• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Project Freya - A New Contribution to the Team

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Enkidu

Member
Joined
Dec 1, 2003
Location
Boulder, Colorado
Well folks -

We are in a challenging period as a team. We have had some stress and some divisiveness on our forums of late. We have lost some respected members of the team and community, and we have struggled with our differing opinions.

I for one have been hoping for a new and inspiring focus - something that reminds us why we are all here: The fun and challenge of overclocking - and the passion we feel for distributed computing projects such as SETI.

To that end - I'm launching a project of my own :) Over the last week or so I have accumulated all the parts I need to build a new rig. "Wow, a new rig, boy that's real exciting, Enk" you say? Anticlimactic? Perhaps - but it is a particularly fun overclock; one that is (quite honestly) far, far over my head. I will probably fry a motherboard along the way, make some major mistakes, but I'll learn from them - probably mostly from people on this very forum with far more experience overclocking than I :)

I am dubbing the project "Project Freya" (I intend to name the new machine freya). Here are the specs:

Asus PC-DL Motherboard
Two 1.6ghz low voltage Xeons with D1 Stepping
2x256 (512mb) memory
Two Intel Wind Tunnel HSFs
550 watt Sparkle power supply
An Antec P-180 Case (limited edition SPCR "Black" model)

What's so special about a couple of 1.6ghz xeons? Well, the low voltage D1 stepping beauties have, with some frequency, clocked to over 3.0ghz without any heroic cooling techniques - on nothing but air! I'll be happy if I can get them to 2.8ghz.

This is not a new thing I have discovered, there are a number of places that detail how to do this with this board and these processors, but the project requires soldering on the motherboard - a vdimm mod, a vcore mod and so forth - accordingly it will be quite fun :)

I'll be working on this over the next 2-4 weeks - as I get everything laid out I will start posting some photos and, if all goes well, links to a new cruncher for the team that is doing over 1000cs/day (That's my goal anyway, we'll see how close I can get).

I hope everyone has a wonderful week - let's remember why we're all here: Each other, the hobby, and our respective projects!

Enk
 
Last edited:
Looking good Enk :D, just an one q: you will have 2x256Mb RAM in there? Because if its 1x512 I would strongly suggest going to dualchannel, dual xeons each with HT would easily get bottlenecked with only 3.2Gb/s (theoretical) memory bandwidth.
I've seen alot of people hit 3.2Ghz ( 800FSB ) out of those 1.6lv, so gl with overclocking.
 
Glad to see that not everyone is affected with the current issues we have as a team.

I wish the best of luck to your project - I am looking forward to see some pics and results. And I hope this little project will be an eksample of what Overclockers.com is all about. Because it seems that some way down the road - we forgot why we were here.
 
lol when i first read project freya, i thought you had quit the SETI team to make your own distributed computing project xD thank god I was wrong :p

good luck with that 1000cs/day :eek: that would be an awesome contribution

Careface*
 
Hamm3r said:
Looking good Enk :D, just an one q: you will have 2x256Mb RAM in there? Because if its 1x512 I would strongly suggest going to dualchannel, dual xeons each with HT would easily get bottlenecked with only 3.2Gb/s (theoretical) memory bandwidth.
I've seen alot of people hit 3.2Ghz ( 800FSB ) out of those 1.6lv, so gl with overclocking.
I did indeed get 2x256 for that very reason, thanks (I've updated my original post to reflect this now)

Dk Jedi Allianc said:
Glad to see that not everyone is affected with the current issues we have as a team.

I wish the best of luck to your project - I am looking forward to see some pics and results. And I hope this little project will be an eksample of what Overclockers.com is all about. Because it seems that some way down the road - we forgot why we were here.
Thanks - and indeed helping to "re-center" the team is one of my goals :)


Careface said:
lol when i first read project freya, i thought you had quit the SETI team to make your own distributed computing project xD thank god I was wrong
Oops - Sorry for the confusion Careface - I'm not going anywhere :)

Enk
 
Good luck with that dually! My dually has never hit 1k cs a day. 810ish a day was about the best I could do. Haven't done the v-dimm mod either... one of these days... lol. I would like to know how Freya turns out, so keep us updated! :attn:
 
Update:

The case arrived today, so I got everything unpacked and ready to rumble.

Unfortunately I ran into a major snag - as I was attempting to mount the intel Processor Wind Tunnels to the PC-DL - I realized that the screws that come with this HSF do not come with bolts or a mounting plate so there is no way to secure the HSF ot the board. The PC-DL comes with its own little mounting aperatus that would allow me to mount the heatsync, but without the Wind Tunnel housing that goes over the heatsync, I'd have no fans on the CPUs, and that is a bad thing.

Not sure what I'm going to do yet - I may have to take the screws into a hardware store and get some nylon washers and a few nuts.

Here are the two side by side - on the left is the "first stage" intel Processor Wind Tunnel - note the little gold screw that - when placed through a mounting hole, has nothing to secure to, this is what I will need a nut for (or a mounting plate, or something I haven't thought of). On the right are the mounting brackets that come with the board, note they use the little plastic push-through mounts, no screws.

dscf0174.JPG


Here is a similar shot with the sync on (and in the case of the wind tunnel on the the left, the fan as well). Only thing missing here are the retention brackets (and of course the fact that if I were to turn the board over everything would fall off).

dscf0176.JPG


Unfortunately, I'm heading out of town for about a week on thursday so this puts a prety big delay in my efforts. If anyone has any ideas - let me know :)

Enk

[Edited to add photos]
 
Last edited:
Looking forward to seeing what it will do. At best I was getting about 1200RAC with my pentium D 820.
 
Very cool but I'm not ready to graduate to water cooling yet :) - as well as wanting to figure out how high I can get these on air.

I'll keep it bookmarked though!

Enk
 
I run mine at 3ghz (14x215) with the stock IWTs. I can get it to boot into windows at 3.3g, then it BSODs, and I can run programs at 3.2g, but prime fails in a matter of minutes (if that long). This is done at 1.5v. 1.6v didn't seem to make much of a difference for stability. One of these days I will have to finish this system... lol

Wish you Luck on your dually. :thup:
 
Just a quick update -

Since I'm back from holiday - I swung by the hardware store and got the parts I needed - I have the board and all components in the case and am successfully POST'ing at 3.2ghz (16x200) - wow!. Running idle (using bios monitor) at about 44c.

I have not yet installed windows (I think my drive is bad, so having to try one of the 10 others I have laying around) so I don't know what the machine will run at under load.

Given 44 C idle I'm guessing I'll need to run it at 2.8-3.0 instead of 3.2 - but since my stated "Project Goal" was 2.8ghz or higher, I'm doing pretty good so far!

I'll post picture series later along with benchies when I have them.

By the way - HAPPY NEW YEAR!

Enk
 
Last edited:
1.56v (remember these are the 1.3 low voltage CPUs). Everything I've read says I need 1.6v to run stable at 3.2ghz and at 1.56v it has indeed locked up a couple of times (once during a windows update, forcing me to reload, ack! hehe).

The one mod I didnt finish (I am afraid of the solder point) is the one that would push the board to 1.6v. I do have a second set of these identical CPUs, so in a few months I'll build another one and give the solder point a whirl. If I can get that second one working well, I'll come back and do it to freya.

The major remaining question I guess, is what speed I ran run at given 1.56v. I'll have to play with them for awhile I imagine.

Enk
 
Back