out my little mistake (all the while completely failing to offer any help whatsoever that has anything to do with the topic).
 
265.jpg
 
Those timings are indeed pretty awesome, didn't look at 'em first time I glanced at the screenshot... Fully stable? Voltage?
 
I haven't really tried much to see how high i can go, i just kinda set it there for the time being

cpuz3000.JPG
 
i'm very envious of dfx's pictures.. seeing as I'm sitting on a 3500+ (939) and a ASUS A8V revision 1.02... so my overclockability is nill :( no agp/pci lock for me
 
intercollector said:
Nice memory timings. Mind posting what kind of cooling you have, temperatures, and benchmark results?
Click to expand...
i'm more interested in his RAM
 
Are the a64 DTR proccies supported well by this mainboard?

I'd really like a 1MB l2 cache, but unsure on the support. DFI forums are giving me some really weird jsp errors when I try to access them.

Torn between a 1mb cache or the 200 extra mhz... Cool and quiet doesnt thrill me, but the dtr's seem to be the only 1mb caches commonly available in the a64-3200 price range (about $200)
 
AMD T-type said:
i'm more interested in his RAM
Click to expand...
1:1 265HTT? Could it be anything else but the BH-5 on an extreme amount of voltage? The new Samsung TCCD is great but I've never seen anyone post CAS2.5 with it over 250HTT.
 
to put this clearly to ppl...

memory speeds above 200mhz on a64 setups does NOTHING.
 
at this point in A64's cache doesnt matter, its the mhz that count, 2.4 1MB l2 vs 2.6ghz 512kb, the 2.6 is better
 
Bullitt said:
Are the a64 DTR proccies supported well by this mainboard?

I'd really like a 1MB l2 cache, but unsure on the support. DFI forums are giving me some really weird jsp errors when I try to access them.

Torn between a 1mb cache or the 200 extra mhz... Cool and quiet doesnt thrill me, but the dtr's seem to be the only 1mb caches commonly available in the a64-3200 price range (about $200)
Click to expand...
They work well with a bios upgrade. Or so I've seen.
 
sadvocate said:
why's that?
Click to expand...

the a64 isnt bandwidth starved like the p4. there is just too little gain in performance by ocing the mem. the whole system runs async anyway, so just keep the memory at 200 and oc the fsb for the higher overall clockspeed.

so you can save some cash and get some regular cas2 pc3200 (doesnt really even have to be 2-2-2)
 
Jason711 said:
the a64 isnt bandwidth starved like the p4. there is just too little gain in performance by ocing the mem. the whole system runs async anyway, so just keep the memory at 200 and oc the fsb for the higher overall clockspeed.

so you can save some cash and get some regular cas2 pc3200 (doesnt really even have to be 2-2-2)
Click to expand...

ah, I see what you mean, gains not worth the money. was changing HTT along with RAM speeds so I couldn't verify RAM performance independant of the other changes - but what you say goes along with what I've heard from others, async ram doesn't really hurt performance.
 
There's still a performance gain, it's just not very tangible unless the memory is making leaps and bounds when you OC it... 'specially if you have to lower the timings (which diminishes the advantage of the OC). Anand's last memory article illustrates this thru some of the tests.
 
Jason711 said:
to put this clearly to ppl...

memory speeds above 200mhz on a64 setups does NOTHING.
Click to expand...

but it cant hurt.... i score higher in everything running 230DDR over runnig a ratio makeing it liek 190DDR....

Im sure that guy with 265MHz DDR is runnign faster than most of us.... if those timing are real...but i have a hard time beliving that, iv never seen memroy runnig 2.0 CAS past 255, must be BH-5,
 
Something isnt right, i dont think even BH5 is capable of running 265 1:1 @ 2-2-2 timings.
 
Summoner said:
Interesting stuff, but in the real world you arent going to be running 3.4-3.7v through your memory and gaming and so on. If you did, i cant imagine the life of your A64 is going to be too long with that amount of stress on the memory controller ;)
Click to expand...
Couldn't have said it any better....emphasis on real world. xtremesystems are in fact quite extreme. gr8 site though. running 265 1:1 @ 2-2-2 timings are envious I'll agree but also look at how many guys over there that fry hardware :D
 
lil layzie said:
Yea, with a simple bios update.
Click to expand...



Which BIOS update tho?
got a link? because Im really thinking about doing a DTR with this mobo...since it seems alot of people have good luck with the DTR's and OCing
 
Summoner said:
Interesting stuff, but in the real world you arent going to be running 3.4-3.7v through your memory and gaming and so on. If you did, i cant imagine the life of your A64 is going to be too long with that amount of stress on the memory controller ;)
Click to expand...

The vdimm is independent of your memory controller. The chip would last just as long at XXX fsb at high dimm voltages, as it would at low dimm voltages.

If you mean to say that the tight timings at such a high FSB would lead to premature death of the part... then I would say that is the same fear that all overclockers have lived with since the beginning of overclocking. You run the risk of burning up a chip when overclocking anyway, but since most of them are designed to run for 15yrs at stock speed, it is nearly impossible to guess how long they will operate in overclocked form.

I can tell you that I've had my xp2500 running at 2.6ghz for over a year, with no adverse affects. By the time you'd end up killing it, its time to upgrade anyway. :)
 
Hmm, i could have sworn i read somewhere that there is a max "safe" limit for vdimm when running an A64...

I'll go search around for a bit :)
 
You must log in or register to reply here.
Back
Top