Richland & Kabini rumours

[Nemesis 1]

Offtopic response: Haswell no longer being 10w means its a weaker Haswell. Core sizes means ALOT! Less room for batteries! omg...and other equipment! Sheesh. Atom actually looks better than Haswell in that market! This guy eats Intel marketing like its nothing.

On topic, why is the APU's so sensitive to RAM? More so than Intel's iGPU. And do you guys see this "trend" improving even more?

I just saw some G.skill RAM OC'ed to 3900MHz, and made me wonder.

Actually it is ontopic . You guys made it topic when you brought intel into this topic , It wasn't me . hjust read from first page easy to find the guility party .

So are you thinking that I am thinking that a 7 watt haswell will have some performance as a 55 watt haswell. OH i am sure they perform = LOL. So whats that got to do with anything . Is Arm going to surprize a low voltage intel haswell . Not likely arm amd or NV. Its the halo tablet part . the bigg dollar one . Don't worry about haswell silvermont is what you should worry about as we shall see both platforms around the same time.

 

[SiliconWars]

Actually it is ontopic . You guys made it topic when you brought intel into this topic , It wasn't me . hjust read from first page easy to find the guility party .

Yep it was mrmt in the 5th post http://forums.anandtech.com/showpost.php?p=34422004&postcount=5 , ending with "Can't this company do anything right? How they always are late in developing their own ideas and always end as the cheaper alternative to Intel products?".

I'm sure he's one of yours...who do you think mentioned intel first?

 

[Nemesis 1]

Name of the thread: Richland and Kabini rumors. Atom is not Richland and Atom is not Kabini. Thank you.

Than you should not have talked about Atom or haswell . The moment that happened it opened up this topic . or do you think AMD should have its own forum . The moment you start to compare Intel or Nv or qualcomm . Its an invite . If your dogging another product and expect people who like that product to stay out of that topic . your looking for excuse to create topics and butcher anothers products with talk and don't want any rebutle.

Have you ever been apart of a forum that changed its formate so as to keep opposing sides apart ? No need to answer I recall all of it . I also recall who the biggest offenders are .

NOw here at AT we had rollo and I like the guy but god he got me banned alot . I asked many times if he worked for Nv . The only reason rollo isn't here is because he lied to us.
ATI/NV had hellish battles here . ME being 100%ati that put mme up against the power brokers here . Not a good place to be . We battle here all the time hard . But what do we get from it . Guys research and we build a knowledge base. YOU can't just say a thing and get away with it when talking hardware . We like to see links . and we do the usual off the wall stuff . That keeps us going threw slow times . At is the best hardware forum around . Not so much software but hardware .

 

[MightyMalus]

Because the actual graphics execution units are more numerous and powerful than their Intel counterpart. If the graphics can do more work per second, then it needs faster access to memory in order to a) get more work to do and b) return its results. It's why you'll still see DDR3 memory on low-end discrete GPUs, but GDDR5 exclusively on anything faster.

Thank you for the explanation!

 

[SevenThirty]

On topic, why is the APU's so sensitive to RAM? More so than Intel's iGPU. And do you guys see this "trend" improving even more?

I just saw some G.skill RAM OC'ed to 3900MHz, and made me wonder.

Intel IGP have access to L3 cache. This is oen of the reasons that it runs better on i7 than on i3/i5. That, and HD 4000 is inferior generally in laptops, so removing a bottleneck is less noticable.

 

[Abwx]

An official number....

AMD_Unprocessed Lisa Su "Kabini delivers 50% more performance than Brazos 2.0" #AMDsurround #CES

 

[NTMBK]

AMD's CES talk just finished, Richland is officially 32nm.

 

[ShintaiDK]

AMD's CES talk just finished, Richland is officially 32nm.

So its basicly..nothing new? No CGN either. D:

 

[Hitman928]

Richland is trinity 2.0, no gcn cores either (unless I misunderstand the question during the q&a).

Presenter said 28nm is "coming up very nicely" for AMD. I suppose that means yields are looking good. I hope so as they have 3 new processor lines based on 28nm due by the end of the year.

Steamroller APU in 2H 2013.

 

[Rvenger]

AMD's CES talk just finished, Richland is officially 32nm.

Ouch. And I am not impressed by the roadmap at all. Intel will be on 14nm when AMD hits 28nm. Double the size of Intel's process tech.

 

[ShintaiDK]

Maybe its me. But I dont understand the reason for Richland at all, so short lived. Unless there is something else AMD wont tell us.

 

[Abwx]

She said 20-40% better perfs for Richland A8/A10 so there s something
that has been enhanced , either the GPU or the CPU or both but it s
unlikely that this is coming from a frequency bump.

 

[mrmt]

Yes and the same intel delivered 90nm and 45nm. The same GF delivered 65nm and 45nm. The same TSMC delivered 40nm and 28nm.

That would be my point.

Are you implying that Intel had issues with 90nm? Dothan had no issues with 90nm. Intel's problem at that time was a netburst, a flawed concept much like Bulldozer is today. Not even 65nm could fix Netburst, and it was the same 65nm node that gave us Conroe, the chip that wiped AMD line up overnight. Same with TSMC 40nm, a process that became a very interesting once the ramp up issues were solved.

What I said about GLF is that it does not a foundry on par with TSMC, much less with Intel. I never said that their 32nm process is botched to a point that ditching SOI and doing a half node shrink would yield a 40% increase in clocks. It's you who is implying that, with what basis I don't know.

 

[Rvenger]

Perhaps they took the bottleneck out of the PD architecture. lol

 

[Hitman928]

Maybe its me. But I dont understand the reason for Richland at all, so short lived. Unless there is something else AMD wont tell us.

Most likely you won't see Kaveri really hit until sometime next year. 2013 Will most likely be more of a paper launch with very limited availability. Richland seems like more of a solution update than a chip update with bundled software and additional power saving features for more perf/w or more performance when you need it and better power management when you don't for a longer battery life.

AMD could also have been surprised by the quickness of the 28nm ramp and thereby decided to push Kaveri and keep Richland at limited supply. Less likely, but remotely possible.

 

[ShintaiDK]

She said 20-40% better perfs for Richland A8/A10 so there s something
that has been enhanced , either the GPU or the CPU or both but it s
unlikely that this is coming from a frequency bump.

Only thing that they can really do is expand the chip with more SPs besides adding more cores. I guess it also explains the shortlived Richland. Its simply a product to use up all the wafers they are forced to pay for at Glofo if I had to guess.

 

[Abwx]

Perhaps they took the bottleneck out of the PD architecture. lol

I think that it s mainly a GPU improvement , but then without a slightly
better CPU it would be diffucult to extract more perfs unless this was only
a matter of bandwith...

 

[mikk]

Maybe its me. But I dont understand the reason for Richland at all, so short lived. Unless there is something else AMD wont tell us.

I guess Kaveri won't come in 2013. Keep in mind AMDs notorious slow execution when it comes to a new architecture. If Richland comes middle of 2013 most likely we won't see Kaveri this year.

 

[NTMBK]

Ouch. And I am not impressed by the roadmap at all. Intel will be on 14nm when AMD hits 28nm. Double the size of Intel's process tech.

That presentation just said that Kaveri is due out 2H13, so... no.

Maybe its me. But I dont understand the reason for Richland at all, so short lived. Unless there is something else AMD wont tell us.

I think it's just better bins on Trinity, tbh, so they can clock the laptop parts better without blowing the budget. On the plus side, looks like I voted right in that poll.

 

[piesquared]

Ouch. And I am not impressed by the roadmap at all. Intel will be on 14nm when AMD hits 28nm. Double the size of Intel's process tech.

It's going to take some awfully funny math to make that work considering 14nm won't be hear until at least the middle of 2014, more likely into 2015. And Kabini/Temash will ship in the first half of 2013.

 

[piesquared]

That presentation just said that Kaveri is due out 2H13, so... no.



I think it's just better bins on Trinity, tbh, so they can clock the laptop parts better without blowing the budget. On the plus side, looks like I voted right in that poll.

Or Resonant Clock Mesh...

 

[Abwx]

I think it's just better bins on Trinity

Or Resonant Clock Mesh...

The claimed performances improvements numbers are too high
to be just provided by such marginal tricks...

 

[Evleos]

Regarding Richland:

(4) Testing and projections develop by AMD Performance Labs. The score for the 2012 AMD A10-4600M on the PCMark 7 Overall benchmark was 1965 and the 2012 AMD A8-4555M was 1650, while the "Richland" 2013 AMD A10-5750M was 2175 and the 2013 AMD A8-5545M was 1850. PC configs based on the "Pumori" reference design with the 2012 AMD A10-4600M with Radeon(TM) HD 7660G graphics, the 2012 AMD A8-4555M with AMD Radeon(TM) HD 7600G graphics, the 2013 AMD A10-5750M with AMD Radeon(TM) HD 8650G graphics and the 2013 AMD A8-5545M with AMD Radeon(TM) 8510G Graphics. All configs used 4G DDR3-1600 (Dual Channel) Memory and Windows 7 Home Premium 64-bit. RIN-4

(1) Testing and projections develop by AMD Performance Labs. The score for the 2012 AMD A10-4600M on 3DMark 11 was 1150 and the 2012 AMD A8-4555M was 780 while the "Richland" 2013 AMD A10-5750M was 1400 and the AMD A8-5545M was 1100. PC configuration based on the "Pumori" reference design with the 2012 AMD A10-4600M with Radeon(TM) HD 7660G graphics, the 2012 AMD A8-4555M with AMD Radeon(TM) HD 7600G graphics, the 2013 AMD A10-5750M with AMD Radeon(TM) HD 8650G graphics and the 2013 AMD A8-5545M with AMD Radeon(TM) 8510G Graphics. All configurations use 4G DDR3-1600 (Dual Channel) Memory and Windows 7 Home Premium 64-bit. RIN-1


3DMark11 2012/1H2013 platforms respectively, scores 1150/1400, using the same memory bandwidth.
PCMark 2012/1H2013 platforms respectively, scores 1965/2175, using the same memory bandwidth.

 

[Evleos]

(2) Test conducted in AMD Performance Labs using FutureMark 3DMark Vantage P as a metric for GPU performance and PCMark Vantage as a metric for CPU performance. The AMD Z-60 APU-based system scored 455 and 1914 in 3DMark and PCMark respectively. The AMD A6-1450 APU-based system delivers scores of 981 and 3123 in 3DMark and PCMark respectively. The platforms tested were the AMD Z-60 APU with AMD Radeon(TM) HD 6250 graphics, 2x 2GB DDR3-1066, Microsoft Windows 7 and the "Larne" reference platform with an AMD A6-1450 Quad Core 1.0GHz APU, AMD Radeon(TM) HD 8280 Series graphics, 2GB DDR3-1066 system memory and Microsoft Windows 7. TEM-2

Tablet platforms:

2012: 455 in 3DMark and 1914 in PCMark
2013: 981 in 3DMark and 3123 in PCMark

(3) Test conducted in AMD Performance Labs measuring productivity performance with PCMark Vantage. The "Kabini" A6 APU-based system scored 5271 while the "Brazos" APU-based system scored 2807. "Kabini" PC config is based off the "Larne" reference design with 2013 AMD A6-5200 APU with AMD Radeon HD 8400 graphics, 4G DDR3 1600, and Windows 8 64bit. "Brazos" PC config is based off the "Renmore" reference resign with 2012 AMD E2-1800 APU with AMD Radeon HD 7340 graphics, 4G DDR3 1333 and Windows 7 Ultimate. KBN-3

Netbook/AIO/cheap ultrabook platforms:

2012: 2807 in PCMark Vantage
2013: 5272 in PCMark Vantage

So I guess they're not cherry picking their benchmarks after all, considering how general 3DMark and PCMark are.

 

[MightyMalus]

Perhaps they took the bottleneck out of the PD architecture. lol

I laughed so hard when I read that!

But anyways. I am not actually surprised. I was already expecting this. Its the same info we been talking about for awhile. What's the surprise?

Richland has always been a "minor" update. Now its a 20% to %40 update.
And the SoC will be a big update against Hondo. And it is.
Kaveri has been said its a 2014 chip.
What's new?
Actually, I wasn't expecting Richland to improve so much...sure it don't have a GCN Igpu?