AMD Piledriver rumours ... and expert conjecture

[Reynod]

We have had several requests for a sticky on AMD's yet to be released Piledriver architecture ... so here it is.

I want to make a few things clear though.

Post a question relevant to the topic, or information about the topic, or it will be deleted.

Post any negative personal comments about another user ... and they will be deleted.

Post flame baiting comments about the blue, red and green team and they will be deleted.

Enjoy ...

 

[viridiancrystal]

rumours of AMD 8xxx series : "Orlando"

8870 : same performance as a 7950, same number of transistors as a 7950. Lesser power, lesser price compared to 7870. BS ?

Lower price than the 7870 at launch maybe. Probably not where it is now. They could just rebrand everything down a tier and make a new chip for the 8970. I hope they aren't that lazy, but who knows 😛.

 

[cgner]

Nvidia has done it with 400-500 cards. They might do the same with 600-700 cards, just tier them down one step, get a new top GPU, and there u go. Like I said, it would not be surprised if AMD did the same.

 

[fazers_on_stun]

AMD is dead....

amd is not dead

After reading all the non-PD posts, I am dead... 😛

 

[jdwii]

I'm sure the 8970 will be around 20% faster while using less power then the 7970. Remember the real reason why nvidia has better Performance/ Power consumption this gen they waited(lost lots if sales doing so) until TSMC had a more advance 28nm die and Amd is still using their older 28nm die design now with the Radeon 8000 series they will have their newer design.

 

[de5_Roy]

if you guys're gonna talk about radeon 8k series, try this thread...
http://www.tomshardware.com/forum/346081-33-radeon-8000-series-complete-discussion
btw, radeon 8k will need to bring in some serious improvement if they want to compete with kepler 2.0/gk110/whatever (afaik no maxwell until 2014). kepler's upper mid range asic (gk104) competes head to head with highest end gcn asic while being more pwoer efficient.
edit 2: and please, decent launch drivers on time, amd. please.

 

[esrever]

It seems like rumors and leaks always slow down as the launch approaches.

 

[amuffin]

Everybody is leaving...

http://www.marketwire.com/press-release/amd-announces-chief-financial-officer-transition-nyse-amd-1702366.htm

 

[gamerk316]

And their stock plunged 12% as a result.

 

[jdwii]

And their stock plunged 12% as a result.

If Amd had such great people they would be in a better position with Intel. Even though this looks bad, Amd needs to do this to involve and improve their products. The only person i think they should of kept was the one that dealt with the console players since i think he was doing a great job. They need to get new enginners or some of the great ones back and go back to hand craft designs.

 

[amuffin]

I don't think they have the money! 😛

 

[sarinaide]

Wow they lose a Hector Ruiz pocket lining cronie and its the end of the world. Did nobody remember like the x number of Intel and ex AMD engineers they brought back?


Again all these engineering samples are fake, I mean photoshoped, I mean a bag of balls. No FMA3 and FMA4 supported instruction sets? latter photoshopped in....epic fail.

 

[viridiancrystal]

http://wccftech.com/amd-shows-running-samples-vishera-fx8350-trinity-a105800k-apu/

http://muropaketti.com/amd-esitteli-vishera-koodinimellista-fx-8350-prosessoria-ylikellotettuna

Piledriver still has some overclockability?

 

[-Fran-]

1.45v and 4.816 Ghz... Not bad, but I'd say it still seems a bit high voltage wise. Wonder how heat and power goes if that is true.

Cheers!

 

[cgner]

I don't think they have the money! 😛

another intel fanboy troll

1.45v and 4.816 Ghz... Not bad, but I'd say it still seems a bit high voltage wise. Wonder how heat and power goes if that is true.

Cheers!

1.45-1.50 is where phenom II would work when overclocked. PD should not have to reach that.

 

[viridiancrystal]

1.45v and 4.816 Ghz... Not bad, but I'd say it still seems a bit high voltage wise. Wonder how heat and power goes if that is true.

Cheers!

I'm am hopeful that the voltage is a little higher than it will need to be, and it could be that the chip that was being used isn't very well binned, as I assume they are binning good chips in prep for launch.

another intel fanboy troll



1.45-1.50 is where phenom II would work when overclocked. PD should not have to reach that.

Phenom II's barely over 4 Ghz overclock 😛

 

[esrever]

Im guessing that chip is running as stable as possible cause they don't want a BSOD at a demo with even the slightest chance. It does look like piledriver won't OC as greatly as bulldozer tho, 4ghz base would be hard to push 4.8 on air where the 3.3ghz bulldozer could do 4.5 on air.

 

[viridiancrystal]

Same arch, same wall. However, a 4.5Ghz Piledriver should handily outperform a 4.5Ghz Bulldozer. That is the hope of all of us anyway XD

 

[guskline]

1.45v and 4.816 Ghz... Not bad, but I'd say it still seems a bit high voltage wise. Wonder how heat and power goes if that is true.

Cheers!

I agree! I have a 8150 solidly OC'd at 4533 (21x216) at 1.4 v and could get more if I had higher voltage but the heat when running IntelBurnTest is too high. I wonder if that voltageon the 8350 is set too high or that is a rough engineering sample? 4.8Ghz is great BUT not at 1.452v!

 

[guskline]

Im guessing that chip is running as stable as possible cause they don't want a BSOD at a demo with even the slightest chance. It does look like piledriver won't OC as greatly as bulldozer tho, 4ghz base would be hard to push 4.8 on air where the 3.3ghz bulldozer could do 4.5 on air.

I'm hoping PileDriver is an improvement, but from my recent acquisition and OCing of a 8150 I note the following.
1. The 8150 was stock clocked at 3.6 Ghz (not 3.3ghz) and supposedly could OC turbo on all cores to 3.9 and on up to 3 or 4 cores to 4.2. To run a 8150 STABLE at 4.5Ghz on air requires a top notch Noctua cooler or equivalent. I'm using a Corsair H100 which is their top end liquid cooler and 4.5Ghz is attainable at a stable state and with reasonable temps even when running Prime95 etc.
2. If the PileDriver is running stable at 4.8 Ghz and it should, I'm a little surprised that it needs 1.452v. Perhaps the resonant mesh Tchnology allows a higher OC but rquires more voltage. I noticed a high multiplier, 24.

All things considered, if the PileDriver allows a solid satble 4.8 OC with Liquid Cooling, it's an improvement, just not a significant one. We will have to await the official release.

 

[jdwii]

http://wccftech.com/amd-aseries-trinity-apus-pricing-specifications-detailed/


Amd is out of their mind a A10 should be priced at the I3 not low-end I5's. It's like they want to go bankrupt. Next their going to price the 8 core Piledriver at 320$.

 

[Smeg45]

For an office/Facebook box most don't even need any APU's. A $50 G530 would be enough. I'm still kicking myself for spending an extra $80 overall on an 3870K when I would have been satisfied with a G530.

 

[jdwii]

All i know is some one can get a Pentium Dual core and a Radeon 6670 for 165$. And they will use around the same power consumption and they will have a upgrade path. Plus if this is priced at 165$ the 4 core Piledriver will at least cost 150$. Like chad said a year ago with bulldozer i smell one hell of a price cut on these after the first 3 months.

 

[fazers_on_stun]

Im guessing that chip is running as stable as possible cause they don't want a BSOD at a demo with even the slightest chance. It does look like piledriver won't OC as greatly as bulldozer tho, 4ghz base would be hard to push 4.8 on air where the 3.3ghz bulldozer could do 4.5 on air.

Hmm, my bet is the resonant clock mesh (RCM), which I mentioned many pages ago 😀.

But this is an engineering sample, right? Production hardware should be better..

 

[-Fran-]

I'm hoping PileDriver is an improvement, but from my recent acquisition and OCing of a 8150 I note the following.
1. The 8150 was stock clocked at 3.6 Ghz (not 3.3ghz) and supposedly could OC turbo on all cores to 3.9 and on up to 3 or 4 cores to 4.2. To run a 8150 STABLE at 4.5Ghz on air requires a top notch Noctua cooler or equivalent. I'm using a Corsair H100 which is their top end liquid cooler and 4.5Ghz is attainable at a stable state and with reasonable temps even when running Prime95 etc.
2. If the PileDriver is running stable at 4.8 Ghz and it should, I'm a little surprised that it needs 1.452v. Perhaps the resonant mesh Tchnology allows a higher OC but rquires more voltage. I noticed a high multiplier, 24.

All things considered, if the PileDriver allows a solid satble 4.8 OC with Liquid Cooling, it's an improvement, just not a significant one. We will have to await the official release.

I wonder how the arch in AMD CPUs work voltage wise... And I'd really want to see power figures.

This could just be a coincidence, but I'll throw it for kicks:

Cheers! 😛

EDIT: Damn links, lol.

 

[tpzhei]

I had a 3ghz Core2Duo and a HD 5670 and I was able to play Crysis 2,Crysis Warhead,Metro 2033,Witcher 2 all at 1920x1080 medium to higher settings no AA/AF
your system is stronger than that one
so your system should be fine at 1920x1080 or below at medium to high settings no AA for any game

this is my Youtube channel "Gaming on a Budget" which has videos to prove it can be done

http://www.youtube.com/user/TheSpooky777/videos?view=0

@ king smp

thanks dude. should i get another 8gig ram? or is it overkill