Information

Dear forum members,
We are delighted to inform you that our forums are back online. Every single topic and post are now on their places and everything works just as before, only better. Welcome back!



Discussion

Discussion on Article:
Alleged Specifications of AMD Fusion S-Series “Richland” Point to Clock-Speed Increase.

Started by: vanakkuty | Date 01/29/13 11:04:00 PM
Comments: 20 | Last Comment:  02/03/13 08:55:44 PM

Expand all threads | Collapse all threads

[1-8]

1. 
There is a trick to Richland that will easily differentiate between it and Trinity when it comes to performance. Wait for the launch to know more.
5 2 [Posted by: vanakkuty  | Date: 01/29/13 11:04:00 PM]
Reply
- collapse thread

 
Im bet it has to do with HSA and other plans AMD has to bringing OpenCL into everyday apps.
0 0 [Posted by: the_file  | Date: 02/01/13 03:19:28 PM]
Reply

2. 
show the post
1 4 [Posted by: AnonymousGuy  | Date: 01/29/13 11:04:44 PM]
Reply
- collapse thread

 
You are an Intel fanboy. Just jump up and down on a couch and shout it out. You have demonstrated form. You even use the old "I'm an AMD investor but..." line on 12 Oct 2012.
5 2 [Posted by: fanboyslayer  | Date: 01/30/13 03:15:26 AM]
Reply
 
How empty is your life that you bothered to look at my old posts?

It's also not fanboism when the two companies aren't even competing with each other anymore. AMD is trying just to stay out of bankruptcy at this point. Or did you miss their credit rating drop today? Or the annoucement months ago that they have no roadmap to being profitable in 2013?
1 3 [Posted by: AnonymousGuy  | Date: 01/30/13 07:33:55 PM]
Reply
 
Thanks to modern computers, it is possible to scan your posts in around the same time it takes to do a pee. My name is Fanboyslayer. I hunt fakes like you. You have been shown up time and time again by the mod as well. You ought to slither away from whence you came.
1 0 [Posted by: fanboyslayer  | Date: 01/31/13 06:37:00 AM]
Reply
 
Howcome the intel articles haven't even 20% of quantity of replies comparared to 98% of ALL AMD artiles on this site.

Some people are definatly trying to spread disinfo on this site.
0 1 [Posted by: the_file  | Date: 02/01/13 03:20:58 PM]
Reply

3. 
show the post
2 8 [Posted by: Tukee44  | Date: 01/29/13 11:31:48 PM]
Reply
- collapse thread

 
Hmmm
7 2 [Posted by: fanboyslayer  | Date: 01/30/13 03:06:49 AM]
Reply

4. 
Richland should be a nice performance bump and another great value for those looking for the best bank-for-the-buck.

It's reported that Richland uses the Piledriver instruction set for increased performance, so that may bode well for the desktop Piledriver CPUs due in early '14... or sooner.
3 3 [Posted by: beenthere  | Date: 01/30/13 08:24:59 AM]
Reply
- collapse thread

 
Richland is Trinity with a trick that gives upto 10% CPu gains and upto 20-40% gains for GPU as per AMD's test bed and test suites.

Its a Trinity die with no changes to design, so its Piledriver core and Northern Islands GPU same as Trinity.

You will know more about the trick when it launches.
7 2 [Posted by: vanakkuty  | Date: 01/30/13 11:02:43 AM]
Reply
 
Well if it gives up 10% of the CPU speed, my guess is the CPU underclocks and drops voltage when the GPU load goes > 90% in a game and the GPU's Turbo boost function kicks in to take advantage of the additional TDP headroom. This way in the most GPU-limited scenarios, more of the TDP's headroom is available for the GPU to boost its clocks and the reduction in CPU speed doesn't materially affect the overall system performance since the game is primarily GPU limited.
1 2 [Posted by: BestJinjo  | Date: 01/30/13 09:31:14 PM]
Reply

5. 
so this is just trinity but made whit there new high density library , better turbo and a new clock mesh. (ans a minor revision in the core design). correct me if i'm wrong.
3 1 [Posted by: massau  | Date: 01/30/13 01:51:09 PM]
Reply
- collapse thread

 
As mentioned above as reply to beenthere. There are no changes to design its the same as trinity. High density libraries are not used in Trinity/Richland, it will be used in a future family of chips.

3 2 [Posted by: vanakkuty  | Date: 01/30/13 01:57:39 PM]
Reply

6. 
Richland should yeild some decent gpu performance gains over trinity but it will be kaveri that's going to be AMD's golden goose. It will be the first AMD apu to use 28nm. The first AMD apu to be based on the steamroller arch and the first AMD apu to use GCN for its gpu.
3 0 [Posted by: SteelCity1981  | Date: 01/30/13 11:23:08 PM]
Reply
- collapse thread

 
AMD have always had 'amazing' stuff just over the horizon but never really delivered for me. Intel is no better and that looks to continue with Haswell if the numbers today are right.

I guess the days of significant performance increases are over. The only thing that would get my wallet out is a mid-range priced 8-core (real 8-core, 16-core in AMD world). How long have we had quad-core now for? 2006 I think and they have been mid-range for well over 5 years.
3 0 [Posted by: loadwick  | Date: 01/31/13 11:07:32 AM]
Reply

7. 
It would be nice if it featured ECC as well. I miss the time when AMD didn't arbitrarily cripple features and had ECC across the board. Now there aren't really any great options for a low-cost AMD server with integrated graphics.

The Intel i3/Xeons and boards are both way overpriced, and/or arbitrarily crippled in various ways, like missing AES extensions, VT-d, hyper-threading, and such. I hate to support the company on principle, and it is obnoxious having to scrutinize the Intel ARK for each and every CPU/feature.
3 1 [Posted by: x  | Date: 01/31/13 06:18:17 AM]
Reply

8. 
Generally speaking the mods on this site should add the HSA initiative to the list of major topics to this site. When HSA is done, meaning when AMD has fully implemented it, it will be a huge industry change.
2 0 [Posted by: the_file  | Date: 02/01/13 03:23:36 PM]
Reply
- collapse thread

 
Calculating on the GPU is 5 times more efficient than on the CPU. Whilst this ought to revolutionise computing, it depends on programmers hacking for this platform. This in turn depends on AMD's penetration of the platform. The embedded market is big and will help. Also Kaveri will offload some computation to the GPU itself without needing the special C coding hooks.
0 0 [Posted by: linuxlowdown  | Date: 02/01/13 06:55:23 PM]
Reply
 
"Also Kaveri will offload some computation to the GPU itself without needing the special C coding hooks."

I hope your right because that means --HUGE-- speeds. The implications behind HSA, and they stated in the mission statement and descriptions somewere also, is to give high level languages, INCLUDEING interpreted languages, (they used python as an example) access to low level controlls.
1 0 [Posted by: the_file  | Date: 02/03/13 08:55:44 PM]
Reply

[1-8]

Back to the Article

Add your Comment