PDA

View Full Version : 965 BE boxed parts testing.



chew*
08-22-2009, 08:54 AM
As the title states.

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/med_IMG_1389.jpg

Chip 1 default vcore.
http://chew.ln2cooling.com/965BE/boxed%20parts%20first%20shotb%20%200163.JPG

Chip 2 default vcore, stronger core's it seems weaker NB.

http://chew.ln2cooling.com/965BE/boxed%20parts%20first%20shot%200355.JPG

Chip 3 default vcore, has the strongest NB ( needs least amount of NB_vid )

http://chew.ln2cooling.com/965BE/boxed%20parts%20first%20shot%200362.JPG

chew*
08-22-2009, 08:55 AM
This is a preety beat install of XP Sp1 but its the most reliable, seeing as all my other OS's are corrupt atm, scores won't be the greatest.

Playing around with max benchable core frequencies atm.

Chip 3 ( I think this might me my ln2 chip ;) loves volts )

Not bad for the heatwave i'm dealing with atm.

http://chew.ln2cooling.com/965BE/965%20benching.JPG

http://chew.ln2cooling.com/965BE/965%20benching%202.JPG

Chip 2 Looks to me you air guys are gonna want the default 1.4v chips.

http://chew.ln2cooling.com/965BE/965%20benching%203%20chip%202.JPG

http://chew.ln2cooling.com/965BE/965%20benching%204%20chip%202.JPG

Seemed to need quite a bump here......

http://chew.ln2cooling.com/965BE/965%20benching%204%20chip%202%202.JPG

Seems to be a pattern forming all 4 gig + in cine, FYI if cine passes this at this speed 4.1 3d is a cinch ;)

Chip 1 with 1.425 default is almost identical to its brother chip 3 for max.

http://chew.ln2cooling.com/965BE/965%20benching%205%20chip%201.JPG

http://chew.ln2cooling.com/965BE/965%20benching%205%20chip%201%202.JPG

I think I will go play the lottery now.

Titan7171
08-22-2009, 08:55 AM
Damn I need to get ione of these.......4.1ghz default core...hell yea!

chew*
08-22-2009, 08:56 AM
Reserved 2

chew*
08-22-2009, 08:58 AM
Reserved 3

surfinhicdude
08-22-2009, 09:02 AM
You can send the worst one my way when you are done. :D

chew*
08-22-2009, 09:04 AM
You can send the worst one my way when you are done. :D

The worst ones are the "cherry" pick review samples that AMD sent me that people bust my balls about :ROTF::rofl:

LaMpiR
08-22-2009, 09:08 AM
Nice stack :)
Is it doable for most of them to go over 4Ghz or just those cherry picked?

beta
08-22-2009, 09:13 AM
Nice stack :)
Is it doable for most of them to go over 4Ghz or just those cherry picked?

i wonder what chew will reply to this :rofl:

chew*
08-22-2009, 09:34 AM
Nice stack :)
Is it doable for most of them to go over 4Ghz or just those cherry picked?

These are boxed parts, I spent about 20 min on chip 1 almost done testing chip 2 then I will get to benches......that said.

You be the judge.

FlanK3r
08-22-2009, 10:19 AM
and where is third chip?:) I see 3 x4 965, right?

FlanK3r
08-22-2009, 10:21 AM
btw, send me one :), change with my x4 955 :)

Zeus
08-22-2009, 10:23 AM
Chip 1 has a default Vcore of 1.424V?

Looks good mate, wondering how these cherry picked retails will do. :D (sorry, couldn't resist.)

LaMpiR
08-22-2009, 10:33 AM
Sorry, didn't see the boxed part of the title :)
p.s. hangover sucks :shrug:

RussC
08-22-2009, 10:42 AM
Wow, I agree, that real, default is 1.424?

RussC


Chip 1 has a default Vcore of 1.424V?

Looks good mate, wondering how these cherry picked retails will do. :D (sorry, couldn't resist.)

chew*
08-22-2009, 10:54 AM
Wow, I agree, that real, default is 1.424?

RussC

Yes russ it varies from chip to chip thus amd's loose rating of 140W TDP.

Here's something my review samps can't do......

2 things are needed to scale these chips, voltage and temps.......

I could not boot at 1.55v on my review samps on phase........

Don't try this at home kids....cpu death is imminent especially on air like this.

http://chew.ln2cooling.com/965BE/tolerance.JPG

hikaricloud
08-22-2009, 11:07 AM
That's crazy. Looking forward to great things to come with these chips you got there.

artemm
08-22-2009, 11:25 AM
I could not boot at 1.55v on my review samps on phase........

Don't try this at home kids....cpu death is imminent especially on air like this.


I like where this is going :yepp:

WSP
08-22-2009, 12:50 PM
those boxed parts definitely better than that 'cherry pick' you posted a while ago :D

throw in some max air cooling bench please :)

afireinside
08-22-2009, 12:58 PM
Don't try this at home kids....cpu death is imminent especially on air like this.



:shocked: I was running ~1.67v real on air for a while...

965s look nice. I want one.

FlanK3r
08-22-2009, 01:39 PM
Il try for compúarsion the same with my x4 955 :) and the same vcore, but think, end behind 3900-4000 MHz

FlanK3r
08-22-2009, 03:07 PM
lol, u mean really, 1.728V with AIR? OMG :D...I have fear with 1.55V :). Il look forward next benches chip2 and chip1 (to time u post chip3 at air)

here my chip x4 955 with 1.4V 4000MHz/2800MHz (higher i dont tried)

http://img341.imageshack.us/img341/2541/easyw.jpg ( http://img341.imageshack.us/img341/2541/easyw.jpg higher image)

Lastviking
08-22-2009, 03:40 PM
lol, u mean really, 1.728V with AIR? OMG :D...I have fear with 1.55V :). Il look forward next benches chip2 and chip1 (to time u post chip3 at air)

here my chip x4 955 with 1.4V 4000MHz/2800MHz (higher i dont tried)



I runing my 940BE @ 1.585v 24/7 and 1.625v when i bench..some chips seems to like more vcore than another :) , i dont wory about to high vcore. They always have reboted for me if i push the vcore to much :) , then i lower it abit. my first 940BE hade problems over 1.55v but this chip works fine to about 1.625v with air.

FlanK3r
08-22-2009, 03:43 PM
whau, interest...think, my like 1.45,1.475 and 1.485V mostly..
whats your temps at 1.585 (1.625V)?

chew*
08-22-2009, 04:27 PM
Flanker flop that NB up to 2900-3000 at least for all cores bud at 4 gig and 1.4v

High NB changes alot ;)

VoodooProphetII
08-23-2009, 01:08 AM
Chew* I ended up borrowing a friend's M3A79-T Deluxe (due to my Crosshair III phase that burned) to test the retail 965 I have on some DICE. Now unfortunately I could not get lower then -45C on the pot due to less surface area. I designed the pot myself for more mass, but will need to drill some more holes for surface area. Anyways, I was able to bench Vantage at 4.725Ghz at 1.58Vcore with 3600NB. The chip runs pretty well, but I had problems getting stable above 1.6Vcore, but I could run 06 at 4.88Ghz with the same vcore. I think the problem I have might be my Antec 650Watt trio struggling because of the 140TDP....but I stand to be corrected.

Here's the Vantage link. Unfortunately my electricity is off at home, can't get the screenshot from my PC right now.

http://www.hwbot.org/result.do?resultId=887398

chew*
08-23-2009, 06:44 AM
Chew* I ended up borrowing a friend's M3A79-T Deluxe (due to my Crosshair III phase that burned) to test the retail 965 I have on some DICE. Now unfortunately I could not get lower then -45C on the pot due to less surface area. I designed the pot myself for more mass, but will need to drill some more holes for surface area. Anyways, I was able to bench Vantage at 4.725Ghz at 1.58Vcore with 3600NB. The chip runs pretty well, but I had problems getting stable above 1.6Vcore, but I could run 06 at 4.88Ghz with the same vcore. I think the problem I have might be my Antec 650Watt trio struggling because of the 140TDP....but I stand to be corrected.

Here's the Vantage link. Unfortunately my electricity is off at home, can't get the screenshot from my PC right now.

http://www.hwbot.org/result.do?resultId=887398

Yah I have no problem cranking up the volts on the retails, they accept it and bench higher....My air bench rig uses a 750w PC P&C, my cold rig uses a 1200w + 860 w turbocool ;) No power issues

BTW cinebench is about the most intensive bench you can run becasue it hammers the core and NB which is why I ran it, it tells me alot about a chip quickly.

Benching over 4060 in all other benches should be a breeze with all 3 of these chips.

chew*
08-23-2009, 11:51 AM
So playing with chip 1 right now, Team XS can use some points so I tackled some air scores ;)

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_1stplace%20air%20965be%20wprime%2032m.jpg

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_1stplace%20air%20965be%20wprime%201024m.jpg

http://chew.ln2cooling.com/965BE/pifast%20air%201st%20965BE.JPG

Gonna go rip in a fresh OS to tackle some PI :)

FlanK3r
08-23-2009, 12:16 PM
whau nice, have first screen bigger? Its 8.814s? Great:up:

a603nk
08-23-2009, 12:20 PM
wew... 965 be @ 4Ghz+, amazing

chew*
08-23-2009, 08:25 PM
So apparently I was told I could not do 4 gig + 1800 + tight on a retail chip in this thread....http://www.xtremesystems.org/forums/showthread.php?t=230103 :rolleyes:

Chip 1 Oh and its air not phase......

http://chew.ln2cooling.com/965BE/1800%20tight%20retail.JPG

Sgt.McRuff
08-23-2009, 08:55 PM
Chew what memory where you using there?

Frontl1ne
08-23-2009, 09:10 PM
That's some awesome OCing chew*, thanks for posting the results :D

Zeus
08-23-2009, 10:30 PM
Chew what memory where you using there?

I'd bet my bike that it was Elpida Hyper based mem like OCZ Blades or Corsair Dominator GT.

Nice run chew* but i would have expected 4.2GHz 32M stable from that chip...

chew*
08-23-2009, 11:14 PM
How about we sacrifice 50 mhz cpu for 50 mhz NB ;)

I'm still in the middle of a heatwave + on air Zues....80 degree nights with humidity you can imagine what daytimes like.

Still haven't broken out my TRUE 120 and Delta fan yet :p:

Chip 1

http://chew.ln2cooling.com/965BE/32m%202.JPG

Zeus
08-23-2009, 11:30 PM
That's more like it! :) Very nice NB!! :up:

Lightman
08-24-2009, 12:43 AM
Very nice set of CPUs!

They seem to be very consistent, variations are small.

*chew can you list weeks of production for these chips?
Great work BTW!:up:

chew*
08-24-2009, 07:40 AM
All same week.

Hopefully by now I have got my point across.

http://chew.ln2cooling.com/965BE/IMG_1391.jpg

charged3800z24
08-24-2009, 08:07 AM
Very nice results Chew*. Those chips seem pretty nice. I am trying hard not to buy a 965. I told myself this 955 will hold me till new silicoln;). I I bought more AMD chips this year than a sane person should. So, when you coing to try out the TRUE on them?

chew*
08-24-2009, 08:18 AM
Very nice results Chew*. Those chips seem pretty nice. I am trying hard not to buy a 965. I told myself this 955 will hold me till new silicoln;). I I bought more AMD chips this year than a sane person should. So, when you coing to try out the TRUE on them?

Probably don't need to, although it would help me when I test 3d. The setup I have on the xiga 1284 is just so tolerable noise wise I only use the TRUE/Delta now when its an absolute must.

Getting Ready to put some stuff under phase anyway.

FlanK3r
08-24-2009, 08:41 AM
nice run, with 4140/3000 is it unbelievably (but i see it :-D ), now is hot, today i tried 4050 MHz superpi and it was crash :-/

knopflerbruce
08-24-2009, 01:30 PM
Sweet clocks! Can't wait for some LN2 action;) You seem to know what you're doing so I'm sure at least one of these will reach some impressive numbers.

Just waiting for my motherboard so I can try my 940 on LN2. I envy you too much so I had to get one:p: It'll be weird to have no CB (even if my Claws can go pretty low they DO have a bug, hehe).

chew*
08-24-2009, 05:35 PM
Almost there......no tweaks save affinity.

http://chew.ln2cooling.com/965BE/1m.JPG

chew*
08-24-2009, 08:06 PM
This concludes air on chip 1...... with tweaks i'm preety sure that would have been a sub 16 1m run.

http://chew.ln2cooling.com/965BE/1m%202.JPG

LowRun
08-25-2009, 03:34 AM
So apparently I was told I could not do 4 gig + 1800 + tight on a retail chip in this thread....http://www.xtremesystems.org/forums/showthread.php?t=230103 :rolleyes:


Just been reading the thread two times in a row and couldn't find anyone saying you couldn't do this with a retail :shrug:

chew*
08-25-2009, 06:07 AM
Just been reading the thread two times in a row and couldn't find anyone saying you couldn't do this with a retail :shrug:

Numerous people pointed out the fact that I was not using retail or how retail was my retail......the reason they pointed this out is because they do not think its possible on retail even on phase change ;) and assumed my Retail grade review sample was just cherry picked.....Or I was using a super cherry.

I also got PM's about it.....

Myth busted....It's all about the driver behind the wheel......not the chip being used.

Booj
08-25-2009, 06:34 AM
Nice testing in here Chew* :up: I'm liking the rams nice 'n tight.

I have a 965 and Giga 790fxt ud5p here for a play with a cascade. I'm looking forward to some fun on the green side ;)

chew*
08-25-2009, 06:38 AM
Awsome have fun, :up:

If your using the same IC based modules on giga they hit a lower wall just so you know, board is tuned better for d9 gtr/gts.

Look for 6-6-5-15-11 around 1720 for GTR, and 7-6-5-15-11 around 1800 for gts......GTS will need less volts.

Giga needs to be cold ( only sub 0 etc doesn't need more than a single stage ) or you hit a reboot loop about 1680 clocking most ram.

Booj
08-25-2009, 07:05 AM
Thanks for the tips man :up:

In the short time i've had i've been at ~1600 6-6-5-15 but with trc much looser than what you've got running, i'll be sure to give that a try. No problems here with the board as of yet.. knock on wood.

chew*
08-25-2009, 07:11 AM
Thanks for the tips man :up:

In the short time i've had i've been at ~1600 6-6-5-15 but with trc much looser than what you've got running, i'll be sure to give that a try. No problems here with the board as of yet.. knock on wood.

If you think TRC's tight, I'm running those 4 gig modules at 90 ns as well ;)

If your using GTS don't bother with 6-6-5 they need to much volts to sustain that speed at high clocks.

They will degrade, my crucials are just about toast from trying it. They run 7-6-6 1800 undervolted to give you an idea how much easier it is.

FlanK3r
08-25-2009, 08:23 AM
This concludes air on chip 1...... with tweaks i'm preety sure that would have been a sub 16 1m run.

http://chew.ln2cooling.com/965BE/1m%202.JPG


AIR or now phase? Nice:up:

chew*
08-25-2009, 08:38 AM
Thats air I predict phase will be quite a bit higher, thats the worst chip of the bunch as well :D

hikaricloud
08-25-2009, 08:44 AM
Man, stop making me want this chip! D: That's pretty damn impressive.

EnJoY
08-25-2009, 08:56 AM
sp1? is it that hard to slipstream a disk with sp3 on it? :rolleyes:

FlanK3r
08-25-2009, 08:58 AM
omg...this chip run about 100 MHz higher than my chip :) -->this must be min 3950 MHz stable

Tomasis
08-25-2009, 09:04 AM
It's all about the driver behind the wheel......not the chip being used.

I N D E E D :up:

nice 16sec run :)

chew*
08-25-2009, 09:26 AM
sp1? is it that hard to slipstream a disk with sp3 on it? :rolleyes:

YES :D

Want to do it for me.

Titan7171
08-25-2009, 10:16 AM
I ordered one....It should be here tomorrow!

Buckeye
08-25-2009, 10:37 AM
I ordered one....It should be here tomorrow!

Nice Titan, you will be very pleased with these CPU's :up:

Titan7171
08-25-2009, 10:47 AM
Nice Titan, you will be very pleased with these CPU's :up:

cant wait:up:, hope I can get at least 4.1ghz for 24/7:yepp:

Im going to be putting up to 1.55v for 24/7 use, if they run cooler than the 940's then maybe 4.2ghz?

Also it will be running on M3A79-T Deluxe so lets see how it reacts to this mobo, will I be able to hit the 3ghz NB and HT or not:shrug:

Zeus
08-25-2009, 11:06 AM
I think 4.1GHz for 24/7 is aimed a little too high, at least if you want it to be Prime (Blend) stable.

You're gonna push 1.55V on air?

Most phenom II cpu's i've seen have their sweet spot around 1.48V.

I hope you have better luck though. Looking forward to your result.

Titan7171
08-25-2009, 11:19 AM
I think 4.1GHz for 24/7 is aimed a little too high, at least if you want it to be Prime (Blend) stable.

You're gonna push 1.55V on air?

Most phenom II cpu's i've seen have their sweet spot around 1.48V.

I hope you have better luck though. Looking forward to your result.

yea 1.55v on air is what I ran until spring for 3.8ghz on this 940be...... I dropped it to 1.53v/3.7ghz because of heat issues this summer because I didnt hook up the ac in my room yet:shakes:(its sitting on the floor). If they run a little cooler I might be able to pull it off

I need to get at least 400mhz increase out of this (above what Im already able to run)for it to be worth the upgrade.

FlanK3r
08-25-2009, 11:27 AM
I N D E E D :up:

nice 16sec run :)

with classic DR2 1066 is not possible, maybe 16.2s max:( And in 32M is much bigger diferent

chew*
08-25-2009, 12:45 PM
yea 1.55v on air is what I ran until spring for 3.8ghz on this 940be...... I dropped it to 1.53v/3.7ghz because of heat issues this summer because I didnt hook up the ac in my room yet:shakes:(its sitting on the floor). If they run a little cooler I might be able to pull it off

I need to get at least 400mhz increase out of this (above what Im already able to run)for it to be worth the upgrade.

I think 4.1 is a little optimistic for air for 24/7 maybe water.....but air I don't think so.

4.1 benchable yes thats very realistic.

I know prime passes 3900 like a breeze.......

the 1.4 v chips seem to have a sweet spot at 1.475 however have more tolerance than my 955's and can do 1.55.......the 1.425 default chips seem to like 1.55...

Titan7171
08-25-2009, 01:19 PM
I think 4.1 is a little optimistic for air for 24/7 maybe water.....but air I don't think so.

4.1 benchable yes thats very realistic.

I know prime passes 3900 like a breeze.......

the 1.4 v chips seem to have a sweet spot at 1.475 however have more tolerance than my 955's and can do 1.55.......the 1.425 default chips seem to like 1.55...

Well I shall try......3.8ghz@1.55v on a 140w 940BE/air was also a hardsell when they were first released:yepp: and my 4.26ghz air is still one of the highest for a first run 940 just check the Phenom 2 overclocking thread.

I must succeeed in my quest for 4.1ghz:p:

FlanK3r
08-25-2009, 01:32 PM
Chew: one man in overcloksnet have x4 965 AIR 4020MHz with 2700+ MHz at NB stable. Great chip, think. But what im remember, one your 955 was too at 4 GHz stable...

chew*
08-25-2009, 01:42 PM
So 2 chips got stable out of how many? was it prime blend stable?

FYI I can pass all other stress tests as easily as I bench cinebench at 4 gig so not much of a stress test to me.

I think the main fact is that the 965BE is consistent......

Also don't forget that now that amd has a higher bin I would not expect to see to many outlier chips etc they are gonna tighten the noose since they are binning for the higher chip now.

Here's some work on phase :)

http://valid.canardpc.com/cache/banner/673609.png (http://valid.canardpc.com/show_oc.php?id=673609)

http://valid.canardpc.com/cache/banner/676679.png (http://valid.canardpc.com/show_oc.php?id=676679)

FlanK3r
08-25-2009, 01:47 PM
Hm, u right, Cinebench is not far way system stability,my max stability is 3880MHz and max Cinebench 3950 MHz. 4GHz not possible, in x64 total not :)

Chew:can u for me tested this test stability? 14 000 block size or more for rock stability with 15 rounds. I mean, its hard test, not too long, but hard. I mean, it stress too northbridge.

http://www.svethardware.cz/forum/downloads.php?do=file&id=1670

Zeus
08-25-2009, 02:00 PM
You don't need all sorts of progs to test stabilty really. If it can stand at least one hour of Prime 95 blend without rebooting you're fine.

I can run 4GHz all day long without problems but it will reboot in Prime blend sooner rather than late. Must say i haven't tried over 1.48V though.

chew*
08-25-2009, 02:16 PM
Hm, u right, Cinebench is not far way system stability,my max stability is 3880MHz and max Cinebench 3950 MHz. 4GHz not possible, in x64 total not :)

Chew:can u for me tested this test stability? 14 000 block size or more for rock stability with 15 rounds. I mean, its hard test, not too long, but hard. I mean, it stress too northbridge.

http://www.svethardware.cz/forum/downloads.php?do=file&id=1670

I will take a look at it one of these days. ATm i'm very busy chasing peoples scores on better cooling than I'm on atm :D

Titan7171
08-25-2009, 02:21 PM
So 2 chips got stable out of how many? was it prime blend stable?

FYI I can pass all other stress tests as easily as I bench cinebench at 4 gig so not much of a stress test to me.

I think the main fact is that the 965BE is consistent......

Also don't forget that now that amd has a higher bin I would not expect to see to many outlier chips etc they are gonna tighten the noose since they are binning for the higher chip now.

Here's some work on phase :)

http://valid.canardpc.com/cache/banner/673609.png (http://valid.canardpc.com/show_oc.php?id=673609)

Its not the chip its the overclocker...;)

Ive been overclocking since the AMD t-birds, I got a couple tricks here and there.

chew*
08-25-2009, 02:23 PM
Its not the chip its the overclocker...;)

There are certain limitiations.......I don't want people to get there hopes up on 4.1 24/7 stable chips thats all ;)

chew*
08-25-2009, 02:25 PM
Ok this one is probably not a final.

Didn't put to much effort into it yet.

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_1m%20phase%201.JPG


These I did put a little effort into, had to give bill a reason to bench again :p:

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_32m%20phase%201.JPG

Not only is 1800 on 4 gig sticks done and tight it's also broken :D
Not as efficent but the ram was probably a tad ticked off......

No tweaks other than priority and affinity......

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_32m%20phase%202.JPG

Buckeye
08-25-2009, 02:34 PM
HAHA Good one Chew :rofl:

Keep in mind that I am a first time user of these CPU's and have my share of learning to do.

As I am busy with another project atm., but I will get that chip back on the Cascade real soon and see if I can even come close to you, it won't be easy ;)

Titan7171
08-25-2009, 02:44 PM
There are certain limitiations.......I don't want people to get there hopes up on 4.1 24/7 stable chips thats all ;)

Oh I defininitely understand. We'll see what happens when I get this thing:shrug:

chew*
08-25-2009, 07:31 PM
Been a real long day and I only got through one chip not even sure which chip it is anymore gonna have to wipe paste off tommorow to find out.

Anyway here goes.

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_1m%20phase%202.JPG

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_pifast%201.JPG

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_wprime%2032m%20phase.jpg

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_wprime%201024m%20phase.jpg

Definitely beat my best 955 in 3d...........

http://chew.ln2cooling.com/qdig-files/converted-images/965BE/lrg_3d%20phase%20965%202.jpg

I'm going to bed now, i'm beat........

Buckeye
08-25-2009, 08:32 PM
Yeah go to bed now !!!

Daveburt714
08-25-2009, 08:40 PM
Very Nice results on all those chips Chew*.... :clap:

I'm amazed at how strong the IMC seems to be on all of them. If you don't mind what NBvid range were you ~ on the air runs?

My 550 is a sweet chip on the cores, but the IMC seems to cr*p out whenever I get over 2700 (not that I'm complaining it only cost me $95 :D)...

FlanK3r
08-26-2009, 07:55 AM
Yes, great ,im looking for u and LinX stability

Smartidiot89
08-26-2009, 08:13 AM
Nice Chew*... Is that on phase or dice?

FlanK3r
08-26-2009, 08:34 AM
i think, he talk about phase air...

chew*
08-26-2009, 08:38 AM
Nice Chew*... Is that on phase or dice?

I have been on phase change since post #73

No money for DI atm :(

I need to send my other mach 1 out to be modded.........Wprime 1024M puts this mach 1 into cardiac arrest :shakes:

FlanK3r
08-26-2009, 08:57 AM
Chew, u will test today LinX ?One run is only 18-30 minutes...(its say, 14 000 and more is for stability PC)

chew*
08-26-2009, 09:04 AM
Chew, u will test today LinX ?One run is only 18-30 minutes...(its say, 14 000 and more is for stability PC)

Thats way last on my list of things to do man......I have ln2 prep work to do for the weekend........very little time and lots of stuff to test.

CobraXP
08-26-2009, 09:07 AM
Thats way last on my list of things to do man......I have ln2 prep work to do for the weekend........very little time and lots of stuff to test.
U da man!!! :up:

ELItheICEman
08-26-2009, 09:22 AM
U da man!!! :up:

Gotta second this. In the little time I've spent on XS, most of what I read is the excursions of this guy. I wish I had the money and time to do half of what chew* does :clap:

Keep it up man :up:

Tomasis
08-26-2009, 09:42 AM
with classic DR2 1066 is not possible, maybe 16.2s max:( And in 32M is much bigger diferent

yeah, 16.2s is not bad either. I need save money for 4gb and mobo.. because Im running only 2gb now :rofl:

i was talking about ddr3

G0ldBr1ck
08-26-2009, 03:25 PM
Im just not seeing as much of a gain from 955 to 965 as there was from 940 to 955. Still tho, its good to see that there is some improved headroom none the less.

chew*
08-26-2009, 03:30 PM
Im just not seeing as much of a gain from 955 to 965 as there was from 940 to 955. Still tho, its good to see that there is some improved headroom none the less.

When comparing to my best binned 955 chip not so much ( 1 in 1000 chip ), when comparing to the majority of chips the gaps rather large take a look.

NB improvement is really what I notice the most.

http://www.xtremesystems.org/forums/showpost.php?p=3980472&postcount=293

G0ldBr1ck
08-26-2009, 03:37 PM
Well It looks like a good improvment on the NB but still there was a larger amount of headroom moving from 940 to 955 on the core. Just your avarage joe overclocker saw ~200mhz. If your getting an average ~100mhz from 955-965 on phase. I would guess the average joe will be seeing ~50mhz on air.

Dont get me wrong, the chips are awsome. Theres just allot of folks wondering if a move from 955 to the 965 would be worth it and im just evaluating what im seeing. If anyone will milk out those extra Hz it will be you im sure.
Maybe you could give us your idea of what the average joe might expect for gains?

chew*
08-26-2009, 03:52 PM
Well It looks like a good improvment on the NB but still there was a larger amount of headroom moving from 940 to 955 on the core. Just your avarage joe overclocker saw ~200mhz. If your getting an average ~100mhz from 955-965 on phase. I would guess the average joe will be seeing ~50mhz on air.

Dont get me wrong, the chips are awsome. Theres just allot of folks wondering if a move from 955 to the 965 would be worth it and im just evaluating what im seeing. If anyone will milk out those extra Hz it will be you im sure.
Maybe you could give us your idea of what the average joe might expect for gains?

Well from an upgrade point of view from 955 to 965 the gains are not rather large especially if you have one of the "good" chips.

I think that for a new purchase however and upgrading from chips that arent 955's or "good" 955's the 965BE will be the "safer" bet as the binning is most likely getting "tighter", etc we wont see to many more stellar 955's from this point on.

As far as gains, well benching at 4.gig is less strained now, preety sure most people can get to 4 gig benching rather simply.

I'm seeing roughly if I take averages 50-100 gains on core, 100-200 gains on NB, better voltage tolerance and oddly enough they seem to run cooler.

The chips seem a little more balanced.....I'm finding they don't like to validate as high or at least exceed previous valids however they seem to operate higher, meaning suicides are not stellar but effective clocks at what matters, tasks, gaming etc are improved.

Smartidiot89
08-26-2009, 04:10 PM
When 955BE was released we saw an improvement with pretty much all Phenom II's that was released in stores after it, even the "older" ones such as 720BE etc. as long as they were manufactured after the point which were 955BE was release.

This is AMD and Globalfoundries improving their 45nm SOI technology which still has alot in it and I really love the results and observations you've come up with Chew* :)

A new CPU for me will wait thou, until the new stepping arrives, but as Chew* says the 965BE has a much tighter binning and will reach best results most of the time unless you happen to get a golden CPU on the lower range

chew*
08-28-2009, 06:05 PM
It's DI time

Special Thx to Aaron schradin of ln2 cooling.com for the use of his pot yet again. This is the pots first time testing under DI. Looks like it does rather well ;)

http://chew.ln2cooling.com/965BE%20DI/IMG_1407.JPG

http://chew.ln2cooling.com/965BE%20DI/IMG_1408.JPG

http://chew.ln2cooling.com/965BE%20DI/IMG_1409.JPG

http://chew.ln2cooling.com/965BE%20DI/IMG_1416.JPG

Booj where's your backup :p:
http://chew.ln2cooling.com/qdig-files/converted-images/965BE%20DI/lrg_1m%20DI.JPG

Bill :D
http://chew.ln2cooling.com/qdig-files/converted-images/965BE%20DI/lrg_DI%20pifast.JPG

http://chew.ln2cooling.com/qdig-files/converted-images/965BE%20DI/lrg_DI%20wprime%2032m.jpg

http://chew.ln2cooling.com/qdig-files/converted-images/965BE%20DI/lrg_DI%20wprime%201024.jpg

Was working on optimizing my 3d run when I realized I was low on DI :( stopped 3d to boot a diff OS , and run 1m 32m and pifast, ran out for 32m.

http://chew.ln2cooling.com/qdig-files/converted-images/965BE%20DI/lrg_DI%20%203d06.jpg

Validated 5300 on 1.52.1 cpu-z :shakes: but it did 5300. need to load 1.52.2 on vista next round.

Buckeye
08-28-2009, 08:33 PM
Very nice stuff there and on DICE :)

You have a very lot of condensation where you live, now I can see why you prep the way you do.

So the temps you show on your meters are at ISH and base of pot ?
Pretty darn close if so.

Phantom pot seems to hold up pretty well under DICE, plus I bet its very nice to have such a big tall pot for that.

Looks like I have my work cut out for me as you beat my scores on DICE and now moving to LN2 HAHAHA :rofl:

Nice job Chew :up:

chew*
08-28-2009, 08:59 PM
Yep one probe on IHS other on pot, that was temps during 3d run.

I was actually taking video for research of temp delta for Aaron during 1024m run, I ran it quite a few times which ate up quite a bit of DI tonight.

Yep condenastion where I live is no joke. Asus board had a pool of water on it :D

Board is damaged or I would have played multi card, honestly surprised it was holding up as well as it did.

Booj
08-28-2009, 09:02 PM
Booj where's your backup :p:

hehehe i'll see what I can cook up ;)

Great results there on dice :yepp:

Buckeye
08-30-2009, 07:04 AM
Hey Booj I see your using a Cascade for your runs ?

Do you have any pictures on that unit ?

Booj
08-30-2009, 06:48 PM
Sure mate, check out Kayl's build log (http://www.xtremesystems.org/forums/showthread.php?t=97321&page=11) for this unit.

I love it, Kayl is a top notch builder. The cold bug switch using a hot gas bypass is simply excellent. This one is tuned for high heat loads too, it makes i7 benching a breeze.

Buckeye
08-30-2009, 06:58 PM
Oh yeah thats a sweet unit :up:

Mine works well on Intel chips also, when they dont cold bug on me :)
Its handled my 965XE and 920 D0 just fine, but some other chips I have arnt worth the bother for even DICE.

No hot gas bypass on mine but dont need it on AMD chips anyway. But some tests can bring it to its knees if I am not careful :)

ELItheICEman
08-31-2009, 06:21 AM
Nice run! I noticed you didn't really go higher than 1.400V on the CPU - did more volts not help at all, or did you just not try it?

chew*
08-31-2009, 07:11 AM
Nice run! I noticed you didn't really go higher than 1.400V on the CPU - did more volts not help at all, or did you just not try it?

I'm was using an Asus, vcore is reported wrong ;)

ELItheICEman
08-31-2009, 07:46 AM
Ah, that's lame. What was your vCore for those runs? under 1.6V?