AMD Single-Channel World Record: 3516.66MHz - Conrad's Socket 754 Monster

[EMC2]

/me passes out to the participants and popcorn to the viewing gallery

Wanna borrow my 32-gallon sippee cup for your run Conrad?

Good hunting to all 3 of you

Peace

[Below Ambient]

nice job dude... id sure like to see what you could do with a 939

[conrad.maranan]

Today marks the worst day I've had in a very long time. "How bad can it possibly be, Conrad?" Well, let me start by saying that I flushed some good hardware down the drain.

Early this morning, I mounted my phase change unit to the X850 XTPE. Sweet. Eveerything is running good, so I bump up the VGPU to 1.75V and am pushing 700MHz easy on the core; no artifacts until I get to Nature - and that's where the small triangles come up in the end. Fine. Let's bump up the VDD and VDDQ a little bit since the memory is running nice and cool now. 2.6V for VDD and 2.62V for VDDQ. As soon as I start ramping up the memory clocks in RivaTuner, the computer shuts down. After about three hours of trying to get the card to work again, it just wouldn't wake up.

Screw it. Let's not let 10 pounds of dry ice go to waste. Let's push the Newark 3700+ to new heights.



Crappy AGP card in place and everything is good to go. So I open up ClockGen and CPU-Z and start pushing the clocks. I make it up to around 3660MHz and the system shuts down just as I'm getting ready to save my screenshot in Photoshop. Okay. Let's go into BIOS and give the VCORE a little bump. 1.475 + 126% and the system boots into Windows no problem. I start working my way up from 12x265MHz in ClockGen - 5MHz increments in 30 second intervals just to make sure everything marinates well. At 270MHz, I decided to take a quick snapshot of my block temps. Check out how well the chilly1 DI/LN2 container just holds temps.



As soon as I applied my clocks for 310MHz, I was greeted with a BSOD. The dry ice/acetone concoction is starting to make me nauseous, so now's a good time for a quick cigarette break. Five minutes later, I power up the system and a tiny spark shoots out just next to the base insulation and one of the board's capacitors. A few droplets of condensation just managed to spoil my fun. Even after disconnecting everything and taking a hair drier to the board, it never started up again.

I should've captured a screenshot after every adjustment was applied. I ended my day with nothing but dead hardware...and absolutely nothing to show for. This blows because my chip obviously doesn't have the cold bug.

In closing, I'm tempted to pick up another board on Monday from mWave. But I doubt I'm going to drop another $450 for an AGP card. If anything, I'll just pick up an X800XL. I need a new hobby.

[bachus_anonym]

OH man!!! I'm sorry to hear that, Conrad Mobo and VC dead? This is a disaster!
Try not to worry, man! Tommorow is a new day

EDIT: Are there any visible signs of damage on the mobo and vc itself? If there are, maybe it can be possible to fix it

[Supertim0r]

OH ! sorry to hear this Conrad !
Xtreme Overclocking is an expensive hobby for sure
maybe you could rma your dead hardware if it's not "highly modded"
keep us updated !

[EMC2]

I feel it bro

On the bright side, looks like your CPU is indeed very sweet

Does the MB cause the PS to shut down when you attempt to boot, or does it accept power without causing a PS shutdown and just not make it thru POST? If it's PS shutdown, might be able to remove the cap where the arc occurred and resurrect the MB...

Future note... snapshots with digicam of screen

And remember always - you still have the important things in life

Peace

[afireinside]

Ouch... thats horrible man, especially the video card Let the mobo dry out for a bit more. I had my first DFI sit in a puddle for 2 hours and after 2 days drying it worked fine. If you got water under the caps though, its toast. Happened to me as well my first run Just got a black screen and it never ran again, even a week later with completely different hardware. Plastic sheeting is a must...

[ozzimark]

DAMN! this is sad conrad.. dead hardware always breaks my heart, esp when it's being used for such a good cause

This blows because my chip obviously doesn't have the cold bug.

but on a positive note.. it does seem this way

And remember always - you still have the important things in life

qft, it's always better to be sad with someone than to be sad alone

[NickS]

Dang man. That sucks, but I guess...well..you're lucky anyway. I can't get 2400MHz on my Newcastle even with 1.83v stable. Guess I have a bad chip. 2200MHz causes games to pause sometimes. But, I got a nice 39s SuperPi @ 2.4GHz.

3.5GHz on a s754--that is my next processor lol. When are they going to be released?

[conrad.maranan]

Are there any visible signs of damage on the mobo and vc itself? If there are, maybe it can be possible to fix it

I have examined everything with a magnifying glass. There is nothing that appears to be shorted out. Everything is packed back up in boxes for now. I'll cross my fingers and fire everything up again in a few days. Maybe they'll just miraculously come back to life in the same way that my VX did.

maybe you could rma your dead hardware if it's not "highly modded"

I hope you're aware that's against our Forum Policies. I would never do such a thing, my friend. Please refrain from suggesting such in the future.

Does the MB cause the PS to shut down when you attempt to boot, or does it accept power without causing a PS shutdown and just not make it thru POST? If it's PS shutdown, might be able to remove the cap where the arc occurred and resurrect the MB...

And remember always - you still have the important things in life

The power supply does not shut down. The fans all spin and the LEDs on the board can still turn on. I just don't get a display to screen, even with other AGP cards installed.

As for the important things in life: today marks my first wedding anniversary. I'm not even going to think about my bad luck with hardware today.

Let the mobo dry out for a bit more. I had my first DFI sit in a puddle for 2 hours and after 2 days drying it worked fine. If you got water under the caps though, its toast.

I'll let everything sit for a few days. Hopefully, it'll all just come back to life again.

DAMN! this is sad conrad.. dead hardware always breaks my heart, esp when it's being used for such a good cause

If you were to tabulate all of the hardware I have RMA'd for replacement and purchased because of failures, you'll realize how disappointed I am. I'm almost discouraged at this point, but I...must...find...reason...to...go...on...

3.5GHz on a s754--that is my next processor lol. When are they going to be released?

Send a PM to Vandi423. He's the source for Newark chips.

[ozzimark]

I have examined everything with a magnifying glass. There is nothing that appears to be shorted out. Everything is packed back up in boxes for now. I'll cross my fingers and fire everything up again in a few days. Maybe they'll just miraculously come back to life in the same way that my VX did.

i'm crossing my fingers for you, this stuff better come back!!

(speaking of which, i vote for an addition of a good luck smilie!)

[D U R A N D A L]

Holy Crap! And I thought I had reached my limit!
Those Newark Mobiles are really something!

[chilly1]

HAppy to see teh container working, well.. Did you get a temp in bios?.. Did you use the seal string?

[twilius_basic]

Hey Conrad, just got to read up on this thread Sorry to hear about your losses sounds like you had it going until hardware started going kaput.

Like Chilly asked, can you tell me what sort of temps you saw in the BIOS, when you were reading -75C at the base? I would really appreciate that

I hope to try Chilly's container on my Newark sometime soon too, I hope it goes far like yours did (too bad you couldn't get any screens )

Cheers,

-William

[conrad.maranan]

Reggie - Yes. I did use both the seal string and the foam insulation tape that you included in the last parcel. I guess I skimped out too much and didn't make use of enough insulation. I have learned my lesson the hard way, sir.

William - The temp reading I had in BIOS fluctuated from -55C to -50C. For the most part, it was glued to -55C. chilly1 tubes are the only way to go. Just look at how close the block temps remained to the actual temperature of dry ice.

And here's the bottom line: My Newark 3700+ is another dead component to add to my list of graveyard-bound hardware. I picked up another DFI board early this morning from mWave. Before going back home, I drove out to my sister-in-law's house to borrow the Clawhammer I installed in her system a while back. I fired up the brand new board with my Newark and it was a no-go. The Newark was removed and the Clawhammer took its place. Everything fired up smooth with the Clawhammer.

$118.66 for the new board was my last hope and attempt to break into the 3.7GHz zone. This system is done with and I'll be moving on to something else when time permits. So, gentlemen. I leave you all with 3516.66MHz as the top 754 number to beat. It has been a fun and frustrating ride for me, so I wish you all the best.

[xenolith]

Sorry to see your saga end so tragically, Conrad. Your Newark was a fine piece of hardware that gave its life admirably. I hope you plan on giving it a proper resting spot?

I have a feeling 3516.66MHz for a s754 will be a long lived WR.

[conrad.maranan]

I have a feeling 3516.66MHz for a s754 will be a long lived WR.

Well if twilius_basic travels the dry ice route, he'll take me down easily. 3.7GHz would have been a reality with a slight adjustment to VCORE. And if his chip really enjoys playing in the cold, I can only imagine what LN2 would do.

I'm Googling for custom acrylic trophies right now. I'd like to have the chip embedded in clear acrylic (or something similar) and have a high-resolution photo of my CPU-Z verification right next to it. It'll make a nice paperweight.

[SDX]

I'm Googling for custom acrylic trophies right now. I'd like to have the chip embedded in clear acrylic (or something similar) and have a high-resolution photo of my CPU-Z verification right next to it. It'll make a nice paperweight.

LOL, post pics when its done, that would be awesome!

[MaxxxRacer]

now that is Xtreme conrad.... a bit creepy... but xtreme none the less.

[EMC2]

It was a wild and wooly ride Conrad, you did very well bro

Sorry to hear about your rocket fizzling I'm sure your next project will go even better though with the knowledge learned on this ride

Peace

[twilius_basic]

Well I'm really sorry like the other guys that your adventure has ended thus You did good though m8

I will of course be using your Newark as my goal when I try dry ice However, I must warn you that I'm not very good at suicide-screens. My SS tends to only be a little higher than my bench-stable speeds.

I have my X850XTPE here, hopefully tomorrow or day after I can test it. If it clocks good on air, then I will hopefully try DI on CPU+GPU sometime soon, and strive for the stratosphere.
I know I won't make it, but I'll try if this X850 is any good

Good luck with whatever you do next even when we have disappointing/frustrating experiences, at least we can learn from them, to do better next time

[bias_hjorth]

Awesome Conrad. Glad to see the s754 is still kicking serious butt.

Well done

[NickS]

Send a PM to Vandi423. He's the source for Newark chips.

Thanks man! Will do!

[[XC] moddolicous]

Sorry to hear that conrad. Hopefully u will take pictures next time.

[twilius_basic]

Conrad, can you offer me any tips?

I (very) unexpectedly got a few lbs of DI today, so I figured I'd try a little out on my Newark and just get a feel for things

Well so far, my feelings have been disappointment I don't have a temp probe, so I have to go off BIOS readings, but after remounting a couple times, I get -45C or so at stock vcore, -33C or so at 1.375 x 126% = 1.73v.

Does that sound about right? I've got the screws tightened down as tight as I can freakin get them.

Btw since I'm on my crapped up WinXP install, I've just been messing around, not trying for high scores or anything. Since this latest remount has better contact, I was able to run Car High at 3306mhz (before it just locked up when I tried to open 3DMark2001). Not too bad I don't think But you remember how I said I suck at suicide-shots? Well it holds true --- I still can't even match your 3.5ghz screeny Everytime I get close to it I get BSOD. More voltage doesn't seem to help.
Condensation isn't a problem atm either.

I'm sure I'll be out of DI before I get a response, but never-the-less I'd like some advice for next time if you don't mind

*EDIT* Also, SmartGuardian is currently reading 205-206C (a.k.a. -49C to 50C) IDLE while typing.... do you remember what it said when you were running? Of course load temps go up way higher.