In terms of manufacturing and performance, since Samsung recently inked to fab some Qualcomm systems, and GloFo/Samsung have a partnership, is there not potential for a somewhat rapid improvement in process to use an advanced, up and running and potentially available capacity 16nm process from Samsung within a short period of time? (<18 mo) to bring them closer to the new Atoms that are due.
Why did you ever both to write this article? I hope 2 pages of bashing AMD got whatever grievance you had out of your system. I will look forward to more professional articles in the future.
Begs the question why cover it at all if this is a totally uncompetitive product that is effectively dead on arrival. Why spend resources covering a product nobody will ever buy when there are numerous products lots of people buy they haven't covered.
Cover it because it is finally done. Plus 14 SATA 3 connectors and dual 10 GigE is nothing to scoff at. Add to that a good size of DDR4 ECC RAM and we have something to think about. I suggest you price shop dual 10 GigE for less than $400 and see what you get. I think those items make it worth considering despite the proc being uncompetitive on several fronts. Don't be confused; this is a niche system. This is not a jack of all trades setup.
Still even for real use (business) or just for myself for fiddling around I would go with the Xeon-D. It's just so much more powerful. Yes it costs more but not much. Even if its $500 vs.$ 1000. Software you run on that will cost 10x times of that amount.
it all depends on what you what you want, this is aimed at a use case with lots of hdd's & ram (thinking zfs here) plus failover 10GbE. so for a high-end high-density soho nas it looks perfect... that's a niche but it's one the a1100 is a few hundred dollars cheaper... just wish the huskyboard would expose all 14-sata ports, as without them it's pointless...
Hmmm, an ngnix/centos based caching server or redis memcache is, in terms of software, free. Maintenance costs money but that's it. Storage-wise, there is GlusterFS and openstack doesn't cost money either. Plenty you can run on these machines at extremely low costs...
I'd hardly call that bashing. The question he should have asked is how the heck is this company still in business when it executes this poorly? Now that is bashing.
Did you read the same article I did? I found it to be mostly-balanced, and if there was any bias it was in *favor* of AMD. It seems quite likely this will find a niche in NAS systems, and seems like a good groundwork for future iterations.
I will look forward to more intelligent comments in the future.
This feels more of an empty press release for investors. Your perf guess seems low,maybe it does quite a bit better with the final silicon and software. Power is crazy but A57 on 28nm , kinda to be expected and they can't really clock it at 1.2Ghz for 1.5W per core since perf would be too low. Don;t really think AMD expects more than pilot programs and further work on the software from this part. Still if you look at how A57 vs A72 do in phones, makes you regret they don't , at least, have A72 here on 28nm.
I don't know that it's fair to compare the power numbers when AMD is including 2 10gb network controllers in their SoC and Intel doesn't. The better comparison would be Xeon-D, and it'd TDP is close to 50% higher.
Consider that the actual 10 Gbe ethernet interface is of course not on the SOC die (just the MAC), the integrated 10 GBe components are probably only a few Watt.
This thing probably sucks but that isn't surprising. I never ever got the micro server hype. It does also not make sense when you can run stuff on VMs on beefier CPUs and get better performance/watt and $.
no it does not suck. it is based on ARM while it is not fully comparable to a x86 socket.
AMD had to release this socket on the default A57 to pave the road for there next gen k12 arm. With this platform they are able to give initial go for drivers/support etc.
I would disagree with that comment about performance per watt. In certain fairly common use cases, such as a storage server compared to the C2000 series you would not only have to add a raid card to the Intel setup, but also provide for dual 10gb nic ports. Both of those cards will add both cost and wattage to the total overall system...
Dear Anandtech, I'm very disappointed about this review. I have from long time a Atom server and the basic concept of power consuming used in the standard review is not right when you are evaluating a system below 50W. A single hard disk can consume from 5W (if you use low performance , high density disk ex: western digital red 2.5”) to 10W (if you use high performance hard disk). We need also the RAM that for each bank have a power consumed more or less of 3W. In my system ( ATOM with 2GB DDR2 RAM, LSI controller and 3 Hard disk 3.5” Raid 5) the power consumed of the only motherboard is around 35W and the hard disks have the same consumption. The power consume of the CPU is reported as 10W. I don’t want speak about this processor but 40W or 20W is the same thing for me because if we add 4 hard disks to our hypothetical system, the power consumed from the storage is greater than the CPU.
20W more or less per server node is a lot in a system like the HP Moonshot where have 40+ nodes in a high density system. It means that your 1 KW cluster now needs 1.8 KW.
And I do make the point that it less of an issue in a storage rich system and that AMD might have a chance there.
So an upgrade to A72, 14nm and USB 3.0 or 3.1 with actual SATA and PCI will make a great chip someday, now that the road was opened.Maybe a 16 core A72 will do even better...
Is an A1120 going to use 25W on average in all likelihood? Considering it's clocked the same as the A1150 which has a 32W TDP but double the cores, I doubt it.
With AMD's limited resources and Intel's aggressive, and at times, illegal anti-competitive practices, AMD cannot compete with Intel on Intel's terms, so they need investigate how to success where Intel (and Nvidia) isn't dominate (i.e. where Intel / Nvidia isn't focused).
AMD identified "micro servers" or "low power" servers as a potential market segment where it isn't obvious that Intel or other processor manufacturers will dominate this market.
History has repeatedly shown that new technology is almost always released in a fashion where it isn't polished, and in many cases actually new (different) technologies as opposed to a new generation of an existing technology, don't actually dominate over older technology within a single product development cycle, but it is still informative to follow progress to know what works or is promising, and what doesn't work.
"Thanks to the Linaro "96boards enterprise edition", a 300-400$ SoC + board should be available soon and make it much cheaper to build software for the 64 bit ARM ecosystem."
"the board will feature a 4-core AMD Opteron A1100 Series processor with two SO-DIMM memory slots, PCIe®, USB, SATA, and Gigabit Ethernet capabilities."
theres the AMD problem right there, the core SOC has 2x 10GbE and yet they will not release a cheapish 300-400$ board with the only real 10GbE option you are likely to buy gimped, sad as they could actually make a profit on these 10GbE SOC in the 2016 world consumer markets
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
37 Comments
Back to Article
Lonyo - Thursday, January 14, 2016 - link
In terms of manufacturing and performance, since Samsung recently inked to fab some Qualcomm systems, and GloFo/Samsung have a partnership, is there not potential for a somewhat rapid improvement in process to use an advanced, up and running and potentially available capacity 16nm process from Samsung within a short period of time? (<18 mo) to bring them closer to the new Atoms that are due.Kevin G - Thursday, January 14, 2016 - link
In 18 months we should be looking at the arrival of AMD's K12 chip if all goes according to plan.nandnandnand - Thursday, January 14, 2016 - link
K12 is expected to be on a 14/16nm process, right?medi03 - Friday, January 15, 2016 - link
Most likely 14nm (Samsung's process, with GloFo on board)Jtaylor1986 - Thursday, January 14, 2016 - link
Why did you ever both to write this article? I hope 2 pages of bashing AMD got whatever grievance you had out of your system. I will look forward to more professional articles in the future.Lonyo - Thursday, January 14, 2016 - link
I could 1 page of bashing AMD for being late and non-competitive in raw number terms. Then 2 pages of explaining potential use cases.Lonyo - Thursday, January 14, 2016 - link
*Count.Jtaylor1986 - Thursday, January 14, 2016 - link
Begs the question why cover it at all if this is a totally uncompetitive product that is effectively dead on arrival. Why spend resources covering a product nobody will ever buy when there are numerous products lots of people buy they haven't covered.eanazag - Thursday, January 14, 2016 - link
Cover it because it is finally done. Plus 14 SATA 3 connectors and dual 10 GigE is nothing to scoff at. Add to that a good size of DDR4 ECC RAM and we have something to think about. I suggest you price shop dual 10 GigE for less than $400 and see what you get. I think those items make it worth considering despite the proc being uncompetitive on several fronts. Don't be confused; this is a niche system. This is not a jack of all trades setup.beginner99 - Friday, January 15, 2016 - link
Still even for real use (business) or just for myself for fiddling around I would go with the Xeon-D. It's just so much more powerful. Yes it costs more but not much. Even if its $500 vs.$ 1000. Software you run on that will cost 10x times of that amount.bernstein - Friday, January 15, 2016 - link
it all depends on what you what you want, this is aimed at a use case with lots of hdd's & ram (thinking zfs here) plus failover 10GbE. so for a high-end high-density soho nas it looks perfect... that's a niche but it's one the a1100 is a few hundred dollars cheaper...just wish the huskyboard would expose all 14-sata ports, as without them it's pointless...
jospoortvliet - Thursday, January 21, 2016 - link
Hmmm, an ngnix/centos based caching server or redis memcache is, in terms of software, free. Maintenance costs money but that's it. Storage-wise, there is GlusterFS and openstack doesn't cost money either. Plenty you can run on these machines at extremely low costs...tijag - Friday, January 15, 2016 - link
http://begthequestion.info/Shadowmaster625 - Thursday, January 14, 2016 - link
I'd hardly call that bashing. The question he should have asked is how the heck is this company still in business when it executes this poorly? Now that is bashing.nandnandnand - Thursday, January 14, 2016 - link
It's not bashing to tell the truth. AMD is striking out once again.eanazag - Thursday, January 14, 2016 - link
I wouldn't consider this a strike out. This is a walk or a bunt for first. This is not as bad as their x86 desktop line.makerofthegames - Thursday, January 14, 2016 - link
Did you read the same article I did? I found it to be mostly-balanced, and if there was any bias it was in *favor* of AMD. It seems quite likely this will find a niche in NAS systems, and seems like a good groundwork for future iterations.I will look forward to more intelligent comments in the future.
jjj - Thursday, January 14, 2016 - link
This feels more of an empty press release for investors.Your perf guess seems low,maybe it does quite a bit better with the final silicon and software.
Power is crazy but A57 on 28nm , kinda to be expected and they can't really clock it at 1.2Ghz for 1.5W per core since perf would be too low.
Don;t really think AMD expects more than pilot programs and further work on the software from this part.
Still if you look at how A57 vs A72 do in phones, makes you regret they don't , at least, have A72 here on 28nm.
cygnus1 - Thursday, January 14, 2016 - link
I don't know that it's fair to compare the power numbers when AMD is including 2 10gb network controllers in their SoC and Intel doesn't. The better comparison would be Xeon-D, and it'd TDP is close to 50% higher.JohanAnandtech - Friday, January 15, 2016 - link
Consider that the actual 10 Gbe ethernet interface is of course not on the SOC die (just the MAC), the integrated 10 GBe components are probably only a few Watt.eldakka - Friday, January 15, 2016 - link
With 14 SATA ports, I wonder how this would perform as a ZFS-based NAS server?beginner99 - Friday, January 15, 2016 - link
True but you could just stick an expansion card into a Xeon-D server, if you need the CPU speed.beginner99 - Friday, January 15, 2016 - link
This thing probably sucks but that isn't surprising. I never ever got the micro server hype. It does also not make sense when you can run stuff on VMs on beefier CPUs and get better performance/watt and $.The_Assimilator - Friday, January 15, 2016 - link
"But there is more than meets the eye or we would not bother to write this article."And then at the end:
"So the new AMD SoC has no performance/watt advantage and no price/performance advantage over Intel's offerings."
AMD has failed, again.
Minion4Hire - Friday, January 15, 2016 - link
That's not the end. You have failed to read two entire pages if you think that is the end of the article.duploxxx - Friday, January 15, 2016 - link
no it does not suck. it is based on ARM while it is not fully comparable to a x86 socket.AMD had to release this socket on the default A57 to pave the road for there next gen k12 arm.
With this platform they are able to give initial go for drivers/support etc.
bill.rookard - Friday, January 15, 2016 - link
I would disagree with that comment about performance per watt. In certain fairly common use cases, such as a storage server compared to the C2000 series you would not only have to add a raid card to the Intel setup, but also provide for dual 10gb nic ports. Both of those cards will add both cost and wattage to the total overall system...Xeus32 - Friday, January 15, 2016 - link
Dear Anandtech,I'm very disappointed about this review.
I have from long time a Atom server and the basic concept of power consuming used in the standard review is not right when you are evaluating a system below 50W.
A single hard disk can consume from 5W (if you use low performance , high density disk ex: western digital red 2.5”) to 10W (if you use high performance hard disk).
We need also the RAM that for each bank have a power consumed more or less of 3W.
In my system ( ATOM with 2GB DDR2 RAM, LSI controller and 3 Hard disk 3.5” Raid 5) the power consumed of the only motherboard is around 35W and the hard disks have the same consumption. The power consume of the CPU is reported as 10W.
I don’t want speak about this processor but 40W or 20W is the same thing for me because if we add 4 hard disks to our hypothetical system, the power consumed from the storage is greater than the CPU.
JohanAnandtech - Saturday, January 16, 2016 - link
20W more or less per server node is a lot in a system like the HP Moonshot where have 40+ nodes in a high density system. It means that your 1 KW cluster now needs 1.8 KW.And I do make the point that it less of an issue in a storage rich system and that AMD might have a chance there.
mosu - Friday, January 15, 2016 - link
So an upgrade to A72, 14nm and USB 3.0 or 3.1 with actual SATA and PCI will make a great chip someday, now that the road was opened.Maybe a 16 core A72 will do even better...silverblue - Friday, January 15, 2016 - link
Is an A1120 going to use 25W on average in all likelihood? Considering it's clocked the same as the A1150 which has a 32W TDP but double the cores, I doubt it.Torashin - Friday, January 15, 2016 - link
Whatever happened to their Freedom Fabric?zodiacfml - Friday, January 15, 2016 - link
It is late but they don't have a choice. Sell it without profit or at a loss then do better in the next generations.blzd - Friday, January 15, 2016 - link
With AMD's limited resources, was this really worth investing in?mctylr - Friday, January 15, 2016 - link
With AMD's limited resources and Intel's aggressive, and at times, illegal anti-competitive practices, AMD cannot compete with Intel on Intel's terms, so they need investigate how to success where Intel (and Nvidia) isn't dominate (i.e. where Intel / Nvidia isn't focused).AMD identified "micro servers" or "low power" servers as a potential market segment where it isn't obvious that Intel or other processor manufacturers will dominate this market.
History has repeatedly shown that new technology is almost always released in a fashion where it isn't polished, and in many cases actually new (different) technologies as opposed to a new generation of an existing technology, don't actually dominate over older technology within a single product development cycle, but it is still informative to follow progress to know what works or is promising, and what doesn't work.
nils_ - Wednesday, January 20, 2016 - link
That article needs some serious proofreading...BMNify - Thursday, January 21, 2016 - link
"Thanks to the Linaro "96boards enterprise edition", a 300-400$ SoC + board should be available soon and make it much cheaper to build software for the 64 bit ARM ecosystem.""the board will feature a 4-core AMD Opteron A1100 Series processor with two SO-DIMM memory slots, PCIe®, USB, SATA, and Gigabit Ethernet capabilities."
theres the AMD problem right there, the core SOC has 2x 10GbE and yet they will not release a cheapish 300-400$ board with the only real 10GbE option you are likely to buy gimped, sad as they could actually make a profit on these 10GbE SOC in the 2016 world consumer markets