Slashdot Mirror

← Back to Stories (view on slashdot.org)

Banana Pi 24-Core ARM Server Running Ubuntu Breaks Cover (hothardware.com)

Posted by msmash on from the up-next dept.

MojoKid writes: ARM-based server processors have threatened to take on Intel in the data center for some time but not much has materialized thus far in terms of significant deployments. However, a new breed of low cost ARM server implementations may be in the works with a many-core platform called Banana Pi. The latest Banana Pi device being teased is something very different in the form of a 24-core ARM server that speculation suggests might be sold as a Banana Pi server board or as a finished server product.

A video has surfaced that reportedly shows a 24-core ARM Cortex-A53 processor with 32GB of RAM, though the OS only sees 29.4GB of that RAM. The OS is Ubuntu 18.04.1 LTS with MATE desktop. Unless the processor used in this device is something unannounced, and that seems unlikely, the chip itself would likely be a SocioNext SC2A11. The same processor is used in the Linaro Developer Box. The demo shows the server fully loaded at 100% CPU utilization building a Linux kernel and reportedly the system also supports NVMe storage as well as TensorFlow workloads for machine learning. Not much else is known about the system at this time but it's an interesting development in the Linux server space to be sure.

88 comments

  1. What's the advantage to using this? by Anonymous Coward · · Score: 0

    It's not clear what the advantage is over other competing systems like the Ampere eMAG. Is there a compelling reason to choose this server over the eMAG or existing Intel-based systems? I assume the benefits to ARM-based systems are the low power usage and that they provide an alternative to Intel's near-monopoly in this sector. Of course, AMD also has some new offerings, and have been some Opteron-based systems available previously. Also, is it really notable that this runs Linux? Isn't that pretty much assumed for most servers?

    One potential use for servers is high-performance computing, especially considering the mention of software like Tensorflow. Unfortunately, I wouldn't choose this for any attempt to build such a system. It's not clear how these servers could be interconnected, but Infiniband support, especially QDR Infiniband, is definitely preferable over ethernet, even as an optional add-on. Also, the amount of RAM per core is pretty low, which limits what could run on such a system. It would be more interesting to see a higher-end alternative offering more RAM per core and possible support for Infiniband.

    I'm just trying to understand the appeal of these systems. Is it a step along the way to ARM-based systems that are cheaper to install and operate? Or is there something else I'm missing here?

    Also, off-topic, but what's with the Nazi ASCII art? Slashdot admins are surely aware of this, and surely there's someone on the staff with enough Perl knowledge to tweak the lameness filter so it rejects those posts. At a minimum, shouldn't these get deleted like the APK posts do?

    1. Re:What's the advantage to using this? by ShanghaiBill · · Score: 2

      It's not clear what the advantage is over other competing systems

      1. Power consumption.
      2. Density.
      3. Cost

      For power consumption, you need to look at both what the server consumes, and the power used to cool the datacenter.

      Density means you can pack more into a rack, saving on expensive floor space.

      Cost savings aren't clear. TFA doesn't mention prices. But it will likely be less expensive than an Intel system of equal capability.

    2. Re: What's the advantage to using this? by Anonymous Coward · · Score: 0

      What I would like is a clear metric on how efficient these are per watt vs a 65watt Ryzen 2700. If i run sixty odd watts of these things, do I get more or less performance vs these army ants.

      I can infer that an Odroid ux4 is twice a pi3, and a good little atom is about twice that but half a peak mobile Celeron and that an old i3 is twice that and the ryzen is about triple that, but there's a lot of imprecision in taking five thumb rules in series.

      Also, what doesn't the "editor" ever stick in a price estimate on these things. If it is a hundred bucks i want a dozen for fun. If it's twelve hundred i don't care for one.

    3. Re: What's the advantage to using this? by Joce640k · · Score: 1

      An editor's job is to get you to read articles surrounded by adverts, not to provide information.

      --
      No sig today...
    4. Re: What's the advantage to using this? by Anonymous Coward · · Score: 0

      Density also applies to software. If these ARM based servers can run more 'code as a service' per server then cloud providers may be interested in scaling them alongside x86 workloads. How well is x86 virtualization on ARM these days?

    5. Re: What's the advantage to using this? by Anonymous Coward · · Score: 0

      At 32Gb RAM onboard it obviously won't cost just a hundred bucks. I would say US$ 500 minimum... and that would be the bare unit, without case, heatsink and power adapter.

    6. Re:What's the advantage to using this? by Guspaz · · Score: 1

      Can you really achieve good power consumption (in relation to performance) or density when you're using a six year old processor from 2012? This thing is positively ancient in CPU terms, something like three to generations behind. It's the contemporary of the iPhone 5 or the Galaxy Note 2.

    7. Re:What's the advantage to using this? by Tough+Love · · Score: 1

      1. Power consumption.

      Except for the weakest home and niche servers, nobody cares about that. It should be 1. Power performance

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    8. Re:What's the advantage to using this? by Anonymous Coward · · Score: 0

      HAHAHAHAHAHAHAHAHAHA What else is there to say about this?

    9. Re: What's the advantage to using this? by Anonymous Coward · · Score: 2, Informative

      It doesn't have any ram on board, the editor is retarded as usual

      It has 32GB of eMMC on board and 4 dimm slots for ram

    10. Re:What's the advantage to using this? by Anonymous Coward · · Score: 0

      If AWS is any clue to how ARM will be, it will be like 3x as expensive as other options. (Why AWS? Why do you promote it as being "cheap", when t2/t3 are even cheaper for just a micro instance?)

    11. Re:What's the advantage to using this? by Anonymous Coward · · Score: 0

      1. Power consumption.

      Except for the weakest home and niche servers, nobody cares about that. It should be 1. Power performance

      It's the other way around, nobody cares much about power savings for one or two computers in a household but when it's thousands of servers in a datacenter it is a big deal.

    12. Re:What's the advantage to using this? by willaien · · Score: 1

      Not really. Data centers spend a lot of money cooling their systems. If they can get a denser, more power-efficient setup that can do the same thing, they'll definitely be interested in that.

    13. Re:What's the advantage to using this? by Anonymous Coward · · Score: 0

      Actually businesses (mainly datacenters), would be greatly interested in this.

      Less power means they generate less heat.
      Less heat means less power being used by cooling (centralized A/C units), or less A/C units all together.
      Less A/C units mean less repair/maintenance costs, and less annual service contracts.
      Less power consumption also means less backup generators (smaller backup generators), and less annual service contracts for said generators, less repair.. yatta yatta yatta.
      Less heat (usually) means longer MTBF.

      AND/OR

      They can host significantly more servers/services, with the same annual cost or even less cost.

      Either way they are saving thousands to MILLIONS of dollars a year. That's not chump change.

    14. Re:What's the advantage to using this? by AHuxley · · Score: 1

      Its the fun of moving existing quality code from "Intel".
      Showboating on a new type of system rather than just been another person working on a large exiting projects.
      Who wants to be part of a very large project when new projects await on ARM?

      --
      Domestic spying is now "Benign Information Gathering"
    15. Re: What's the advantage to using this? by Anonymous Coward · · Score: 0

      An editor's job is to get you to read articles surrounded by adverts, not to provide information.

      Silly, an editor's job is to get you to read adverts, which happen to be surrounded by articles.

      Remember: Adverts pay the bills, not articles. You could care less if folks read your articles, as long as they read your adverts and buy from your advertisers.

    16. Re:What's the advantage to using this? by Tough+Love · · Score: 1

      You have a certain amount of work to do. You want to get it done using the least amount of power possible. That is power performance. That is also how you spend the least money cooling systems. See now?

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    17. Re: What's the advantage to using this? by Anonymous Coward · · Score: 0

      You've missed the point.

      Data centres don't care about absolute power consumption. They care about performance per watt, or power efficiency.

    18. Re: What's the advantage to using this? by Anonymous Coward · · Score: 0

      The metric may depend on workloads. E.g. one workload may work well on Intel and ARM, another not so. Memory bandwidth, numerical libraries, and speculative execution are some of the significant factors.

    19. Re:What's the advantage to using this? by willaien · · Score: 1

      Then I misunderstood what you meant. "power to performance ratio" or something like that would have been clearer. The phrase "power performance" just sounds like you're talking about souped up gaming machines or something where you don't care about power usage, just performance.

    20. Re:What's the advantage to using this? by Tough+Love · · Score: 1

      Let's call it energy efficiency, the standard terminology.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
  2. Already using ARM in the datacenter by cerberusss · · Score: 4, Informative

    have threatened to take on Intel in the data center for some time but not much has materialized

    Counterpoint: since May 2016, a French hosting company called OVH has been offering ARM-based servers under the brand Scaleway. I've been happy with it, mainly because of the price: EUR. 2,99 per month (US$ 3,41). I mainly use it to host my personal websites, a self-hosted TT-RSS instance, a private wiki, that sort of stuff. You get an ARM board with four cores and 2 gigs of memory. And no worrying about the latest Intel problems like Spectre, Meltdown and Management Engine holes.

    --
    8 of 13 people found this answer helpful. Did you?
    1. Re:Already using ARM in the datacenter by Anonymous Coward · · Score: 4, Informative

      You're correct that Meltdown and IME are confined to Intel processors. However, AMD and ARM processors are vulnerable to Spectre.

    2. Re:Already using ARM in the datacenter by arglebargle_xiv · · Score: 2, Informative

      And ARM has TrustZone as its answer to Intel's Management Engine. Not sure which is worse in terms of insecurity.

    3. Re: Already using ARM in the datacenter by Anonymous Coward · · Score: 0

      Intel is even more vulnerable to Spectre than AMD is, so its not which aare they vulnerable to but more...

      Intel: Every vulnerability AMD has, plus IME, Meltdown, etc.

    4. Re:Already using ARM in the datacenter by Roman+Mamedov · · Score: 4, Informative

      > a French hosting company called OVH has been offering ARM-based servers under the brand Scaleway. I've been happy with it

      Scaleway is a brand of the French hosting company Online.net, not of OVH, who are their main competitor.

    5. Re:Already using ARM in the datacenter by KiloByte · · Score: 2

      Counterpoint: since May 2016, a French hosting company called OVH has been offering ARM-based servers under the brand Scaleway.

      Yeah, but both the CPU speed and I/O are atrocious compared to what one would expect. I'm still paying for one but use it solely as secondary DNS -- it can't deliver more than ~30Mbps of static content, and CPU-wise my Odroid-U2 from 2012 runs circles around it.

      Likewise, Graviton is laughably slow even for the oomph-to-price ratio Amazon touts. Sorry but there's no datacenter-type ARM worth even looking at today. Kind of like x86 phones... The split around use cases is pretty entrenched, with only realistic battle going around lowest-end laptops.

      --
      The creatures outside looked from Alt-Right to Antifa; but already it was impossible to say which was which.
    6. Re:Already using ARM in the datacenter by Anonymous Coward · · Score: 0

      The RISC V is not a processor. It is only an instruction set specification.

      As such it is free of spectre vulnerabilities as is any other ISA specification. Including Intel's.

      It would be perfectly possible to create implementation of a processor to the RISC V specification that was vulnerable to spectre.

      Admittedly, so far, no one has done that. And now that the world is more aware of the risks of such covert channels they are unlikely to do so.

    7. Re:Already using ARM in the datacenter by AmiMoJo · · Score: 4, Informative

      Trustzone is very, very different to Intel Management Engine.

      IME is a complete system-on-chip running a hidden OS that is required to boot the machine. It provides services like VNC access to machines that are powered down or haven't booted the OS yet, and completely compromises the real CPU/OS.

      Trustzone is just an extra processor execution mode that offers some security features under the control of the OS running on it. It's basically for secure storage and code validation (for signed binaries etc.)

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    8. Re:Already using ARM in the datacenter by Anonymous Coward · · Score: 1

      Ryzenfall is not actually a problem... you need to be root to get root access!

    9. Re:Already using ARM in the datacenter by Anonymous Coward · · Score: 0

      *Some* ARM processors are vulnerable to Spectre. Most are not.

    10. Re:Already using ARM in the datacenter by OrangeTide · · Score: 1

      both Spectre and Meltdown can apply to non-x86 processors like POWER9. There are patches to help on POWER9 but it did have a performance cost.

      --
      “Common sense is not so common.” — Voltaire
    11. Re:Already using ARM in the datacenter by arglebargle_xiv · · Score: 1

      Security-wise they're no different, both are bug-riddled super-privileged zones of operation that can be used to compromise the main system, and that the main system has no oversight of. Using holes in the secure-because-we-say-it-is TrustZone to compromise the otherwise secure main system is particularly amusing,

    12. Re: Already using ARM in the datacenter by Anonymous Coward · · Score: 0

      It is a problem because of the persistence issue.

      I note that the ryzenfall series of vulnerabilities were more serious than the IME vulnerabilities, but it's an Intel disaster and AMD handwave. The fanboys and AMDs responses when there's an issue is really holding them up in serious data centre applications.

    13. Re: Already using ARM in the datacenter by Anonymous Coward · · Score: 0

      This was entirely factual. You AMD fan boys reflexively downvoting anything critical of AMD really aren't helping AMDs credibility in business applications.

      AMD downplays and stonewalls, and eventually just goes quiet, while using you to do their PR for them. Intel by contrast has been quite open and keen to communicate particularly with their larger customers. If you're wondering why Intel continues to dominate despite AMD having a competitive offerings, this is one of the big reasons.

    14. Re:Already using ARM in the datacenter by AmiMoJo · · Score: 1

      Do you have any info on these issues? As far as I am aware the only known vulnerabilities were in the software using TrustZone, and all the open source implementations have been fixed. And you can of course disable it.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    15. Re:Already using ARM in the datacenter by pnutjam · · Score: 1

      I looked at their site and they seem to offer both bare metal and "cloud" (virtualized?) at that price point. The bare metal has slightly better specs, but I suppose you would sacrifice some reliability.

      I'm curios how that compares to the server I get from time4vps using openvz (12 EUR/ quarter):
      CPU: 1 x 1.70 GHz
      RAM: 512 MB
      Storage: 512 GB
      Bandwidth: 4 TB
      Port speed: 100 Mbp

      It's a "storage" vm, so they don't offer backup. I use it to backup my home system with borg and I run a calibre server and website on it. It's been very stable and I've been very happy with the performance.

      --
      Cheap storage VM.
    16. Re: Already using ARM in the datacenter by Anonymous Coward · · Score: 0

      I thought these were unstable for a long time and requires a special kernel patch? This was why they are not expanding this product line and just resell servers that people cancel?

    17. Re:Already using ARM in the datacenter by arglebargle_xiv · · Score: 1

      I thought this stuff was well-known... maybe they've fixed it now, but at the time TrustZone had absolutely no security mitigations, no ASLR, no DEP, no non-executable heap or stack, as one pen-tester put it you could "hack like it was 1999". So for example Motorola's otherwise quite secure Android Razr cellphones were hacked by exploiting various holes in TrustZone and then attacking the "insecure" phone software from inside the "secure" TrustZone (the call is coming from inside the house!). Given that the TrustZone kernel that was used seems to have been written without any regard for security, I wouldn't put much hope in future versions being much better, it'd be like starting from Windows '95 and hoping for a secure system at some point.

    18. Re:Already using ARM in the datacenter by AmiMoJo · · Score: 1

      What kernel? With TrustZone it's just a processor execution mode, you have to write your own kernel. There are open source ones as well as proprietary ones.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    19. Re:Already using ARM in the datacenter by cerberusss · · Score: 1

      How it would compare, I don't know. It's quite hard to compare these ARM thingies. I mean, four cores yes, but what is their performance? I have no idea how to compare those with Intel stuff, and not really a need either... I just picked the cheapest because my needs are not demanding.

      By the way, that's a very interesting VPS. It would indeed be great for backups, or OwnCloud/NextCloud or what have you.

      --
      8 of 13 people found this answer helpful. Did you?
    20. Re:Already using ARM in the datacenter by cerberusss · · Score: 1

      Thanks for clarifying that! Apparently this competition is working well, because what comes out of the French hosting business is quite frankly amazing.

      --
      8 of 13 people found this answer helpful. Did you?
    21. Re:Already using ARM in the datacenter by arglebargle_xiv · · Score: 1

      That's kinda arguing with semantics, I didn't kill Bob, the bullet fired bu the gun killed Bob, I was just a bystander. To use TrustZone you need to run a TrustZone kernel in it, so for example of your phone has a SnapDragon CPU you get the Qualcomm kernel running QSEOS. Which happens to be riddled with vulnerabilities...

    22. Re:Already using ARM in the datacenter by pnutjam · · Score: 1

      Time4vps has hands down the best price for storage that I've been able to find. Yeah, it's openvz (the also offer kvm, but not in storage nodes), but it performs well as far as I can tell.

      I'm on an older plan where I get 1TB disk and 10TB of bandwidth for that cast, maybe a little less cpu...

      If you get on their mailing list, they offer discount coupons with recurring discounts a couple times a year. I have a 30% off code that might still be good if you drop me an email.

      --
      Cheap storage VM.
  3. Likely be a SocioNext SC2A11 by scdeimos · · Score: 2

    No kidding. It's an SC2A11B in the photo of the motherboard.

    1. Re:Likely be a SocioNext SC2A11 by willy_me · · Score: 3, Informative

      The pictured board is not the one being described. It is for comparison purposes. The board is a "SynQuacer E-Series 24-Core Arm PC Motherboard" -- some info can be found here.

      If this new Banana Pi 24-core board is priced affordably, it could be quite interesting. They should be able to offer a much more affordable board then the referenced SynQuacer.

    2. Re: Likely be a SocioNext SC2A11 by Anonymous Coward · · Score: 0

      Wipe that African OS off it and I'd be interested.

    3. Re: Likely be a SocioNext SC2A11 by Anonymous Coward · · Score: 0

      You could of course do that yourself.

    4. Re:Likely be a SocioNext SC2A11 by Anonymous Coward · · Score: 0

      IIRC the SynQuacer board isn't that expensive (for a non-mainstream board at least). I considered getting one for a while, until I finally dug up the CPU's specs and discovered that it only runs at 800MHz. So while there's a lot of cores, performance is still going to be crap. On the other hand the CPU only uses something like 5W.

      The main selling point of the board to me was that it's one of two commercially available AArch64 board with DIMMs that I know of, instead of the usual 2-4GB of soldered-on RAM. I think you can also buy a ThunderX2 board, but those start at something like $7000.

  4. NVMe by ArchieBunker · · Score: 1

    Does it even make sense to use NVMe? Does the processor have that much throughput?

    --
    Only the State obtains its revenue by coercion. - Murray Rothbard
    1. Re:NVMe by Anonymous Coward · · Score: 1

      There may be other reasons to choose NVMe. Physical size for example, as NVMe is much smaller than a 2.5" SSD.

    2. Re:NVMe by Anonymous Coward · · Score: 0

      With slow cores NVMe should be beneficial for the lower software overhead, theoretically speaking of course.

    3. Re:NVMe by thegarbz · · Score: 2

      You're comparing an interface to a form factor. Not all M2 form factor drives are NVMe

  5. Re: by Anonymous Coward · · Score: 0

    Cool, got an UltraSPARC T2, so everything is all good!

  6. How does it compare to my Pentium? by Anonymous Coward · · Score: 0

    More like Pentiummy!

  7. Well Sure! by Anonymous Coward · · Score: 1

    "ARM-based server processors have threatened to take on Intel in the data center for some time ..."

    Yes, since about, oh, 2005 was it? ARM servers have been "threatening" and "threatening" and "threatening" for so long, that threatening is all they can do. Or so it seems.

    All that threatening must make an ARM CPU tired. Why don't you lay down and take a nap?

    1. Re:Well Sure! by NicknameUnavailable · · Score: 0

      The real selling point is more not being riddled with backdoors.

    2. Re: Well Sure! by Anonymous Coward · · Score: 0

      Most data centers are heavily virtualized these days. Both Hyper-V and ESX have mitigations and the ability to flat out disable vulnerable instructions for the VM hardware. Nobody is seriously looking at ARM for large scale, high availability scenarios.
      Early last year we ran a small proof of concept and the actual performance was truly abysmal. There are bottlenecks all over the damn place, the only potential use case we could find was for all RAM workloads which barely touch network or storage and there just isn't enough RAM on these boards to make that worthwhile.

      If you're looking to build a small home lab that feels like a larger data center these may have some promise. But it's a long way from anything that you'd want to put any important workloads on.

    3. Re:Well Sure! by Anonymous Coward · · Score: 0

      You know, your pithy reply sounds good, except that I've been in this business for a very long time.

      Show me a vendor that has excellent security, and I'll show you a vendor that is so green they haven't been tested. Or they prioritize security so much they have neglected features and competitiveness. Or they have any of a huge pile of problems that tends to rule them out from enterprise adoption and success.

      While I too decry sloppiness in the security domain, I have to deal with the world as it is. Enterprise vendors that respond and patch their security issues have been an adequate response, though it is terribly vulnerable to zero-days. I'm not conspiracy minded enough to view systems like IME and ILO and iDRAC as backdoors. These are very practical for large-scale server support and they do not exist in a vacuum, absent any customer demand.

  8. Why virtualization by DrYak · · Score: 4, Interesting

    How well is x86 virtualization on ARM these days?

    1. qemu-user-mode is a thing and has been for probably the past decade
    (And specialized software like DOSBOX has had a great time on ARM gaming devices)

    2. Why the hell do you need x86 virtualization on an ARM server for the cloud ?!?

    It used to be something important back when most businesses ran Windows servers in-house, running proprietary binary x86 Win32 applications.

    Nowadays the cloud is mostly Linux, and is running mostly interpreted languages (like Python, Javascript, etc.)

    Linux is opensource, you can compile the user land on any architecture that pleases you (e.g.: look for ArmBian, RaspBian, and other such ARM-specialized Debian derivatives, etc.), and then subsequently Node.js and Python3 will happily execute whatever code you throw at them, no matter if they run on a different Arch than the dev's laptop on which they were written.
    If your business use software that you write in an actual compiled-language, you can cross compile or compile on a ARM machine before deployment as part of your devops.

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
    1. Re:Why virtualization by wierd_w · · Score: 2

      There's also Eltechs Exagear desktop, which works similarly to QEMU-usermode. (Faster in some applications, buggy as shit in others. Buyware.)

      Why would you need it though? Unless you have a very specific service daemon that staunchly needs x86, it would be better to use native binaries.

    2. Re: Why virtualization by Anonymous Coward · · Score: 0

      There's a lot of proprietary Linux software out there, used on clouds, that is Intel only, particularly in areas such as engineering.

    3. Re:Why virtualization by Anonymous Coward · · Score: 0

      Horrible. It's not something any sane person would do. Ideal for Linux though where you can recompile and run.

      If you wanted decent performance for x86 apps run the damned things ON x86. Yeah you can probably have them run but they'll be slow and inefficient compared to native x86.

      That's not where this is going to be a win. Medium performance/low power. Where you say want a web server to always be up, don't want to pay the power bills for x86 and never expect insane loads.

  9. Spectre: Depends. by DrYak · · Score: 5, Informative

    However, AMD and ARM processors are vulnerable to Spectre.

    ARM:
    Not all ARM processors do speculative execution. Some cheaper/simpler/lower power/lower cost CPUs do execute things normally (in order, no speculative, etc.)
    And that's precisely the case of the Cortex-A53 used TFA's server board. No speculative execution, no way to spectre them.
    There's no way to exploit them if the exploited functionality doesn't exist.
    (BTW: that used to also be the case on some older Intel Atoms and Xeon Phi)

    AMD:
    Spectre V1 i.e. "speculative execution working as it is supposed to work" (aka.: the behaviour that was already known to exist and was criticized from day 1 of the invention of speculative execution, but was dismissed back then because "what could you possibly learn by loading things into the cache ?". Cue in cache access timing a few decades later and suddenly it seems a bit more nefarious) is the only one that has been successfully demonstrated on AMD CPU.
    And even that one is still an application reading its own memory to which that it already has access to begin with. Nothing freaking and scary, just go learn to keep 3rd party executed code and critically important data separated (e.g.: don't run your password management plug-in in the same web browser process as the JavaScript in tabs that runs any shit you pull from the internet), because you're just one bug away from disaster, be it Spectre or not.

    Most of the freaky ones (Meltdown) aren't affecting, because AMD engineers don't have a tendency to throw all security through the window just to scrape a few cycles in some benchmark.
    Others (like Spectre v2) haven't been successfully demonstrated in practice: AMD CPU *do* indirect branch prediction, but have much more complex predictors which are difficult to use reliably and any way the CPU speculates a lot less ahead so there isn't that much you can do here, the later being also true for the couple of other variants that also exist on recent AMDs.

    TL;DR: there are orders of magnitude of difference between how Intel CPU are affected by Spectre and every body else in the CPU market.

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
    1. Re:Spectre: Depends. by drinkypoo · · Score: 1

      TL;DR: there are orders of magnitude of difference between how Intel CPU are affected by Spectre and every body else in the CPU market.

      Is that true? What about IBM POWER? That's vulnerable to MELTDOWN as well, is it as vulnerable as Intel to SPECTRE-type attacks?

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  10. Re: by Anne+Thwacks · · Score: 1
    Cool, got an UltraSPARC T2, so everything is all good!

    It may be cool, but what about the 85dB noise level?

    I would love the people making these ARM boards to make them for factor compatible with Oracle motherboards so they can upgrade old Sun/Oracle kit. It would be an interesting way to compare performance if you could swap them in the exact location, just switching the cables to the unit above or below.

    --
    Sent from my ASR33 using ASCII
  11. WARNING: Horrible hardware support! by Anonymous Coward · · Score: 5, Insightful

    I did a lot of work with "Banana Pi" singe-board computers, and everything thhat comes out with that name, has atrocious hardware support, atrocious hardware and atrocious support.

    The hardware is usually barely documented, and does *not* qualify as open hardware, the support forums are censored in Chinese face-saving tradition and full of false promises, and in many, the board design can only be described as *failed*. (E.g. the power regulator not supporting the storage power requirements when powered via USB port. Or the storage and gigabit network interfaces maxing out at 400 MB/s even with tuning.)

    Unless you want to be in a world of madness and pain, don't buy them.

    (And I'm not one of those who think that China is crap in general. I was very happy with my Blackview BV6000, in all aspects.)

    1. Re:WARNING: Horrible hardware support! by drinkypoo · · Score: 1

      I did a lot of work with "Banana Pi" singe-board computers, and everything thhat comes out with that name, has atrocious hardware support, atrocious hardware and atrocious support.

      That's how I feel about AllWinner in general, which is who makes the SoCs they use. I guess my very old AllWinner CPU finally is supposed to work with a mainline kernel, and the only binary blob you are supposed to need after boot is for the GPU, but for literally years the mainlining effort appeared stalled and there were no meaningful updates.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    2. Re:WARNING: Horrible hardware support! by bluefoxlucid · · Score: 1

      I've been hoping for a RISC-V SOC excluding networking, with 1 core 4-way SMT (yes I know), SATA-III, and PCI-e. The PCI-e would allow manufacturers to add a discrete ethernet/wifi/bluetooth chip (some of us want no such thing on the board) and USB 3.1 controller (upgradeable by swapping the chip rather than redesigning the SOC); the headers on the board would be USB-C, of course.

      With the new modular designs, it'd be easy to put the USB3.1 on the SOC and simply cut that part off and stamp in a USB4.0 or whatnot later. If you want some boards without networking, the two-chip design is probably less expensive than producing two types of SOC at every combination of frequency and RAM.

      --
      Support my political activism on Patreon.
    3. Re:WARNING: Horrible hardware support! by Anonymous Coward · · Score: 0

      Yep , they burnt me. I was an early purchaser of the M3 before a lot of the details came out.. "sata" my ass. I guess having the plug on board is convenient... but that is about it.

      And drivers? ya, good luck with that one.

    4. Re:WARNING: Horrible hardware support! by Anonymous Coward · · Score: 0

      Oh snap!!! I came here and posted the same thing!!!

  12. OMG ARM is North Korea! by Anonymous Coward · · Score: 0

    I bet 16 of those 24 cores are just photoshopped in. ;)

  13. What a fucking shithole is this by Anonymous Coward · · Score: 0

    Holy shit

    1. Re:What a fucking shithole is this by Anonymous Coward · · Score: 0

      Maybe you should emigrate from the U.S.A

  14. mate on a server? and weak IO as well. by Joe_Dragon · · Score: 1

    mate on a server? and weak IO as well.

    only 2 sata and pci-e X16 at X4??

  15. Re: by Anonymous Coward · · Score: 0

    Cool, got an UltraSPARC T2, so everything is all good!

    It may be cool, but what about the 85dB noise level?

    I'm thinking that not-so-old SPARC hardware looks pretty cool, and wouldn't mind going full immersion Solaris for kicks. How about water cooling and dumping eight of those fans.

  16. why by Anonymous Coward · · Score: 0

    Why can't the OS see all the RAM? Genuinely curious.

  17. Re: by Anonymous Coward · · Score: 0

    Have you checked Solaris license fees? I gave up some old Sun workstations when the formerly free Solaris 10 (from Sun for developers) got "monetized" after Oracle's takeover. Getting too far out of date would leave it open to lots of other attacks, and incompatibilities..

    RO

  18. How long does it take? by RogueWarrior65 · · Score: 1

    The demo shows building a Linux kernel but how long does that take? How long does Yocto take to build on such a machine?

  19. Re: by Anonymous Coward · · Score: 0

    Dang, 10 was what I was thinking about using too. But I'm on my usual 'bsd box now and it's just so comfortable, if it aint broke...

  20. But its a banana by Anonymous Coward · · Score: 0

    Traditionally they dont have good driver support. Other hardware issues tend to creep in too. Also, the cost of this will have to be silly...

  21. I have a bananna pi r2 by Anonymous Coward · · Score: 0

    It's a decent piece of hardware but it years of user support to get anything working. Seriously go look at the forums.
    http://forum.banana-pi.org/c/Banana-Pi-BPI-R2
    Maybe it's better now but last year was the last time I tried to use it and it was such a pain that I dropped it from my project.
    Lot of things along the lines of "Hi guys here is my latest image, docker works and now wifi works as long as you don't do X and Y, and here are instructions how to enable the vlan feature on the switch... btw you'll need to reconfigure the bootloader and then reboot every time you want to reconfigure a vlan."
    I'm not even sure if the guy worked for them or was just a determined hobbyist but to use the advertised features you needed to either find the right driver and get it working yourself or wait for a guy named Frank to post an image with the features you wanted. Sometimes he struggled to get certain features to play nice.
    There were other problems I saw people talk about and I ran into a bunch of my own but I've forgot them.
    I still plan on using it to replace my home wifi router at some point.

  22. Nice Work by Anonymous Coward · · Score: 0

    Finding pick up lines

  23. The ARM Factor by Xnet+Project · · Score: 1

    Thomas Dorr made an interesting note on this project.

    According to Thomas's post this looks to be similar to the following:

    https://www.96boards.org/produ...

    With a few modifications this could look to be ARM SOC's being put into mainstream PC use.

  24. Year End Budget by Anonymous Coward · · Score: 0

    I bet this is just something where they had money left in their 2018 R&D budget and didn't want it reduced next year so they found a way to use up that money quickly.