• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

AMD Demoes Trinity already.

And got a [H] Gold Award too. So theyre off to a good start.
 
What socket is that going to use? I mean it can't use AM3+ since there is no available pins for video.
 
You'd have to redefine tick-tock. With Intel the tick is a new manufacturing process and the tock is the architecture built around it. If you wanted to do something like this on AMD, it really depends on where you want to start. You could choose tick to be the CPU architecture and the tock to be the integration of that into an APU which is what I'm leaning towards. In this sense, you'll probably see a tick and tock simultaneously. Example: Llano could be considered a tock for Stars while BD would be it's own tick.

Either way this is forcing a different company's model and just ends up being confusing >_>
 
K10-based Llano was a "low risk" first product. It wasn't meant to last a long time. I wouldn't call the replacement so soon a tick-tock cycle, but it's something that's definitely needed sooner than later on both the desktop and mobile. K10 is hopelessly uncompetitive in CPU performance at the price points AMD want to target Llano.

That Trinity will get a faster GPU is also great. The 50-100% lead Llano's GPU enjoys (using commodity memory) over HD Graphics 3000 right now will significantly narrow when Ivy Bridge is released in several months. An upgraded GPU in Trinity should restore a similar lead vs the IB GPU.

If you want to explore tick-tock, GF won't have a 22nm process until 2013-2014, and who knows when it will be production ready (mid to late 2014 is optimistically realistic IMO). There are many changes coming for AMD at 20/22nm: it's a bulk (strained) silicon process instead of SOI, AMD is moving to restricted design rules (Intel did that a couple of generations ago) and GF is dumping gate first (and using gate last like Intel)... all while moving to a new core at an accelerated pace? No offense to hope, but that's either total fantasy or a recipe for "unexpected delays." :p
 
SicKlown42012 said:
FMx(notice that only part of it is capitalized) is just a place holder name. They have not publicly named the socket for the next Fusion/APU(Trinity) or CPU only part(Komodo) and their internal name hasn't been leaked. Both are shown as FMx, but they will not share a socket.
Click to expand...

I hope they announce this by he time the Lynx boards start showing up as I would seriously consider grabbing a Micro ATX or Mni ITX board and throwing an A8 in there. I may go ahead and get one for a sff or HTPC build regardless but I would feel a lot better about it if I knew that a year from now I could just drop in a new Trinity CPU.
 
pxc said:
No offense to hope, but that's either total fantasy or a recipe for "unexpected delays." :p
Click to expand...


Woot more delays and so we can expect 5 more years for every advancement now from AMD (like the bulldozer) because at AMD they don't know what they are doing?

Everything is a total fluke, wait yes wait wasn't AMD ahead of certain design problems when they released the 4770 , yeah that was AMD ...
 
SicKlown42012 said:
FMx(notice that only part of it is capitalized) is just a place holder name. They have not publicly named the socket for the next Fusion/APU(Trinity) or CPU only part(Komodo) and their internal name hasn't been leaked. Both are shown as FMx, but they will not share a socket.
Click to expand...


AMD tradition, current APU is socket FM1 trinity will be FM2. its not like they would give it some crazy name.


Pieter3dnow said:
Woot more delays and so we can expect 5 more years for every advancement now from AMD (like the bulldozer) because at AMD they don't know what they are doing?

Everything is a total fluke, wait yes wait wasn't AMD ahead of certain design problems when they released the 4770 , yeah that was AMD ...
Click to expand...

quit feeding the troll.
 
pxc said:
K10-based Llano was a "low risk" first product. It wasn't meant to last a long time. I wouldn't call the replacement so soon a tick-tock cycle, but it's something that's definitely needed sooner than later on both the desktop and mobile. K10 is hopelessly uncompetitive in CPU performance at the price points AMD want to target Llano.

That Trinity will get a faster GPU is also great. The 50-100% lead Llano's GPU enjoys (using commodity memory) over HD Graphics 3000 right now will significantly narrow when Ivy Bridge is released in several months. An upgraded GPU in Trinity should restore a similar lead vs the IB GPU.

If you want to explore tick-tock, GF won't have a 22nm process until 2013-2014, and who knows when it will be production ready (mid to late 2014 is optimistically realistic IMO). There are many changes coming for AMD at 20/22nm: it's a bulk (strained) silicon process instead of SOI, AMD is moving to restricted design rules (Intel did that a couple of generations ago) and GF is dumping gate first (and using gate last like Intel)... all while moving to a new core at an accelerated pace? No offense to hope, but that's either total fantasy or a recipe for "unexpected delays." :p
Click to expand...


Speaking of unexpected delays, one delay you didn't expect apparently is that of the Ivy Bridge. It's been delayed to Mar-April of 2012 and not in "several" months like you claim.

GF may not have 22/20nm ready in 2012 but TSMC will have 28nm ready in several months, which puts it ahead of Intel for a short time.
 
sirmo said:
GF may not have 22/20nm ready in 2012 but TSMC will have 28nm ready in several months, which puts it ahead of Intel for a short time.
Click to expand...
AFAIK AMD is not using GF for 28nm CPU production, and AMD has not mentioned making 28nm Bulldozer derived or Llano derived chips. Notice in that link that AMD's first 28nm CPUs, next gen Bobcat models, aren't coming until next year. The next BD is still on 32nm and after that 20nm. The target for the 20nm APU is late 2013, which is the typical optimistic date. I can't remember the last time AMD was not several months late on anything.

I guess not many people know, but Intel and Micron have a 25nm fab (IMFT in Utah) and 25nm NAND Flash memory have been shipping for a while, like a year. Maybe I'm not up on what constitutes "ahead", but year old 25nm is earlier/ahead of not ready 28nm, right? And in 2012, 22nm TriGate is more advanced than a 28nm planar process?

Maybe you could quibble over the definition of several, but March/April 2012 is certainly several months away from now. :rolleyes: Want to suggest a more appropriate word, or is that just content-free bickering?
 
Last edited:
Several: More than two but not many (actual definition)

Ivy Bridge is about 9-11 months away (almost a year away). Only a biased Intel troll would use a term several months away to describe Intel next gen CPUs release timeframe.
 
sirmo said:
Several: More than two but not many (actual definition)

Ivy Bridge is about 9-11 months away (almost a year away). Only a biased Intel troll would use a term several months away to describe Intel next gen CPUs release timeframe.
Click to expand...

PXC isn't exactly everyones favorite AMD optimist but he's far from a troll. He simply looks at things from a logical perspective, and it's just unfortunate that shit always looks so gloomy for AMD.

Though, he was kinda harsh on Llano.
 
I find the name calling instead of making technical points kinda funny. That says it all. :D

TheBlueChanell said:
kinda harsh on Llano.
Click to expand...
I have really mixed feelings on it. As a product overall, it's definitely a win for AMD, but not a home run.

I took a lot of abuse here for pointing out what Llano's CPU competition would be at release and that part of Llano is a definite shortcoming as we see in the reviews. The graphics performance is good, but still relatively low end for gaming. I don't think that's a major problem, but it didn't really live up to the expected hype.

One major disappointment is GPGPU performance. What AMD provided to reviewers, a GPU accelerated video encoder meant to demonstrate the advantage of the APU, was incredibly underwhelming. The 400SP GPU encoded video slightly slower than software based CPU encoding on a dual core i5 SB CPU. Worse is that i3/i5/i7 SB models have dedicate hardware "QuickSync" which is 3x-4x faster than even that on the same video encoding test.

I would make a ConFusion joke about that, but it's kind of sad. It should perform much better and I'm still looking forward to when that Fusion promise is fulfilled. Right now Llano's on-die GPGPU capabilites are a major disappointment.
 
Last edited:
Yes, the reviews we have seen of llano do show lackluster CPU performance, which was... expected to some extent? It's a little worse than I thought it would be though.

the MX variants should help that some(the best one is 1.9 non turbo vs 1.5) but at the cost of TDP. I think we'll still get decent battery life from the 45w versions despite the 10w extra. Why? power gating :)

That said, all of HP's new laptops that offer llano offer the MX variants. I think that says something.
 
pxc said:
Right now Llano's on-die GPGPU capabilites are a major disappointment.
Click to expand...

QuickSync is not much more than a gimmick, currently.. Sure It's even twice as fast as nVidia's GTX 460, but it's only supported in: Cyberlink’s Media Espresso 6 and Arcsoft’s Media Converter 7.

http://www.anandtech.com/show/4083/...-core-i7-2600k-i5-2500k-core-i3-2100-tested/9

It doesn't support multichannel audio or any codec or other resolution than main profile H.264.

Do we expect Llano to beat GTX 460 in GPGPU?

Nobody but pxc expects this. Yet he's truly disappointed, or maybe he's being a subtle Intel troll?

AMD Llano is a pure win on all fronts, especially for those who actually knew what was in it:

- Great power properties. In CPU tasks it's as efficient as Intel competition. In GPU intensive tasks it's better by factor of 2 or more and this gap widens when you compare it with desecrate GPU solutions.
- Even though it's technically an Athlon II core based CPU and really is a short lived sort of a "proof of concept" AMD has managed to improve the IPC on average by 6%.
- It is much cheaper than the competition.
- 400sp GPU with native support for DDR3-1866 RAM.

What's not there to like, seriously.
 
Last edited:
You must log in or register to reply here.
Back
Top