AMD Beema/Mullins Launch Thread

[Abwx]

That's odd, because Anandtech clearly says that they were produced on TSMC: http://www.anandtech.com/show/7974/...hitecture-a10-micro-6700t-performance-preview Not really sure which is correct.

They have updated their article with this :

AMD nous a confirmé que ces puces avaient vu leur production passer de TSMC pour la génération précédente à GlobalFoundries.

"AMD confirmed to us that thoses chips have seen their production going from TSMC for the previous generation to GF".

http://www.hardware.fr/news/13677/amd-annonce-beema-mullins-maj.html

I guess that Anand just assumed that it was strictly the same chip without bothering to ask AMD.

Edit : I read that he also mislabeled some slide and made wrong assumptions
based on his mistake...

 

[AtenRa]

All these reviews tell us absolutely nothing. Not one single useful piece of information about battery life. Absolutely nothing. For all we know these are going into 2 cell tablets that run for 2 hours before they die. (Or < one hour of gaming?) It may be able to play a game of DOTA, but we dont know if it will even last long enough to finish one.

The MSI Temash 11.6" IPS Tablet last year had almost 4 hours Battery life in Full load Full brightness, even with the same battery you will get higher run time with Mullins and 2x the performance.

 

[Enigmoid]

The MSI Temash 11.6" IPS Tablet last year had almost 4 hours Battery life in Full load Full brightness, even with the same battery you will get higher run time with Mullins and 2x the performance.

You mean this tablet?

http://www.notebookcheck.net/Review-MSI-W20-A421-Tablet.108285.0.html

With a 600 mhz unbearably slow CPU (downclocked from 1 ghz). And a subpar display (IPS and brightness are nice but 768p on a 11.6" tablet is pretty bad- 135 ppi is not going to cut it).

On a 34 watt-hour battery?

 

[seitur]

I am actually surprised. It seems like it might be first in a longer while CPU product from AMD that is actually good.

Need more tests though. Also for desktop Beema.

 

[lefty2]

They have updated their article with this :

"AMD confirmed to us that thoses chips have seen their production going from TSMC for the previous generation to GF".

http://www.hardware.fr/news/13677/amd-annonce-beema-mullins-maj.html

I guess that Anand just assumed that it was strictly the same chip without bothering to ask AMD.

Edit : I read that he also mislabeled some slide and made wrong assumptions
based on his mistake...

... and Beema/Mullins GPU has 38% less leakage current as a result. So, does that mean GlobalFoundaries 28nm is better than TSMC 28nm?

 

[AtenRa]

You mean this tablet?

http://www.notebookcheck.net/Review-MSI-W20-A421-Tablet.108285.0.html

With a 600 mhz unbearably slow CPU (downclocked from 1 ghz). And a subpar display (IPS and brightness are nice but 768p on a 11.6" tablet is pretty bad- 135 ppi is not going to cut it).

On a 34 watt-hour battery?

As i have said, Mullins could have the same or better run time and 2x the performance. So i believe that 4 hours of gameplay is fine, but we have to wait for actual products to reach the market in order to really have an idea and actual data.

 

[Enigmoid]

As i have said, Mullins could have the same or better run time and 2x the performance. So i believe that 4 hours of gameplay is fine, but we have to wait for actual products to reach the market in order to really have an idea and actual data.

The problem is that AMD needs more than twice the performance. Dual jaguar (yes I'm calling it jaguar since I see no reason to call it puma or even puma+ considering there were not core performance changes and IPC didn't change at all) at 1.2 ghz (twice the 600 mhz in that failed tablet) is still miserably weak compared to baytrail 3740.

The tablets look good and If the price is right I will most definitely buy one. Held out on baytrail because I couldn't seem to fine anything with a 10" 1080p screen for around $300.

 

[krumme]

Damn...That sure was the surprise of the month. - lol

If 3.95W tdp (2.8W sdp) get you there, it puts AMD entire big core future relevance in perspective. So little ipc for so much power, mm2 and development cost. 2.2GHz with 80% trinity ipc and 85% of Llano - thats dangerously close!

 

[cbn]

Makes me wonder, though, why AMD is bothering with ARM microserver chips when Puma+ seems so well suited to that market. Hmm...

I am assuming Puma+ SOC has two SATA ports just like Kabini (BGA and AM1) and just like Kyoto.

Two sata ports?

Two.

Why only two?

(Yes, I understand Laptops at most only require two SATA. One for 2.5" SSD or 2.5" HDD and one for optical drive.......but isn't there so much more potential beyond this considering the Kabini AM1 cat cores have AES-NI and are ECC capable)

Is adding two more SATA (to bring to total to four) something that drives up silicon die size too much?

How many mm2 for each SATA on 28nm?

I would just love to have a cat core MINI-ITX with ECC and four native SATA ports. Compared to what something from Intel costs with those features it would probably be a real bargain.

It's made for laptops and tablets, not microservers. If you really want more SATA ports then attach a SATA controller over PCIe, though then you still have a wasted GPU.

If using Linux, I would rather have native SATA for all four ports.

Regarding it being made for laptops and tablets first. Yes, I do understand this, but Brazos (another mobile targeted cat core platform) had support for six SATA ports via some hudson chipsets:

http://en.wikipedia.org/wiki/Comparison_of_AMD_chipsets#Fusion_controller_hubs_.28FCH.29

.....and this on a PCH (built on 65nm) that was only 28mm2 according to the Anandtech review article--> http://www.anandtech.com/show/4134/the-brazos-review-amds-e350-supplants-ion-for-miniitx

The Hudson FCH is very tiny measuring approximately 4mm x 7mm for a total die size of around 28mm2.

I'd imagine on 28nm. having four SATA would have resulted in very little die size increase.

That's something the PCIe x4 slot could be used for. Though it'll add to cost.

I do think AMD made the right call here, these chips are targeted at mobile. The desktop versions are just a happy side-effect. Most laptops, especially in this price range, won't need more then two. Added SATA ports would just waste power.

For the laptops and tablets, AMD could just fuse off the unwanted SATA. Based on what I have read from past reviews reducing I/O in this fashion would reduce power as well.

 

[krumme]

... and Beema/Mullins GPU has 38% less leakage current as a result. So, does that mean GlobalFoundaries 28nm is better than TSMC 28nm?

Well i think its fair to give gf the credit of doubt here despite their history (man that was hard to write). But sure only time will tell. (Now i feel a little better).

But for amd moving this to gf is a huge gain for the wsa agreement and their economy. Add this product give them a stronghold for the low end laptop and desktop and it looks like a solid win. No way are they going into tablets with this because of Intel paying their way no matter the product and we still dont know battery life for this. My guess is Intel still holds a very comfortable lead here but amd gained the performance crown at the laptop and desktop level.

 

[AtenRa]

The problem is that AMD needs more than twice the performance. Dual jaguar (yes I'm calling it jaguar since I see no reason to call it puma or even puma+ considering there were not core performance changes and IPC didn't change at all) at 1.2 ghz (twice the 600 mhz in that failed tablet) is still miserably weak compared to baytrail 3740.

The tablets look good and If the price is right I will most definitely buy one. Held out on baytrail because I couldn't seem to fine anything with a 10" 1080p screen for around $300.

I was talking about the Quad Core A10 Micro-6700T. I dont have Data for the Dual Core.

 

[bullzz]

@krumme - i think the 3.9W TDP is for the dual core chip. not the ones benchmarked here

 

[Abwx]

at 1.2 ghz (twice the 600 mhz in that failed tablet) is still miserably weak compared to baytrail 3740.

.

BT TDP is guaranted for base frequencies, that is about 1.4GHz, anything above will increase TDP well above the official one, it s not like thoses chips are really economical, it s just that Intel consider that 0W for 1 minute + 10W for 2 minutes = 3.3W on average.

Such a definition can help on benchmarks but battery life wise it would litteraly unspec the CPU.

 

[TimCh]

@krumme - i think the 3.9W TDP is for the dual core chip. not the ones benchmarked here

Yes, the one reviewed has a tdp of 4.5W.

 

[Fjodor2001]

Damn...That sure was the surprise of the month. - lol

If 3.95W tdp (2.8W sdp) get you there, it puts AMD entire big core future relevance in perspective. So little ipc for so much power, mm2 and development cost. 2.2GHz with 80% trinity ipc and 85% of Llano - thats dangerously close!

So can't they make a new big core uArch generation based on the small cat cores?

 

[Arkadrel]

The problem is that AMD needs more than twice the performance. Dual jaguar (yes I'm calling it jaguar since I see no reason to call it puma or even puma+ considering there were not core performance changes and IPC didn't change at all) at 1.2 ghz (twice the 600 mhz in that failed tablet) is still miserably weak compared to baytrail 3740.

Except the 2.8 SDP Mullins is FASTER than the Baytrail 3740.

Yep thats just "miserably weak".

Lets see:

web browseing tests:
Sunspider 1.0 = Mullins faster than 3740
Mozilla Kracken benchmarkk = mullins faster than 3740
WebXPRT = mullins faster tahn 3740

Cpu tests:
Cinebench 11,5 single thread = mullins faster than 3740.
Cinebench 11,5 multi thread = mullins faster than 3740.



You make it sound like this new chip from AMD is weak compaired to the Baytrail,
when infact its faster both in CPU tasks and in GPU ones.

Check out the review Anandtech did of it, or any of the other review sites that benchmarked it.

 

[Khato]

Except the 2.8 SDP Mullins is FASTER than the Baytrail 3740.

Indeed, though it's also not like they're in the same power category when producing those results. Just wait and see how much AMD actually improved their power consumption versus what they're claiming via the TDP/SDP ratings. Because all evidence available thus far indicates that they cut TDP on Beema/Mullins versus the previous generation by cutting the guaranteed base frequency in comparison while adding a turbo-mode to improve short duration benchmark scores. Which isn't a bad thing by any means - it's awesome that they finally caught up in that regard as it makes for a huge improvement in user experience.

 

[Abwx]

Indeed, though it's also not like they're in the same power category when producing those results. Just wait and see how much AMD actually improved their power consumption versus what they're claiming via the TDP/SDP ratings. Because all evidence available thus far indicates that they cut TDP on Beema/Mullins versus the previous generation by cutting the guaranteed base frequency in comparison while adding a turbo-mode to improve short duration benchmark scores. Which isn't a bad thing by any means - it's awesome that they finally caught up in that regard as it makes for a huge improvement in user experience.

This is completely contradicted by this :

AMD claims to have achieved a 19% leakage reduction across the CPU cores and a 38% leakage reduction in the integrated GPU.

http://techreport.com/review/26377/a-first-look-at-amd-mullins-mobile-apu

One must take account that it s no more the same process and that either 30% higher absolute frequency or 40% lower power drain at the same frequencies is what apparently differ from the two iterations, not negating that there was some rework that helped extract more perfs from the new silicon.

 

[Fjodor2001]

How does the new 15 W AMD A6-6310 stack up against the 15 W mainstream Haswell CPUs such as the i5-4200U? Have any comparisons been published?

I'm thinking both with regards to performance and battery life.

 

[inf64]

This gotta be the best AMD product since K8 launched . It's fabbed @ GF (confirmed by Anand now) so it shows how good the design fits their process node tech (vs TSMC's).

If only Excavator core would bring some serious x86 perf. boost for mainstream desktop performance segment (up to 65W) and AMD would be positioned really well for 2015-2016.

 

[Abwx]

How does the new 15 W AMD A6-6310 stack up against the 15 W mainstream Haswell CPUs such as the i5-4200U? Have any comparisons been published?

I'm thinking both with regards to performance and battery life.

AMD put the 6310 up to the Pentium 3556U, looking at their numbers it could eventualy go against a product that would be a step above.

http://forums.anandtech.com/showpost.php?p=36309598&postcount=55

 

[Arkadrel]

]How does the new 15 W AMD A6-6310 stack up against the 15 W mainstream Haswell CPUs[/B] such as the i5-4200U? Have any comparisons been published?

I'm thinking both with regards to performance and battery life.

Good question.
Closest thing I could find was this:

Haswell Pentium 3556U vs a A6-6310.

Seems to do okay against it in Pcmark, 3Dmark 11, and Basemark CL.


Im actually surprised how energy effecient these chips are considering its AMD.
AMD usually cannot go toe two toe with Intel when it comes to performance/watt, these look like they could be closer than anything else AMD has.

 

[cbn]

So can't they make a new big core uArch generation based on the small cat cores?

That would be great especially if it kept compatibility with socket AM1:

1. Quad "smaller cat" cores like Jaguar and Puma+ (quad core, two wide OoO design)

2. Dual "larger cat" cores (call then Lion, Tiger, etc) (dual core, three wide OoO design)

 

[Abwx]

2. Dual "larger cat" cores (call then Lion, Tiger, etc) (dual core, three wide OoO design)

Steamrollers cores decoders are 4 instructions/cycle capable, that s double of Jaguar, widening the integer execution pipeline by half, that is one ALU + AGU, with the relevant Store/load management as well as increasing by only half the FP exe unit would overall increase the IPC by a theorical 50% , undoubtly out of reach given the inherent bottlenecks but even half of this value would render any Jaguar eventual upsizing as too time consuming in comparison, not counting Steamrollers pipeline as being extremlely versatile and easily ISA updatable.

 

[Idontcare]

This gotta be the best AMD product since K8 launched . It's fabbed @ GF (confirmed by Anand now) so it shows how good the design fits their process node tech (vs TSMC's).

It is definitely improved, but we can't assume the improvements are due to GF over TSMC given that we have the classic project management triangle thing going on here in which allowing for more development time combined with more resources (money) results in a higher quality product.

For all we know these chips would have been the same, or better, had they been developed and equally optimized for TSMC's node versus GF's.

What we can say with certainty is that to whatever extent GF's node compares to TSMC's, it doesn't appear to be holding AMD's design back in this case.