AMD - Piledriver chitchat

[CAT-THE-FIFTH]

SR to be released this year:

http://www.xbitlabs.com/news/cpu/dis...e_in_2013.html

Edit!!

It is from this presentation:

http://translate.google.co.uk/transl...d%3D1364293981

It looks Excavator is not canned either.

Ruby Reloaded:

http://www.pcgameshardware.de/Grafik...1&i_id=2057243

Crytek seems quite pally with AMD ATM!!

Edit!!

One of the chaps working for OcUK,has hinted the Piledriver refresh might be clocked very high??

He mentioned 5GHZ??!! I wonder though if that is the stock maximum Turbo clockspeed or its OC capability??

The A10-6800K is already clocked at a maximum of 4.4GHZ under maximum Turbo.

[watercooled]

Meaning the AM3+ PD replacement? 5GHz does seem very high, especially for a base clock. If it wasn't for the already fairly high clocks of the uarch I'd say it's unlikely on 32nm but who knows. It's probably some sort of boost clock though.

[CAT-THE-FIFTH]

Some Richland scores leaked:

http://translate.google.co.uk/transl...ed=0CDYQ7gEwAA

If these scores are true the A10-6700 is faster the A10-5800K in 3DMark and consumes less power too.

[HalloweenJack]

I do expect it to be faster since its clocked higher as is the gpu side


edit:

http://www.3dmark.com/fs/240666


my kids pc , 1055T X6 , with an Asus 6670 GDDR5 card!

gpu score of 1309 vs 1212 for the APU

[CAT-THE-FIFTH]

I do expect it to be faster since its clocked higher as is the gpu side


edit:

http://www.3dmark.com/fs/240666


my kids pc , 1055T X6 , with an Asus 6670 GDDR5 card!

gpu score of 1309 vs 1212 for the APU

Its clocked 10% higher than the IGP in the A10-5700 and 5% higher than the IGP in the A10 5800K. It also has a CPU clocked 10% higher than the A10-5700 and around the same as the A10-5800K.

So basically as good,or better performance than an A10-5800K in the same TDP as the A10-5700.

BTW,if that 3DMark11 score translates into realworld performance,that would be actually quite decent for an IGP:

http://www.tomshardware.co.uk/radeon...w-32169-5.html

A good pictures of a test chip using interposers:

http://semiaccurate.com/2013/04/02/g.../#.UVrpYVfC7lh

Interesting:

http://blogs.barrons.com/techtraderd...cuitry-issues/

http://blogs.barrons.com/techtraderd...k-and-on-time/

[Biscuit]

Bet they have some serious runt frame issues with that much of a crossfire array ;D

[CAT-THE-FIFTH]

Bet they have some serious runt frame issues with that much of a crossfire array ;D

The runt of framerate litter causing problems!!

Edit!!

I just realised they used over 70 cards!!!

[CAT-THE-FIFTH]

More PC gamer rage incoming:

http://www.forbes.com/sites/erikkain...ect-gaming-pc/

Also,about the PS4 UE4 demo:

http://www.eurogamer.net/profiles/briankaris/comments

[Biscuit]

More PC gamer rage incoming:

http://www.forbes.com/sites/erikkain...ect-gaming-pc/

Seems like mostly damage control to me as his previous comments where met with some hostility

[shaithis]

The whole thing reads like a gigantic backtracking......and no wonder, he didn't know what he was talking about to begin with!

Goes to show how out of touch VPs can be and why they probably shouldn't make public statements about technical stuff.

[CAT-THE-FIFTH]

It looks like Adobe Premiere Pro is finally getting OpenCL support:

http://techreport.com/news/24616/ado...o-premiere-pro

http://www.anandtech.com/show/6881/o...ro-for-windows

Some of the performance gains look impressive.

Edit!!

These AMD reference tablets look great too:

http://www.xbitlabs.com/news/mobile/...in_Action.html

[CAT-THE-FIFTH]

I saw this mentioned on overclock.net:

http://us.battle.net/wow/en/forum/topic/4079871855

!!!!

Can anyone confirm this??

Supposedly,you can do it using this method:

http://www.overclock.net/t/1377248/x...#post_19688225

[watercooled]

That's pretty disturbing if true, although it seems like it could be methodology errors as surely a performance hit that large would've been picked up by AMD or others?

[DanceswithUnix]

Supposedly,you can do it using this method:

http://www.overclock.net/t/1377248/x...#post_19688225

It's been a long day, where is the method?

[DanceswithUnix]

Hmm, Jaguar+

http://www.fudzilla.com/home/item/30...aguar%20-cores

AMD seem to have really embraced the whole change to highly synthesizable cores and are cranking out core updates quite fast.

[kalniel]

Some titbits hinting at steamroller perhaps:

http://www.rockpapershotgun.com/2013...s/#more-148706