Board arrived today, built and just a fast bench before dinner.
This is at 5088MHz
More to come.
Printable View
Board arrived today, built and just a fast bench before dinner.
This is at 5088MHz
More to come.
Heh, the new board arrived... :)
Have fun Mr. Mainboard Killer! :rofl:
Can I get some boinc numbers :p:
I saw ~30K reported for 2 days on your account but you claimed ~33.5K for a full day @3.5ghz so I take it your stats didnt show a full blown 8 day run:shrug:
Congrats on the new board:up: and PiSlaughter:eek:
Nice MM:D
Finally you can do some real WCG testing:rofl:
About time you got somewhere with it.
Says 1628MHz to me! :sofa:
Pretty good SPi time for such a clock, though! :rofl:
Good thread as I'm interested in this board.
Are you using the Intel cooler, MM?
Awesome Dave - I'm excited for you. :clap: rock and roll
I would call so we could talk bios but :sofa:
Leave cpuz open and do a print screen right when SPi finishes and you can catch the correct cpu speed.
Damn Ruskies!:rofl:
No, will list specs
Yes,yes,yes, damned Texans!:rofl:
Specs:
Intel Burrage Board
2600K D2
TRUE w/113CFM Delta 120x25mm
2x4 gig GSKIll Ripjaws DDR3-1500
EVGA 260GTX
PCP+C 610w Silencer
36 gig WD Raptor
ok, WCG BENCH AT 4578MHZ
Beats my 980X at 4.4ghz.
Beats my 920 @ 4.35 of course:rolleyes:
Seems to have an edge just between 8/9series 4c/8t i7's and 970/980's 6c/12t
How about power consumption MM:) I know it has to be good, Im sold:yepp: going to buy a new cruncher in a week:D
So 4.5GHz takes at least 1.35v. I was hoping it would do it with less than 1.3v.
What is the highest it will do WCG at stock voltage?
Based on reviews I've seen, the 2600K should be generating anywhere from 150-200W at full load. When I get my chip and board, I would like to keep VCore no higher than 1.3v. Once you start pushing the voltage, you run into diminishing returns.
Funny... my E6750@3,5GHz:
Whetstone is about 1k less but dhrystone is almost 4k more than SB...Quote:
ti 11. tammikuuta 2011 11.19.34 Number of CPUs: 2
ti 11. tammikuuta 2011 11.19.34 3242 floating point MIPS (Whetstone) per CPU
ti 11. tammikuuta 2011 11.19.34 18198 integer MIPS (Dhrystone) per CPU
But whetstone is more important(?) I think... And 6 cores more than this...
wow - very nice movieman!
is the burrage board this one?? http://www.intel.com/cd/products/ser...iew/461597.htm
Got mine up & running last night...ok...it was 2am this morning.:rofl:
I just started to play, still need to update bios. Think I have a bad mount, the cpu temps are kinda high(im not telln). had it to 4.1 on stock volts & was running boinc. These are some crazy fast cpu's....Wow!!! I had to ge to sleep, didnt want to make a mistake & screw something up.
Now to get the kids off to school & get this thing tweeked!!
Ps. ram didnt show up until 6:10 last night.:mad: ups jerks!:p:
I'm seeing 120w right now. Machine has been on WCG solid the past 12+ hours claiming almost 51,600, no validated results yet.
Yes
Agreed, just a benchmark to get an idea is all it's worth
Boy my PhII X6 is looking like a power hog more than ever.
Got some validated results.
Scoring all over the place but looks from 41k-51K/day..
NICE!:D
Is that all you got?
http://img262.imageshack.us/img262/1559/capturelq.jpg
:sofa:
Keep in mind that kill-a-watts aren't any precise when used on PFC based psu's. This is something jonnyguru said, but I don't remember where. He said that plenty of times the machine reported less power than what the computer was actually using, which couldn't be because we would have free energy if such thing could happen:rofl::rofl:
You are going to make me sell my 980X to get one of these.
What :eek: Movieman using a single socket board :shocked: 2012 must be around the corner :rofl::ROTF::rofl:
win7-64 Enterprise
Persoally I think you need the call..:rofl:
No, 2011.. socket 2011 that is!:rofl:
Now this is a slick little puppy for power vs electrical draw..
50,000PPD with a machine that draws 120w..
Hell, thats a no brainer..:D
I really think your kill-a-watt is in error. @ 4.1 Auto v-core (2.32) mine was drawing 214 watts with Intel burn Test. 180 watts with WCG. I'm running at least 2 more fans than I need too and 1 extra HD. Vid card is a 9600 GSO. :shrug:
I did run 10 loops of Intel Burn test @4.3 with stock v-core last night. I got a blue screen after being told my system was stable. :ROTF:
I also discovered I couldn't raise the v-core. A bios update today to 1.6 fixed it. :)
My 2600K is still idling @ 4.1. It ran 21.5 hours Sun.-Mon and 22 hours today. I wasted 2 hours futzing around trying to get my v-core to respond. :(
Look at the Sandy Bridge Goodness show itself!!! My first ever 200K day. :) And look at the Boinc boost! Ignore 1/9. I had 2 computers offline. ;) Better numbers to come. My TRUE backplate will be here by the weekend. :up:
http://img3.imageshack.us/img3/3980/...ridgebonus.png
Uploaded with ImageShack.us
http://img24.imageshack.us/img24/462...dgejustone.png
Uploaded with ImageShack.us
^^ So you like SB huh?:rofl:
I played last night and ran at these settings for 12 hours stable on WCG:
For those looking to OC their 2600k's...do we need a custom cooler or will that stock one hold pretty good?
Now this was to find out what "11/10th's was with 1.525vcore set in bios..
WAAAY over what I'd normally do just to do a :p: to my good buddy Hicks!:D
Were do I feel that the "sweet spot" is?
44x 101 on this Intel board for 24/7 WCG work with 1.325vcore and thats where I'm going to stay for now..
Thing is a beast at those settings, runs cool and solid.
Oh Hicks? This buds for you pal!:rofl:
Movieman as a 980X fan I must admit that the Sandy Bridge seems a very interesting proposition.
If you later confirm (I am eager to get this info) that at 120-150 Watts consumption your 2600K can run stable over a week with air cooling and say with acceptable core temps (max 70 C) and have an average over seven days of 50k or more then I will bow very low. This means that a 4 core would equal a 6core at nearly half the power consumption and one third of the price. It is just astounding. It will not change my standard platform that will remain until 2012 at least the X58-980X couple.
I hope that these improvements will be integrated in the future generation that will replace the X58-980X platform with octo and decacores.
Nevertheless I would like to see an exact clock per clock performance test on WCG between the 2600K and 980X. Say with both at 4.2 Ghz with air cooling.
Now I can understand the major breakthrough in power consumption, but I do not understand how 4 cores beat 6 cores at clock per clock. This means that the 2600K can process more instructions per clock and or more floating point operations per clock and the improvement ratio would be of 50% which is enormous.
My guess is that the CPU's running over 1.3v are pulling close to 200W out of the wall socket. I don't trust the Kill-A-Watt to measure power consumption. The only thing I trust is my monthly utility bill to tell me how much electricity is being used.:D
MM make sure you have that kill-a-watt set to Watts. I know it sounds remedial but i used mine the other day and was like :shocked: but had forgotten it defaults to amps or something like that, not watts
For info the comparative specs by Intel between 980X and 2600K. From that it is impossible to expect such performance level from the 2600K.
http://i874.photobucket.com/albums/a...80X-2600K1.jpg
http://i874.photobucket.com/albums/a...80X-2600K2.jpg
http://i874.photobucket.com/albums/a...80X-2600K3.jpg
My westmeres at only 3.7GHz draw 175W headless, I'd believe 150W for SB without a ridiculous over clock.
Maybe your little trick worked, Dave. How many here quickly bought a SB after reading about 50k PPD @ 120w?:rofl:
edit: poop flew out of my mouth.
This new power consumption info has me thinking.. I've not seen hard power numbers for crunching at 5ghz (or the volts needed to get there) but if SB can crunch at 5ghz (or close) at under the watts the other i7's use then it's epicness will be restored in my eyes
Hey! i edited that out haha.
I just saw where a 2600k was pulling 200W at load @5.1 ghz and a 930 @ 4.2 was using 230w (http://www.xtremesystems.org/forums/...d.php?t=264847)
I didn't read thoroughly enough to find vcores used.
Like you say MM a beast of a machine.
I don't believe that 230w figure either. I ran at 4.2 for months and was always between 260-265w.
did you have a video card, and fans and stuff? cuz that 230w figure was only for the board, cpu, and ram. so 260 for a full system might be about right
edit: my killawatt says 242w right now for my i7-950 @ 3.5
Looks like I'll be doing another test this weekend....meter vs Kill a Watt clone. I have to say though I ran a pair of gx2's on a ud5 @ 3.0 (stock volts and cooler) for a total of 721watts (kill a watt)on my corsair tx750 with no problem so I do not expect it to be that far out
Yup, full system, I was getting about 35k a day from that. So if these produce 45k at 200w, that is pretty damn impressive.
An example of beastiness:
http://img838.imageshack.us/img838/4310/capturetj.jpg
Attachment 111097
This a 980X running a 4113.
Thought has crossed my mind.
Thanks Dave. I was feeling idiot with my caculations trying to understand how you could ever be at 120 Watts. Now all seems to make sense.
At 4.4 100% load the 980X draws nearly 90-100 watts more. The 2600K solution is at least 32-35% less power hungry that a 980X which is still a very impressive number.
Now there is still a point unclear to me. Is the internal GPU of the 2600 still active and usable in the K version of the chip with unlocked multiplier?
If it is usable then some part of the power cunsumption reduction is due to the lack of a graphic board that is necessary for the 980X solution. To be fair to the 980X I should subtract some watts due to a basic graphic board.
Do you agree Dave?
Dave here under is your post on WCG forum:
But now the numbers in the posts do not match.Quote:
Watts draw at 100% load is not the puny 120w I thought but at a highly overclocked 4500+MHz isapp 212W..
At default 3500MHzthe draw at 100% load is 193W
My apologies for the false info
Once you have 196Watts at 4.4 Ghz and then it is at 3.5 Ghz.
So if I take the draw at 4.4 Ghz at 212 Watts the difference with the 980X is narrowing but still very sensible. Let's see what happens WCG performance wise.
Retro77 and Dave both logged 5600 boinc today, I still think they can average 6,000 :up:
sorry but this is too funny :rofl:
I know! My wife says she forgives you though, Dave. After all, you are a mere mortal :)
Are there any budget air coolers that will allow for a healthy overclock? I'm going to be selling of the 920 and motherboard and getting a SB rig. My 920 is watercooled right now, but I don't have the mounts for a 1155 board. I also don't really have the time to break down my watercooling loop to switch over to a new block (I currently have an original Fuzion), so I figured air was the way to go until I have time to redo my loop properly.
Statistics
Date ..... Total Run Time Points Generated Results Returned
1/14/11 0:008:06:18:09 ...... 44,062............... 51
So thats app 8.25 core days.
44062 divided by 8.25 then multiplied by 8=42,726 PPD
Not bad for a quad.:up:
I think today is a benching day..
Just looked at my therometer on the back porch and it's minus 2 F..:rofl:
Was just checking some stats and interesting numbers.
Average credit per CPU second 0.009550
http://boincstats.com/stats/host_gra...wcg&id=1463367
That's as high as I've seen for any machine except for the few exceptions where someone ran the machine for a few minutes and then shut down.
I think the key to where this fits into the game is that it just will run higher on air than other cpu's.
I'm set at 4634MHz for the last 36 hours with temps at just under 60C with 1.325vcore.
For me personally I'll shut down Sammy, cut the 90w that takes, add this with it's 212w draw and for a net increase of 122w draw Ill go from 7000PPD to 42,000PPD
Now to wait for socket 2011:D
So, rig on porch.... keyboard, mouse and remote reset switch on desk inside?
Thats about -19C boys and girls
I had to find some advantages to my 980Xs. I cannot just let a newcomer beat me too easily.
Here it is. Points wise you may be near to the 980X performance but runtime wise I can post 12 and more runtime days/day per cpu. Thats the unbeatable advantage of an hexacore.:D
Hello my friend.
Yes, you are right.
The 980's are the king of the single socket.
Where I see these being a good choice is for people thinking of building a lower cost cruncher.
In the US you can buy a 2600K for app $330.00 while a 980x costs close to a $1000.00
Then the cost of the boards is almost the same but you only need 2 sticks of DDR3 memory vs the 3 on the X58 chipsets.
Maybe it's usefullness is for those that want a top producing system but don't want or can't afford the top Hex's..
So far I like it except that flashing the bios on this Intel board is causing me to pull out what remaining hair I have.
There are 4 ways to update bios and I've tried all of them many times in the past 2 days and the only thing I've accomplished is to waste time that could have been crunching.:D
Dave,
I took your advise and ordered a 2600K.
Thanks for having me spending even more money than I first planned to (2500k). :ROTF: :rofl:
Has any of you checked what the power comsumption is at may OC on stock voltage(running WCG)?
And if possible headless or with a simple GFX card?
And how many PPD does it do than?