AMD Reveals Ultrathin Prototype, Roadmaps

[vickybat]

AMD's roadmap looks interesting. There's a variety of piledriver cpu's and apu's coming up. But vishera supporting piledriver is 32nm and will launch in 2013 according to the following article. Have a read guys:

Source

 

[coderunknown]

AMD did the right move by not chasing Intel into the smartphone SOC. Hondo or even the 2W APU looks a bit too power hungry for a tablet, but those should be idle for ultrabooks or ultrathin netbooks. And the part i liked, AMD may produce SOC based on ARM architecture. maybe Hondo + ARM.

Also AMD needs to either lower Brazos pricing or offer a better performing core. at 20k, it challenges Intel 2nd gen Dual core. Also high TDP at 17W, low processor clock (& hence performance). Brazos performance won't increase a lot with increase in clock speed, still a 10W, 2Ghz dual core Brazos is what i am expecting as Brazos 2.0. or at least 2Ghz under turbocore.

according to Xbit labs, this is what we are to expect from Brazos 2.0:

At present AMD readies several flavours of Brazos 2.0 offerings for netbooks, notebooks and nettops. For example, models E1-1200 (two cores at 1.40GHz, 1MB cache, Radeon HD 7310 graphics engine with 80 cores at 500MHz, 18W, etc.) and E2-1800 (two cores at 1.70GHz, 1MB cache, Radeon HD 7340 graphics engine with 80 cores at 680/523MHz, 18W, etc.) will target nettops and low-power notebooks.

i was expecting even lower TDP. though if it can challenge Pentium Dual Core, it'll be a win-win situation for AMD.

 

[vickybat]

^^ Did you read that piledriver FX cpu's are delayed to 2013?? They will be 32nm. I thought that was more important.

 

[coderunknown]

that will be a problem for AMD. Intel jumped from 45 to 32 & then to 22nm. AMD is moving from full node to half node and repeating same thing again. but i feel they "may" have taken Intel's tick-tock strategy for the FX series.

1st gen FX: new architecture. new node.
2nd gen FX: new architecture. old node.
3rd gen FX: old architecture. new node (according to me: 22nm, should be around H2 2014).

AMD is more focused on APU now. Which is the correct strategy as Intel will continue to price their processors higher than normal. If a 5k APU can offer ~70% the performance of a 10k intel processor & 5k GPU, peoples will definitely get attracted to such a deal.

Spoiler

*images.anandtech.com/doci/5503/Screen%20Shot%202012-02-01%20at%202.14.16%20PM.png

Piledriver is late but if they can deliver what made Phenom II famous, they are somewhat safe. But Intel too isn't sitting idle waiting for AMD's next move.

 

[vickybat]

^^ I think they are testing their cpu architecture with an APU. Did you see a pattern there? Trinity will first drop in with piledriver cores but vliw4 gpu this year. Next year piledriver FX cpu's will be launched but alongwith that, amd will unveil Kaveri APU's with steamroller( new architecture) cores + GCN gpu. Then in 2014 maybe, we will see steamroller based FX cpu's.

Unusual roadmap i must say. Yes they are stressing more on fusion and eying the general public. They want to generate more revenue and that's important for the future of the company.

 

[coderunknown]

they have already lost the performance game to Intel. Gunning for the highend is meaningless. So they are trying their best to seize back the sub15k (or at least sub10k) market.

Yup, what you said makes lot more sense. Design an architecture. test it with APU. in the meantime tweak it & wait for manufacturing process to mature. then replace the GPU part with more transistors (FX should have almost same die size as APU) and release those as FX. well planned, only if AMD can execute the plans not themselves

 

[sukesh1090]

are you sue that pildrivers are delayed to 2013?

 

[vickybat]

^^ Read the link

 

[sukesh1090]

^^i read that but i got confused as there are lot of architecture mentioned there.if i am right vishera is piledriver isn't it?

 

[vickybat]

^^ Yes, you got that right mate.