AMD updates product roadmap for 2014-15, Kaveri slated for Q1'14

Scorpus

Posts: 2,162   +239
Staff member

amd kaveri gpu roadmap cpu apu kabini volcanic islands beema

AMD has reportedly updated their product roadmap for 2014 and 2015, revealing when their next line of graphics cards and APUs will be launched. The report, sourced from the sometimes inaccurate DigiTimes, states that AMD's next-generation graphics cards, codenamed 'Volcanic Islands', will be released in late September, and new APUs will follow in Q1 2014.

Both the high-end 'Kaveri' and entry-level 'Kabini" APUs will enter mass production in February 2014 for an announcement in March 2014. On the Kabini front, AMD is expected to launch dual- and quad-core parts with Socket ST3 for notebooks and Socket FS1B for desktops. Kabini APUs will have a maximum power consumption of 25W, according to DigiTimes' upstream supply chain sources.

Kabini and Kaveri APUs were originally scheduled for release in the second half of 2013, although this report confirms a delay of the chips into 2014. The rescheduling of the launch will have an effect on AMD's future APUs, delaying 'Beema' (the successor to Kabini) into either the second half of 2014, or early 2015. Beema will allegedly use Socket FS1B like Kabini, although more Heterogeneous System Architecture (HSA) features will be implemented.

By the end of 2013, AMD will have begun phasing out their FM1 and AM3 sockets, with AM3+, FM2 and new sockets being utilized going forward. In 2015, 'Carrizo' APUs will be launched to succeed Kaveri in the desktop market, featuring the Excavator architecture at two TDPs: 45W and 65W. 'Nolan' chips will also be unveiled to replace Beema, if the reports are correct.

While AMD continue to push their desktop, laptop and low-power components, the company also scored an important contract to produce APUs for both the Xbox One and PlayStation 4. The semi-custom processors, based on the Jaguar architecture also used in the aforementioned Kabini chips, will form an important part of next-generation gaming.

Permalink to story.

 
Still not much listed on the release of the steamroller on AM3+ I see. Oh well, guess Intel Ivy Bridge-E it is then unless I start hearing more than just rumors sometime soon.
 
AMD's move of being not on head-to-head collision with INTEL is becoming a rerun tv show.
INTEL would release its newest products which are way faster than the competition at its own time of choosing and AMD would delay again and again its counter offer.

maybe it's time for the ADRENO company to swallow the hand that formerly feeds it. maybe ARM company can manage AMD better.
 
Wow why bother they have givin up to Intel. Hope all the AMD haters are happy now Intel will just keep doing what ever they whenever they feel like doing it.. Wish they were both doing well
 
Seriously, I think AMD is focusing on their APUs with GCN architecture. That will be interesting. I think the Steamroller FX series CPUs should come sooner, if not, I am expecting huge performance gains over current FX series CPUs.
 
What are you talking about, Kabini is already released!
You can find it from various retailers. Kabini and Temash are on the market now, albeit availability is low(no ITX mobos in the states yet, but Japan should have some now)

As for laptops(Kabini) just search A6-5200 or A4-5000 and you should find some for sale.Or you can see them listed on AMD's website(A6-5200 or A4-5000 or Temash A6-1450)

But yeah, Kaveri could be a long time coming, I just want an A6-5200 itx mobo to play with and I am seriously considering the temash 11.6" laptop with backlit keyobard for $400.
 
Seriously, I think AMD is focusing on their APUs with GCN architecture. That will be interesting. I think the Steamroller FX series CPUs should come sooner, if not, I am expecting huge performance gains over current FX series CPUs.
AMD is slowly but surely coming to realize that the CPU is a dead end for them. Their future now lies in the APU.
What is abundantly clear is that the Bulldozer (and by extension Piledriver/Steamroller/Excavator) was designed primarily from a server perspective, and without server sales ( Intel as of latest figures has 95.6% of the x86 server market) AMD has a serious issue with profitability versus R&D with this architectural line- which course involves chipsets as well the CPU µarch if it intends to stay competitive ( PCI-E 4.0 and DDR4 for example). You simply wouldn't have an FX line without Opteron, simply because the latter is where the money is. AMD's canning of Terramar and Sepang (20 and 10 core + PCI-E 3.0) and the complete disappearance of Excavator (CPU) from AMD roadmaps probably continues a story that started with Dirk Meyer's firing. Excavator will likely live only as an APU (Carizo)...and then at least a year late.
 
AMD is slowly but surely coming to realize that the CPU is a dead end for them. Their future now lies in the APU.
What is abundantly clear is that the Bulldozer (and by extension Piledriver/Steamroller/Excavator) was designed primarily from a server perspective, and without server sales ( Intel as of latest figures has 95.6% of the x86 server market) AMD has a serious issue with profitability versus R&D with this architectural line- which course involves chipsets as well the CPU µarch if it intends to stay competitive ( PCI-E 4.0 and DDR4 for example). You simply wouldn't have an FX line without Opteron, simply because the latter is where the money is. AMD's canning of Terramar and Sepang (20 and 10 core + PCI-E 3.0) and the complete disappearance of Excavator (CPU) from AMD roadmaps probably continues a story that started with Dirk Meyer's firing. Excavator will likely live only as an APU (Carizo)...and then at least a year late.

Eh, that stinks :(. I hope AMDs future is brighter since Intel is just stomping all over them now. What about the ARM server chips? Could that help AMD out?
 
Compared to Ivy Bridge-E, you probably shouldn't be expecting much competition from AMD anyway.
Wasn't really expecting huge competition except in the predicted steam roller cores being 20-25% better than Vishera made it sound appealing because at that improvement that would be a huge gain overall. However the Ivy Bridge-E for my needs is starting to sounds sweeter and sweeter.

Seriously, I think AMD is focusing on their APUs with GCN architecture. That will be interesting. I think the Steamroller FX series CPUs should come sooner, if not, I am expecting huge performance gains over current FX series CPUs.
The market for APU's is growing because of the fact having a balance of a decent CPU with a nice enough GPU can satisfy mobile a lot better. At the moment in mobile, if you do anything requiring any graphics power and your not willing to spend at least 800 or more bucks, then an APU is the only real option as HD 5000 is not even good enough.

APU will become popular mostly due to laptops and tablets as they really bring in more power in the GPU area which is a welcomed edition. In all honesty, this is probably the first time (Meaning these last 2 generations) that we have had chips in general that have on board GPU's (Or in the past chipset GPU's) that can actually do more than run a basic desktop setup.
 
Eh, that stinks :(. I hope AMDs future is brighter since Intel is just stomping all over them now.
AMD's future lies in leveraging what they're good at - namely graphics. APU (PC and console) means they are on a more level playing field. It also means that AMD have to push for increased parallelization in coding...unfortunately AMD have neither the cash, nor the inclination to put themselves at the sharp end of the technology. AMD's stance has always been to design the hardware and let others do the work in getting a software environment up and running
What about the ARM server chips? Could that help AMD out?
I damn well hope so, since that was the sole reason for buying SeaMicro. Unfortunately for AMD they also have some stiff competition in that area, even if you set aside Intel's Avoton business(which will operate in the same market) AMD still has to contend with Calxeda and Applied Micro (both already high profile thanks to Boston's Viridis and HP's Moonshot) amongst others that will likely include Nvidia, Samsung, and Qualcomm.
 
AMD's future lies in leveraging what they're good at - namely graphics. APU (PC and console) means they are on a more level playing field. It also means that AMD have to push for increased parallelization in coding...unfortunately AMD have neither the cash, nor the inclination to put themselves at the sharp end of the technology. AMD's stance has always been to design the hardware and let others do the work in getting a software environment up and running

I damn well hope so, since that was the sole reason for buying SeaMicro. Unfortunately for AMD they also have some stiff competition in that area, even if you set aside Intel's Avoton business(which will operate in the same market) AMD still has to contend with Calxeda and Applied Micro (both already high profile thanks to Boston's Viridis and HP's Moonshot) amongst others that will likely include Nvidia, Samsung, and Qualcomm.

Thanks for the info DBZ, I appreciate it.

Wasn't really expecting huge competition except in the predicted steam roller cores being 20-25% better than Vishera made it sound appealing because at that improvement that would be a huge gain overall. However the Ivy Bridge-E for my needs is starting to sounds sweeter and sweeter.
Intel owns the desktop CPU business. It is unfortunate, but real.
 
Intel owns the desktop CPU business. It is unfortunate, but real.

Yea, honestly they do, its hard but honestly if something happens, they could take it back but the way it seems, unless the APU's like the ones in the consoles have good enough CPUs to back up the (Close to) performance of the 7870 (On PS4) then its just going to be good in the mobile area.

Sad I wont get another great chip like an Athlon 64. Though my 8350 has also been great, I keep wanting more.
 
Yea, honestly they do, its hard but honestly if something happens, they could take it back but the way it seems, unless the APU's like the ones in the consoles have good enough CPUs to back up the (Close to) performance of the 7870 (On PS4) then its just going to be good in the mobile area.

Sad I wont get another great chip like an Athlon 64. Though my 8350 has also been great, I keep wanting more.

The 7870 is the GPU on the PS4?
 
Compared to Ivy Bridge-E, you probably shouldn't be expecting much competition from AMD anyway.

Buddy are you aware that AMD is about to up the ante on gaming with top line GPU's running CPU tasks natively? With far superior bandwidth of truly cross chip shared memory equal access. This will give each chip near instant access to code written for/by either CPU or GPU. So they will no longer have to depend on slower API bottlenecks and all code will be routed to the most appropriate processor for the job at hand.

This compared to both Nvidia and Intel having to depend on SOFTWARE API's in order to run CPU tasks on the FASTER.... GPU. AMD has just claimed the crown for the FUTURE OF HETEROGENEOUS COMPUTATION...... before INTEL even has a chance to get their competitor even out the door!

Because AMD were much smarter by buying ATI in the first place. Knowing that all in one SoC's were the FUTURE of Heterogeneous Computing w/ greater bandwidth available in upcoming chips/APU's. So what's Intel got in the GPU department to compete with either Nvidia or AMD's ATI Graphics Processing Prowess? Not a dam thing and so they simply don't stand a chance with their better late than never approach only now stepping up investment in Graphics Processing Department!

Links that provide better explanation of just where AMD is going with HSAIL and HSA GPU/CPU on same die using same memory space for a true generational shift in shared computing workloads;
http://www.extremetech.com/gaming/164817-setting-hsail-amd-cpu-gpu-cooperation

http://devgurus.amd.com/thread/159457
 
AND to add to my previous comment, AMD is rolling the dice by locking up all consoles with this new approach to heterogeneous computing that is about to make game developers very happy. Because games written using HSAIL will for once be fully cross platform compatible. In other words this is bound to make AMD LEAD PLATFORM for developers and will make it a breeze to port these console games to AMD Desktops where Intel won't have that option.

Meaning that if a developer could easily port PS4 Games to AMD's Desktop and Notebook APU's, but they wouldn't work on Intel based dedicated CPU and GPU Desktops. Since only AMD has fully committed their future to APU same die Architecture!
 
AND to add to my previous comment, AMD is rolling the dice by locking up all consoles with this new approach to heterogeneous computing that is about to make game developers very happy.
Obviously they are ecstatic.
Epic Games’ Tim Sweeney and AMD don’t see eye-to-eye on hUMA
Game developers will always take the easy option when it comes to coding- it is why they so readily money-grab from AMD's and Nvidia's gaming programs. The rest of your post sounds like a poorly constructed press release- you got the buzzwords and phrases noticeable with the ALL CAPS trick, although you are missing the all-important bullet points and AMD's patented "Future Thinking™" and "Forward Looking™" .

You also might take note of a couple of other points:
1. HSAIL is primarily aimed at server/compute.
2. HSAIL isn't really a language - it's an intermediary between existing languages
3. Intel hasn't joined the HSA Foundation partly because it has its own approach to unified memory access
4. Nvidia hasn't joined the HSA Foundation partly because it has its own approach to unified memory access (ARM+GPU)
5. The HSA Foundation members are AMD, Texas Instruments, MediaTek, and Imagination. They need a big partner ( Google, Apple etc) to make the leap- if not it will be a repeat of the (broken) promises of OpenCL all over again...remember when OpenCL was supposedly going to sweep all before it ?
 
AMD stealing patented names of an Indian Turbofan jet engine and its core
 
Back