AMD aircooling validation only - get it to the max

[FlanK3r]

1045 Thuban? I dont know much better Thubans than my 1015

[Mad222]

The newest batch in here is still 1023... Not many ppl want to buy 1090T as compared to 1055 because of the price
I think I need to wait @ least a few months to find newer batches in local shops...

Thx ! I'll keep that in mind, buddy

Should be chasing week 1045 madd.
It's better than 1015 core and NB wise.

Me too...!
If you can find 1045 in your country, get one with Vid (with turbo core on) below 1.4v
AMD has made improvements lately. Surprise....
I'll run LinX & Cine later when I get win7 64 bits installed, ok

1045 Thuban? I dont know much better Thubans than my 1015

[BeepBeep2]

Should be chasing week 1045 madd.

It's better than 1015 core and NB wise.

I don't believe my chip is any better than his 1090T

1.325 VID no turbo 4.3 3d stable, it will post 3250 NB but definately wont boot much over 3100 didnt try valid yet though, I must do this.

If you can find 1045 in your country, get one with Vid (with turbo core on) below 1.4v
AMD has made improvements lately. Surprise....
I'll run LinX & Cine later when I get win7 64 bits installed, ok

lol, I think he's seen and tested at least 4 or 5 1045 chips now and known of more, I got this one from Aaron Schradin through him

EDIT:
Maybe I should lend it to flanker or Mad so one of them can pull more magic out of it :p

[Mad222]

BeepBeep

1.325v default Vid ? I guess it's 1.475v Vid with turbo core on then!
Hmm...Dun think you can get much out of it

Usually 3300nb @ around 4.5g with 1.35v cpu nb on air. What ram divider you were using ?

It's much diificult to find a good Thuban & I dun think the 890fx boards are really 100% ready for that because of its' initial design...hope 9 series will do well

I don't believe my chip is any better than his 1090T

1.325 VID no turbo 4.3 3d stable, it will post 3250 NB but definately wont boot much over 3100 didnt try valid yet though, I must do this.

lol, I think he's seen and tested at least 4 or 5 1045 chips now and known of more, I got this one from Aaron Schradin through him

EDIT:
Maybe I should lend it to flanker or Mad so one of them can pull more magic out of it :p

[BeepBeep2]

BeepBeep

1.325v default Vid ? I guess it's 1.475v Vid with turbo core on then!
Hmm...Dun think you can get much out of it

Usually 3300nb @ around 4.5g with 1.35v cpu nb on air. What ram divider you were using ?

It's much diificult to find a good Thuban & I dun think the 890fx boards are really 100% ready for that because of its' initial design...hope 9 series will do well

1:4 1600 divider. In cold ambient on H20 (-6c) I was able to do 1M at 4.8, NB at 3297 and ram at 1886 6-6-6-18 1T

[chew*]

Beep when i discuss chips with madd it only concerns ln2, air does not always mean good on ln2...........

We went through a few chips although not yours but binned by vid.....there are lower vid chips out there, alot lower

my 1100T is better by 100 mhz core wise and 300 mhz NB wise in vantage versus my 1090T........so to each his own.

[Mad222]

Beep, there'll be not much diff. from 10c -0c for Thuban
4.8g with 3297nb isn't bad at all !

1:4 1600 divider. In cold ambient on H20 (-6c) I was able to do 1M at 4.8, NB at 3297 and ram at 1886 6-6-6-18 1T

Exactly...
Even with the same Vid...it behaves totally diff. under extreme cold ! (yet similar on air)

Beep when i discuss chips with madd it only concerns ln2, air does not always mean good on ln2...........

We went through a few chips although not yours but binned by vid.....there are lower vid chips out there, alot lower

[BeepBeep2]

Beep when i discuss chips with madd it only concerns ln2, air does not always mean good on ln2...........

We went through a few chips although not yours but binned by vid.....there are lower vid chips out there, alot lower

my 1100T is better by 100 mhz core wise and 300 mhz NB wise in vantage versus my 1090T........so to each his own.

Well I believe Aaron had this one on air, there was thermal paste on the CPU (albeit one side of the heatspreader) when it arrived, the middle one in the pic you posted. He said he got 4 Ghz on air like cake, which was very much so easy.

I just find it funny that Mad's 5g air chips are all 7g+ LN2 chips

[Mad222]

'Cause I put on my "presbyopic glasses"
Everybody has their own ways to cherry pick...as long as it works, that's good enough

I just find it funny that Mad's 5g air chips are all 7g+ LN2 chips

[Daveburt714]

Just curious as to how you folks are determining the vid of a chip.

This 1090 is slightly better than my 1055. When I do a clear CMOS and go into bios it shows CPU V's @ 1.257, after booting CPU-Z shows voltage @ 1.308 (idle).

For some reason my chip doesn't seem to like high v's on the IMC.
Anything over 1.3 causes a no post and even using TurboV to raise it above that voltage things get wonky...
OTOH, it will do 3.1 (stable) on the IMC with only 1.265v.

[BeepBeep2]

Just curious as to how you folks are determining the vid of a chip.

This 1090 is slightly better than my 1055. When I do a clear CMOS and go into bios it shows CPU V's @ 1.257, after booting CPU-Z shows voltage @ 1.308 (idle).

For some reason my chip doesn't seem to like high v's on the IMC.
Anything over 1.3 causes a no post and even using TurboV to raise it above that voltage things get wonky...
OTOH, it will do 3.1 (stable) on the IMC with only 1.265v.

Open up Core temp, it's right there

That's an aweosme nb speed

@Mad222
That same core that did 1M @ 4800 at -6c ambient will not validate CPU-Z at 4700 at 20c ambient...

[Daveburt714]

Open up Core temp, it's right there

Thanks Beep, according to CoreTemp it's 1.30.

[FlanK3r]

I tested only few aspect. 2 years old OS, max LINX 4300 MHz, max wprime 1024M 4400 MHz, max superpi 1M 4730MHz aircooled, ambinets about 20 C. Think, fresh OS can help me +50 MHz maybe for benchmarks as superpi and +10-20 MHz for LINX or wprime 1024M. Its possible?

[chew*]

Well I believe Aaron had this one on air, there was thermal paste on the CPU (albeit one side of the heatspreader) when it arrived, the middle one in the pic you posted. He said he got 4 Ghz on air like cake, which was very much so easy.

I just find it funny that Mad's 5g air chips are all 7g+ LN2 chips

And madd and myself both know that 7 gig without a strong NB is a useless peice of silicon other than a screenshot valid

Seen many many chips that had a nice NB on air top out at 3600-3800 on ln2..............

[I.nfraR.ed]

Hope my case is not the same. My Thuban does 3600+ on water (2 cores) and I'm going to check it on dice this weekend.
What's the normal CPU-NB voltage for dice, considering it wants about 1.45 for 3400 with 6 cores and 3600 on 2 cores?

[chew*]

Hope my case is not the same. My Thuban does 3600+ on water (2 cores) and I'm going to check it on dice this weekend.
What's the normal CPU-NB voltage for dice, considering it wants about 1.45 for 3400 with 6 cores and 3600 on 2 cores?

you've got about .100 for the safe threshold on ln2 left..........90% of chips refuse to work over 1.575 on ln2......

[BeepBeep2]

Hope my case is not the same. My Thuban does 3600+ on water (2 cores) and I'm going to check it on dice this weekend.
What's the normal CPU-NB voltage for dice, considering it wants about 1.45 for 3400 with 6 cores and 3600 on 2 cores?

I'd stop at 1.45v on DICE, I remember chew* telling me over livesteam that pushing NB volts on LN2 is not very productive

I could be mistaken though.

EDIT:
Nevermind, chew* beat me to it.

[I.nfraR.ed]

Thanks, I'll see if it scales with more voltage. I've used upto 1.5 on water, without any problems (maybe degradation, but haven't seen it yet). Probably I won't go past 1.5 on dice, too. Dice gave me about 500-600MHz increase on the 555 for the total of 3600MHz NB, but it's a different animal.

[FlanK3r]

heh, now understand...Chew thought 1045 is better under LN2 than 1015 ....

[I.nfraR.ed]

Ok, it seems pretty descent, I think. I was testing for stage 3 of the CountryCup and just made a simple validation. Note that the validation was not the idea, it's done with 4GHz+ uncore and memory over 1920 cl6, lol.

http://valid.canardpc.com/show_oc.php?id=1596298

It made 4GHz NB stable at about 1.5V, pretty close to what I expected. This time I had a good contact on the pot, but again high temperatures, I need to try with another liquid.

[Dk_vr]

One result from me too:

http://hwbot.org/community/submissio...20_4055.96_mhz

Could be better but I believe that's the maximum of my Athlon II x4 620.

[FlanK3r]

uffff, finally


http://valid.canardpc.com/show_oc.php?id=1624216

[theoldtimer]

Way to go FLanK3r.

[FlanK3r]

core 0 is good, other are whorse think...with stability and wprime or others hard loading benchmarks is my 1090T better.

[Manicdan]

how in the world did you get 5ghz validated on air