Printable View
We should do the same with Olympians and sports teams too. Nah I'd say it's fair AMD used to brag clock for clock they could beat Intel. That was in the Barton days, 939 was great too. As consumers you should push for performance. A 6 core processor should not lose to the competitors budget dual.Quote:
Originally Posted by rintamarotta
How do u like this at FX?:)...:up: 2510 MHz at RAM
Attachment 117302
Sweet! Flank3r :eek: we want more :shocked:
That number doesn't say much without stating voltage. If it's within 1.4-1.5V then I like. :) Even without given voltage, I like the RAM clock. :)
So 3.7GHz base clock, or is that already with the +0.5Ghz?
Barton not only beat P4 willis clock for clock, it beat them alltogether, at lower price.So there was something to brag about wasnt it.For gods sake PIII`s often beat P4...Quote:
Originally Posted by Glow9
I know youre flamebainting Glow.But lets set something straight. X6 doesnt "lose" to budget dual intel.Yes ,it has higher IPC, so in some instances it will, but not overall, yorure just talking silly.
@Darthshader, its the same sample OBR has.Its OBR`s screenshot.B0 ES.Not any form of retail.So multi is irrelevant.
retail SKU... noQuote:
Originally Posted by DarthShader
that is interesting, i had the impression that they would never release the quad core thuban version..... seems like they now have enough broken cores??? could be a real high OC'er AMD wise.Quote:
Originally Posted by duron
It also stated 'overclock:34%' so it's 3.7Ghz at stock. Wonder if it is final retail or still immature ES.Quote:
Originally Posted by DarthShader
That should be easelly within 1.5volts at least in retail product.Quote:
Originally Posted by DarthShader
I'll leave the screens to Zambezi release but 4.8ghz 1.47volts, Noctua NH-D14 3-fan push-pull/push-pull config.
I doubt any ES's are yet 100% final retail products, there might be some differences but current samples ES's are way better than earlier ones.
Divide 269 by 200 and tell me what you get...Quote:
Originally Posted by undone
I can boot a 21x multi, and it will say 0% OC...
That sounds good. If final silicon can improve a bit on that, BD will be right where it needs, to be very competetive. :)Quote:
Originally Posted by rintamarotta
Maybe im off.But when i push my 1055T with 1.47v ,it starts to get HOT without of course extreme RPM`s on fans, and here we have 8 cores.I know that a lot depends on the design die area etc.Quote:
Originally Posted by rintamarotta
Could you tell us if this voltage is sustainable with not loud high end air cooling ? In some hard load of course like primex64.
rinta, can i move in with you for a few weeks? :D
We are talking 32nm vs 45nm here and a completely different core design...Quote:
Originally Posted by XRL8
For all we know, this CPU may tolerate up in the 75-80c range like i7 does.
Yeah it COULD.However when you look at past node shrinks, it has always come with lowering requiered voltage.32nm at 45nm voltage (at which it is very hard to keep stability due to heat) is more likely to be unstable.Quote:
Originally Posted by BeepBeep2
As for Intel comparison ,intel doesnt use SOI.And runs at lower voltage.I dont think correlation is proper.
When you take all things into consideration its less likely to be free of heat related problems than X6 at same voltage.Of course full load.Power gating should help a lot in more mundane tasks.
I personally deem "overclock" mhz wise ,when cpu is FULLY stable under prime for several hours.And heat is definietely an issue for most cpus.3x 3000Rpm Kazes arent real world "usable" too.Thats why im asking.
Nice, what about NB speeds? Will we easily 4000Mhz nb air cooled?Quote:
Originally Posted by FlanK3r
IMC setup is probably a lot different than what we see in Phenom II I would guess.Quote:
Originally Posted by Barr3l Rid3r
ok, but seem they still working w/ same clock base of 200Mhz... so NB should have compatile numbers. I saw it will have standard 3200Mhz NB clock. But what about headroom?
At this point in time BD NB clocks are only numbers for us. We don't know what was changed in MC in BD. Yes Thuban's like 2.6ghz NB speeds to use 1600mhz ram effectively. But BD MC can be more efficient and might not need those clocks. Or BD MC might be clocked higher than Thuban NB and thus there will be no need to clock it even higher. It might be the case that clocking NB will not give us any benefits, like todays HT link in thubans, which when you clock higher than stock even decreases performance sometimes.Quote:
Originally Posted by Barr3l Rid3r
So I suppose knowing how high can CPU clock go is important for us OCers, but we cannot say the same about NB clocks.
Its litlebit cooler than 1090T OC'd to 4.1Ghz with 1.47volts, will not say more.Quote:
Originally Posted by XRL8
Node shrink helps in that.
I can match that but with 8GB. :p:Quote:
Originally Posted by FlanK3r
hehe, nice bingo :)
Oh man flanker why didn't you tell us this was from OBR.:( Wasting my feelings at last. Perhaps ur still trust OBR at least sth, but I don't.Quote:
Originally Posted by FlanK3r
http://obrovsky.blogspot.com/2011/07...r3-on-amd.html
undone: in this case u can, believe me. If I had BD (or have :) ), until NDA could not show here :(...
I was trying to determine if thats a realistic 24/7 OC, if its the way you say, than its on the verge, but still doable without extreme measures.Thats good to hear.Quote:
Originally Posted by rintamarotta
Anybody know how its with sandys ? I know many people claim 5ghz and even up, but is that stable and is is sustainable is day to day usage ?
XLR:some SB chips are 5 GHz stable with good voltage (1.35V), of course, more of this can 5 GHz stable too, but with 1.45-1.55V. But dont forgot, SOI HK is diferent than Intels HK. I thinking, 1.45V at AMD will be still very safe for 24/7. More dangerous are temps than volts.