Tuesday, October 25th 2011
AMD Trinity Detailed Further, Compatible with A75 Chipset
AMD detailed its upcoming "Virgo" PC platform that consists of next-generation "Trinity" APU (accelerated processing unit), and current-generation AMD A75 "Hudson-D" chipset. A notable revelation here is that the next-gen APUs will be compatible with AMD A75, although it will be designed for a new socket called FM2. It remains to be seen if FM1 and FM2 are pin-compatible.
"Trinity" packs four x86-64 cores based on the next-generation "Piledriver" architecture, arranged in two Piledriver modules. A module is a closely-knit group of two cores, with certain shared and dedicated resources. Each Piledriver module has 2 MB of L2 cache shared between the two cores. In all, Trinity, with its two modules, has 4 MB of L2 cache without any L3 cache.
AMD is talking about a 20% performance improvement over current-generation "Llano" APUs, which use K10 "Stars" architecture cores. Trinity will feature 3rd-generation TurboCore technology that adds a few new power-management and selective overclocking features.
The integrated memory controller will get an overhaul, too. Unlike with K10-based processors that have two independent 64-bit wide memory interfaces that can be configured to work ganged or unganged, Trinity will have a single 128-bit memory interface, the controller will support dual-channel DDR3-2133 MHz memory standard, with DRAM voltages of under 1.5V. Trinity will include a 24-lane PCI-Express root complex, it supports 2-way multi-GPU configurations.
Moving on to the integrated GPU component, AMD promises a 30% performance improvement over Llano's iGPU. The GPU component is DirectX 11 compliant, and features UVD 3 hardware HD video acceleration, with SAMU and native VCE. Featuring AMD Eyefinity technology, this integrated GPU will support up to three displays without needing a discrete graphics card. Eyefinity can be used to step up productivity.
Source:
DonanimHaber
"Trinity" packs four x86-64 cores based on the next-generation "Piledriver" architecture, arranged in two Piledriver modules. A module is a closely-knit group of two cores, with certain shared and dedicated resources. Each Piledriver module has 2 MB of L2 cache shared between the two cores. In all, Trinity, with its two modules, has 4 MB of L2 cache without any L3 cache.
The integrated memory controller will get an overhaul, too. Unlike with K10-based processors that have two independent 64-bit wide memory interfaces that can be configured to work ganged or unganged, Trinity will have a single 128-bit memory interface, the controller will support dual-channel DDR3-2133 MHz memory standard, with DRAM voltages of under 1.5V. Trinity will include a 24-lane PCI-Express root complex, it supports 2-way multi-GPU configurations.
Moving on to the integrated GPU component, AMD promises a 30% performance improvement over Llano's iGPU. The GPU component is DirectX 11 compliant, and features UVD 3 hardware HD video acceleration, with SAMU and native VCE. Featuring AMD Eyefinity technology, this integrated GPU will support up to three displays without needing a discrete graphics card. Eyefinity can be used to step up productivity.
41 Comments on AMD Trinity Detailed Further, Compatible with A75 Chipset
I think the future of AMD is staked at their capacity to succesfully mass produce this APU with proper yields.
While bulldozer isnt kicking I7 butt, piledriver with a 7xxx GPU will blow away anything intel has to offer as system on a chip. Thus any laptop with an intel GPU completely out of the water.
This could (and should) be the future of any low/medium end laptop at a very affordable price.
But again everything hangs on if they can succesfully mass produce it with nice profit margines. (and a decent TDP)
How can they bring Trinity to 20% improvement over Llano ???
EG:Llano mobile quad core was 1.4GHZ,AMD dual core was 1.9GHZ, seeing as these don't use a full 2 cores, but 2 half cores in a module, i think these bulldozer cores will be closer to 2GHZ
hmm may be AMD will improve with this like Phenom I to PII now?
AMD is just riding on the 'may be' factor!
Dunno how long it will last. But at least FM1 will have some upgrade options. MXM couldn't do it...couldn't make the upgradable GPU in a laptop popular, but AMD might be able to do it with APUs.
2) They say that Piledriver will be a +10%/core boost over Bulldozer.
3) They say that the power efficiency problem will be fixed, so that the high clocks of the deep pipeline architecture can be taken advantage of more effectively. So, 4ghz+ stock clocks on Piledriver -- which also means higher per core performance.
4) Llano uses Athlon II cores.
5) It will have an updated, higher performance GPU
6) Llano uses an updated 5xxx GPU branded as an 65xx.
It's quite easy to see how they'll get a performance improvement. Since they don't give details on the comparisons, they can cherry pick whichever comparison they're making. Something vs some previous Llano configuration is a 20% improvement. OR In some particular benchmark, there is a 20% improvement.
If you don't get excruciating detail it's just an advertising statement with no references and no commitments.
[Assuming, of course, that the information from Donanimhaber is correct, and actually comes from AMD]
(gpus are great at floating point calculations, which bulldozer is week at)
I'm really happy I bought my A8-3850 CPU. My only want is for an unlocked multi.
No thx to BD.
www.guru3d.com/article/amd-fx-8150--8120-6100-and-4100-performance-review/4
FX-4100 is on par with the PII X4 810 or A6-3650 at best, both 2.6 Ghz chips. Being a 3.6 Ghz chip it never gets close to the X4 975 BE of same clocks. Neither is close to the 3.4 Ghz X4 965 and even the FX-6100 often struggles to catch 3.6 Ghz ++ Phenom II's. Of course there's some niche tests where it gets close or even scores a pirric win, but those niche benchmarks are not the kind of tasks that a Fusion platform will ever be doing. BD is also a server chip, so server-like tasks it's where they don't completely suck, but there's no salvation for Trinity in that regards, so they really have to improve it over BD. Unless it's fast in home tasks, it will be fail from our perspective. They will still sell to the masses tho, probably, but that's because they are not concerned about efficiency, ROI, such things.
EDIT: I see that Piledriver is still going to use 2 MB of L2 per module. Am I the only one who thinks they would do much much better with a 1 MB, maybe even 512 KB L2 that is "a lot" (comnparatively) faster? Granted, we don't really know if BD cache is slow because of the massive amount of it or the cache size is what it is in order to try to compensate for the slow access. I really think it's the former, based on other AMD chips (and Intel's), but I'm not an enginner. Either way being things as they are for current BD, devoting almost 40% of die sze to cache seems like a total waste.
Put me down for three. The kids are getting older and tire of sharing a computer.
Second of all, it gets VERY VERY VERY VERY VERY DEPRESSING to only see performance increase numbers of 10-20% right now.
Come on AMD, release something worth releasing, they could have had opportunity to get some serious competition with Intel after x6thuban ect ect....
Bullshit 10-20% increase in performance numbers, yeah the probably will meet those paper launch numbers, but they have also set them selves up for failure if they possible don't even meet those paper launch numbers(which could be possible.)
I just want AMD to release something worth while, and stop pushing necessarily deformed baby's right out of there colon, and trying to make market on them.
And yes that Graphics performance it waht improves most when running faster Mem.
Phenom II X8 about 2.7GHz
www.phoronix.com/scan.php?page=article&item=amd_fx8150_bulldozer&num=1
2 x Opteron 2384 should be your Phenom II example
Bulldozer Family architecture was a do or die scenario K8 derived architectures are over bros