What is Richland? AT users vote.

[Abwx]

This new APU uses the same basic silicon as Trinity, but time has been spent tuning the firmware and process technology to enable higher clock speeds and lower power consumption.

At the Consumer Electronics Show in Las Vegas, AMD discussed Richland with our Editor in Chief. He was told Trinity has lots of architectural headroom that wasn't exploited by the products launched thus far.

http://techreport.com/news/24190/process-tweaks-and-power-management-define-amd-richland-apu

 

[inf64]

http://techreport.com/news/24190/process-tweaks-and-power-management-define-amd-richland-apu

I wonder what exactly was "architectural headroom" that was not exploited before and to how much in terms of performance % it amounts(vs Trinity)? What they basically said was that Piledriver core was still not a real PIledriver but what they could debug/test in time for it to launch last year when it launched. So now they had a full year to test and debug and we will see a full potential? Sounds sketchy a bit but let's wait and see. All this applies to Bulldozer and how broken it really was. If Richland brings at least 3% of IPC with those BIOS microcode updates then PD will end up more than 10% faster than Bulldozer which is a lot for basically the same core. And since it still has many flaws that were detailed in many articles around the web, SR's performance jump that AMD stated at previous HC conference does not seem unrealistic at all. Also they will use more of the Richland's TDP budget for x86 workloads which in turn could bring some of the performance they reported(~10% jump in pcmark7 vs previous top mobile part).

When something was broken that much fixing it would yield big results.

 

[Haserath]

I wonder if this is how broken Bulldozer was, how broken the process was, or a combination of both.

40% GPU performance increase seems like a lot from what should be a solid GPU.

 

[inf64]

Well they explicitly mentioned process in that talk so I suppose if SP remains the same(although "re-arranged" as some websites claim) then performance jump comes from higher clocks for both CPU and GPU. AMD gave a ~20% higher number specifically for the new top vs old top mobile part(old being A10 4600M QC @ 2.3-3.2Ghz and 384SP). The 40% figure is for old A8 vs new A8 which I suspect is thanks to much higher clock of GPU that they can now run with more optimized process node on these new parts.

 

[Abwx]

IPC improvement of Piledriver over BD was too much considering
that it was only a slightly tweaked core.

Much came i think from reducing intenal latencies that were set
at high values to allow the early dies to run at high frequencies.

 

[NTMBK]

Sounds like they've got a much better turbo implementation, which helps explain how the BIOS and motherboard tweaks come into it so much. Also why it's really a mobile refresh, and the desktop is not so much of a big deal. I wonder if they're including improved/optimized drivers in that, as well? VLIW4 was fairly shortlived as a discrete GPU architecture (only in the 6900s) and only featured in Trinity on the APU side, so they probably still had some work to do in eking performance out of it from drivers.

EDIT: Maybe also faster memory support, too. The mobile Trinity parts only officially supported up to DDR3-1600, whereas the CPU went to DDR3-1866. If the mobile parts got faster RAM support that would be a good part of that boost.

 

[Gideon]

I actually asked AMD directly how they got the speed increase, considering pretty much nothing changed on the chip, and wondered whether it was due to PCMark being highly suspectible to storage improvements. This is part of the reply I got:

I understand the concern; we didn't move to 28nm and the architecture didn't change, so where is the uptick coming from? Performance gains have been achieved through a combination of increased frequency, improvements in the bi-directional power management between the CPU and GPU, as well as more precise thermal management that allows for running at higher boost speeds more often.

Obviousily i have nothing to back it up other than my e-mail history, so feel free to not believe it.

 

[Idontcare]

so now that we know whats your take now. I struggling abit today . So I having a hard time with what you wrote . So if you would give your take on what happened here . Thanks

Nemesis, I missed this post of yours but I can't make sense of it. What did you want me to do?

 

[Hitman928]

I actually asked AMD directly how they got the speed increase, considering pretty much nothing changed on the chip, and wondered whether it was due to PCMark being highly suspectible to storage improvements. This is part of the reply I got:



Obviousily i have nothing to back it up other than my e-mail history, so feel free to not believe it.

This is pretty much what they said in response to the same question (minus the PCMark specifics) at their press conference on Monday.