Slashdot Mirror

← Back to Stories (view on slashdot.org)

NVIDIA's $99 Jetson Nano is an AI Computer for DIY Enthusiasts (engadget.com)

Posted by msmash on from the how-about-that dept.

Sophisticated AI generally isn't an option for homebrew devices when the mini computers can rarely handle much more than the basics. NVIDIA thinks it can do better -- it's unveiling an entry-level AI computer, the Jetson Nano, that's aimed at "developers, makers and enthusiasts." From a report: NVIDIA claims that the Nano's 128-core Maxwell-based GPU and quad-core ARM A57 processor can deliver 472 gigaflops of processing power for neural networks, high-res sensors and other robotics features while still consuming a miserly 5W. On the surface, at least, it could hit the sweet spot if you're looking to build your own robot or smart speaker. The kit can run Linux out of the box, and supports a raft of AI frameworks (including, of course, NVIDIA's own). It comes equipped with 4GB of RAM, gigabit Ethernet and the I/O you'd need for cameras and other attachments.

127 comments

  1. Nvidia, lol. by Anonymous Coward · · Score: 0

    No thanks, I don't buy from companies that cheat benchmarks, buy reviews and cripple games with shady developer programs. I'll buy anything else instead of Nvidia. They are scumbags.

    1. Re: Nvidia, lol. by Anonymous Coward · · Score: 0

      I love Nvidia postings because they bring out the butthurt Slashfags in social justice boycott mode.

    2. Re: Nvidia, lol. by Anonymous Coward · · Score: 0

      Justice has always been a social construct. Even the code of Hammurabi required weighting things as equal when they were different, based on large scale social acceptance. Every change has required the same basis.

    3. Re: Nvidia, lol. by Anonymous Coward · · Score: 1

      I love salty cunts desperately needing a basic moral education/beatdown and think they're "edgy" for being spineless, nutless little apologist twinks like you, Kendall. There will be consequences, lol.

    4. Re:Nvidia, lol. by vux984 · · Score: 1

      So who else makes these?

    5. Re:Nvidia, lol. by Anonymous Coward · · Score: 0

      Honestly, for quick AI projects I usually just use an FPGA board. You can do very specialized things as well as AI type stuff and DSP algorithms. And the software tools are a bit better (well, I find VHDL quite a bit better). For one off projects I tend to use something like this from Arrow: http://www.novtech.com/chameleon96

      It's easy to buy it and get it going. It has a 40 IDC header which is easy to breadboard/prototype stuff with. I just don't like Nvidia chips. I try to never put them in a design if I have any say.

  2. Another Slashvertisement? by fbobraga · · Score: 0, Flamebait

    Slow down, cowboy!

    1. Re: Another Slashvertisement? by Anonymous Coward · · Score: 0

      To bad the only other thing is a crappy/crappier company which does the same thing..amd? Yeah almost forgot their name.

    2. Re: Another Slashvertisement? by DontBeAMoran · · Score: 1

      NVIDIA, AMD... all made in Taiwan.

      --
      #DeleteFacebook
  3. Managed by Spacely by Anonymous Coward · · Score: 1

    You have to type JETSON! to get it to do anything.

  4. Marketing by Anonymous Coward · · Score: 1

    Sophisticated AI generally isn't an option for homebrew devices when the mini computers can rarely handle much more than the basics

    That's a marketing failure: many people think cheap toys like the raspberry pi are the only game in town.
    You can get small form factor FPGA boards in the $100-200 range.

    1. Re:Marketing by DontBeAMoran · · Score: 1

      And you're a math failure. FPGA isn't even a valid option to a Raspberry Pi. You can get 20~40 Raspberry Pi Zero for that price.

      --
      #DeleteFacebook
    2. Re:Marketing by Anonymous Coward · · Score: 0

      Math failure? You must've been triggered by me calling the RPi a cheap toy. And hey, you said it: "You can get 20~40 Raspberry Pi Zero for that price."
      >FPGA isn't even a valid option to a Raspberry Pi.
      For AI computations, it absolutely is. FPGAs (and GPUs) run circles around any raspberry pi for AI computations, the very topic of this article.
      My point stands: due to poor marketing, very few people know off the top of their heads that there exist actual powerful single board computers. In a month or two people will probably forget about the Jetson Nano as well.

    3. Re:Marketing by DontBeAMoran · · Score: 1

      The Jetson what?

      --
      #DeleteFacebook
    4. Re: Marketing by Anonymous Coward · · Score: 0

      You are absolutely correct. In fact it becomes increasingly clear that FPGAs are the best compute framework for these types of things.

  5. Nvidia Cannot Be Trusted With Anything IMHO by dryriver · · Score: 0, Offtopic

    Nvidia claims that the realtime raytracing tech in its uber-expensive RTX 20xx cards "took 10 years to develop". Hmm. A full 10 years to turn decades old raytracing techniques into working hardware circuitry! Either Nvidia's hardware engineers have an IQ in the double digits and take ages to implement old techniques in hardware. Or good old Nvidia is one of those lovely tech companies that _could_ have given us exactly what we wanted 6 to 7 years ago, but simply opted to sell old tech for another few years instead - it was more profitable that way, you see? GPU buyers are idiots - Nvidia gives its architecture a radical name change every few years (Maxwell, Turing, bla bla bla), unlocks a bit more of the processing potential the architecture had 5 years ago, and boasts - GASP! - that the "new" architecture is now 20% faster. I wouldn't be surprised if Nvidia already has RTX 40xx or even RTX 50xx GPUs running in its laboratory. When will we get those babies? WHEN NVIDIA FEELS LIKE IT OF COURSE. LMAO!

    --
    Why did the chicken cross the road? Because Elon Musk put an AI chip in its head.
    1. Re: Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      Raytracing scales to the number of pixels on the screen. Luckily the slow growth in resolution of workstation monitors (30 year old displays were 1280x1024 and higher) and the exponetial growth of transitory count meant that real-time raytracing would eventually be possible. U fortunately that day isn't today and we're a good 3-5 years shy of proper full 60 fps and full resolution RT.

    2. Re:Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 1

      May want to loosen than tinfoil hat a little bit, buddy. You are moving right into crazy and /. is full up on that.

    3. Re:Nvidia Cannot Be Trusted With Anything IMHO by yodleboy · · Score: 1

      "GPU buyers are idiots" Are you suggesting we should all switch to AMD cards? I think I'll stick with my "5 year old architecture" that the new hotness from AMD can usually barely keep up with while using way more power. For that matter, do you really think Intel, AMD (cpu and gpu), apple, or any other tech company is NOT sitting on technology? There is no GPU conspiracy from Nvidia...

    4. Re: Nvidia Cannot Be Trusted With Anything IMHO by Type44Q · · Score: 1

      No; twenty year old displays are 1280x1024. Thirty years ago, you were lucky if you had an accelerator that could do 800x600@16bpp (I seem to recall those little IBM 12 or 13" VGA monitors could actually do 1024x768... at something like 35hz vertical!)

    5. Re:Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      Jesus Christ, what entitled morons modded you up?

      simply opted to sell old tech for another few years instead - it was more profitable that way,

      Those fucking bastards! Trying to turn their work into profit! They should throw their old fabs in the dumpster whenever they create better technology. Fuck return on investment!

      GPU buyers are idiots - Nvidia gives its architecture a radical name change every few years (Maxwell, Turing, bla bla bla), unlocks a bit more of the processing potential the architecture had 5 years ago, and boasts - GASP! - that the "new" architecture is now 20% faster. I wouldn't be surprised if Nvidia already has RTX 40xx or even RTX 50xx GPUs running in its laboratory. When will we get those babies? WHEN NVIDIA FEELS LIKE IT OF COURSE. LMAO!

      Yeah, talk about morons. Paying money for the best available tech, when all they had to do was create their own GPUs and go into competition with NVidia and destroy them! What are these morons thinking? Are they discouraged just because AMD and Intel can't compete with NVidia? How weak are they??
      I don't know about anyone else, but you've totally convinced me that NVidia - ney - capitalism just doesn't work.
      This whole supply and demand thing is broken. Everyone should get everything at whatever price they think they should pay right?

    6. Re: Nvidia Cannot Be Trusted With Anything IMHO by The+Grim+Reefer · · Score: 1

      No; twenty year old displays are 1280x1024.Â

      Maybe cheap crappy ones. Of the half dozen monitors I had back then the smallest was 19" that was 1600x1200. Up until 4 years ago I still had a pair of Eizo medical grade 1600x1200 21 inch monitors from 1998 that I kept. They weren't cheap, but I got tired of moving them as they weighed 90lbs. each. I had an SGI monitor that I bought in 1997. I don't remember what the resolution was any more, but it was higher than the Eizo monitors. .

    7. Re:Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      "May want to learn English so you can speak in complete sentences, buddy."

    8. Re: Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      You know nothing about, well pretty much anything. Except you're mother gives really good head and you'll be getting a little brother in about 9 months.

    9. Re: Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      Yea it was 1600x1200.

      Considering you're comparing the processing power required for "rasterization" vs. The processing power for "Ray tracing", I'd say you don't understand the difference between the two at all.

      There s a reason they developed separate processing units to streamline ray-tracing rendering. They are literally explaining how translating rasterization-focused processors are poor performers at ray-tracing. Why do you think CG movies needed farms of GPUs and weeks of processing time to batch out footage?

    10. Re: Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      Most monitors able of 1600x1200 would do 1600x1200 at 60Hz, on a monitor with fast phosphors. Pretty useless unless you have a good reason for it.
      I got a mid-2000s 22" saved from trash, liked it a lot (until it quit working) but if you'll run it at 100Hz the max resolution gets pretty low - at or near 1280 wide.
      I had it late enough that I could try Linux Mint in 'hi dpi' mode on it! (2048x1536 60Hz). Funny but impractical. Some very best monitors might do 2560x1920, never seen that, I don't know how of a chore and reliability issue that is.

    11. Re: Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      In 1990 I got my boss to spring for a Super VGA card that would do 800x600 @ 8 bits, and 1024x768 @ 4 bits. Hi-res Windows 3 in 16 colors, woo! It was a *big* deal...though I don't remember it being particularly expensive. MultiSync monitors were fairly new.

      Most people at my workplace were still using 640x480 5 years later. I switched one co-worker's PC to 800x600 and she told me to change it back (too small to read).

    12. Re:Nvidia Cannot Be Trusted With Anything IMHO by Anonymous Coward · · Score: 0

      Either Nvidia's hardware engineers have an IQ in the double digits and take ages to implement old techniques in hardware.

      Mostly likely the patents expired...

  6. Or perhaps it's from GTC? by Anonymous Coward · · Score: 0

    Today is the first day of NVIDIA's GPU Technology Conference in Silicon Valley. It's always accompanied with new launches.
    https://www.nvidia.com/en-us/gtc/

  7. Finally a board with some RAM by drinkypoo · · Score: 4, Interesting

    4GB puts this into the category where it's actually useful for stuff like web browsing. Sadly, the link to the item from TFA is 404, but it looks like it's actually got enough ports on it to be useful for doing stuff without needing a hub, too. Forget building robots with it, you can build kiosks. Do they have an Android build for it?

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    1. Re:Finally a board with some RAM by 110010001000 · · Score: 1

      4GB for web browsing???

    2. Re:Finally a board with some RAM by drinkypoo · · Score: 1

      4GB for web browsing???

      Welcome to 2010, let alone 2019. The browser will commonly eat up 2GB by its lonesome. Mine is using 1.8GB right now, and that's while using noscript and blocking all kinds of crap. If you want anything else to be able to happen at the same time you're browsing, without needing to swap, then you need more than 2GB RAM — which is where most of the cheap boards top out.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    3. Re:Finally a board with some RAM by Anonymous Coward · · Score: 0

      4GB for web browsing???

      Yeah, you're right - seems a bit underpowered for chrome or FF.

    4. Re:Finally a board with some RAM by Anonymous Coward · · Score: 0

      Single board computers with an arm SoC are pretty hot right now for hackers, developers, makers. You've got the Pi with Broadcomm chips, a lot of others with Rockchip and Allwinner SoCs.

      This looks like Nvidia's entry, which could be pretty interesting. If it's well documented, well supported, and has good drivers that lets makers get at the GPU for compute and graphics It'll be worth looking at.

      Computer vision applications in particular need a lot of muscle - But the GPUS from the other vendors are designed to drive tablets, set top boxes, thin clients, media players, etc. Hopefully Nvidia is bringing it's GPU compute tools to the single board computer space.

    5. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      Runs lynx just fine.

    6. Re:Finally a board with some RAM by drinkypoo · · Score: 1

      This looks like Nvidia's entry, which could be pretty interesting. If it's well documented, well supported, and has good drivers that lets makers get at the GPU for compute and graphics It'll be worth looking at.

      That's always the question, isn't it? Nvidia actually controls the IP of their ARM SoC GPU line, unlike GeForce (which is encumbered at least by agreements with Microsoft, if not others) so they have the option to release the driver sources. But will they? Still, odds are it'll be better-supported than anything from Allwinner.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    7. Re:Finally a board with some RAM by Anonymous Coward · · Score: 0

      Yeah, gramps, ARPANET has pictures now.

    8. Re:Finally a board with some RAM by quenda · · Score: 1

      What do you mean "finally"?
      Ebay, Amazon etc have a plethora of ARM-based mini-PCs with 4GB under $99. Well under.
      Typically sold with Android as set-top boxes, they can be used for web browsing, office PCs, etc.
      Some will work with Linux, but remember there is no Linux/ARM version of full Chrome.

    9. Re:Finally a board with some RAM by Shaitan · · Score: 3, Interesting

      Sad but true and there really is no good explanation for it. I just closed everything an opened chrome, clocked in at 350mb. Loaded slashdot as the only page, suddenly it skyrockets to 850mb and then slowly settles back to 550-650mb. Now there definitely isn't enough content in this page to explain 10mb and even with the linked pages you have nowhere near the 200mb it has absorbed in content. It does make me wonder just what the hell it is using so much memory for.

      Compared to the 4mb footprint of Netscape 4 or IE 3.5 you've gained what... some adjustments to javascript and css? That explains maybe 10mb of the increase. A gargantuan cache? That would explain another 64mb maybe. The actual page content? If anything the pages actually have less content with html, css, and light weight icons being the styling of choice these days it's almost all text. A 2mb page would be massive but they use shitty autochurned output for most sites these days so call it 8mb, across all the linked pages that will fill your 64mb cache. With a little breathing room added in that is what about 100mb that can be explained?

    10. Re:Finally a board with some RAM by Shaitan · · Score: 3, Interesting

      It takes A LOT of pictures to explain 2GB. Good luck coming up with something in browsers that explains their memory consumption. Even with the... dear god 3600 lines of text that composes this page (vs the 50-75 you could do it in, maybe 150-200 if you were counting js and css). There just isn't any magic here. Even that ugly video enabled ad that is probably larger than the entire browser should be would only explain a few mb. This page and every linked page still isn't going to explain more than 100mb and the entire browser should itself should use less than that.

      I blame people obsessed with OOP

    11. Re:Finally a board with some RAM by drinkypoo · · Score: 1

      What do you mean "finally"?

      I mean finally featured on Slashdot, where the last several which were slashvertised had only 2GB, or even less.

      Ebay, Amazon etc have a plethora of ARM-based mini-PCs with 4GB under $99. Well under.

      Any idea which are worth a crap?

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    12. Re:Finally a board with some RAM by quenda · · Score: 1

      Any idea which are worth a crap?

      For what purpose? If you are just buying one for personal use, save yourself effort by spending a bit more for Atom instead of ARM.
      The software will be easier.

    13. Re:Finally a board with some RAM by drinkypoo · · Score: 1

      For what purpose? If you are just buying one for personal use, save yourself effort by spending a bit more for Atom instead of ARM.
      The software will be easier.

      Intel can die in a fire.

      Finding these things on Amazon is hard. The search listings wind up mixed with all kinds of unrelated garbage.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    14. Re:Finally a board with some RAM by quenda · · Score: 3, Funny

      Try searching for "4gb kodi box". Then narrow down to those with an SoC that has decent drivers for your OS of choice.
      Are they any good for you?

    15. Re:Finally a board with some RAM by drinkypoo · · Score: 1

      Try searching for "4gb kodi box".

      Heh. I should have guessed. But I didn't. Thanks!

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    16. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      Fuck Slashdot!

    17. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      Ahhhhhh 100100101001 know nothing

    18. Re:Finally a board with some RAM by Anonymous Coward · · Score: 0

      Yeah, but do they have SATA though?

    19. Re:Finally a board with some RAM by Anonymous Coward · · Score: 0

      Yes cache IS NOW ILLEGAL!

    20. Re:Finally a board with some RAM by quenda · · Score: 1

      I'm not sure why that was modded funny.

      The items sold as "kodi" boxes make good cheap general-purpose ARM PCs.
      The Kodi thing means mass-market and cheap.
      If you do actually want to run Kodi on them, it is best to wipe the supplied software, and install LibreElec.

      Obviously this lacks the AI power of the above NVidia, but thread-starter was looking for a general-purpose ARM box with 4GB.

    21. Re:Finally a board with some RAM by Anonymous Coward · · Score: 0

      If you buy this for performance and features (e.g. nvidia CUDA, OpenGL 4.5 etc.) you will obviously use the nvidia proprietary driver. I think that's it, case closed. At least it will be the same as on x86 Linux PC, x86 FreeBSD and Windows.

    22. Re:Finally a board with some RAM by AmiMoJo · · Score: 1

      Web browser memory use must be the most misunderstood topic in tech these days.

      Firstly, unused RAM is wasted RAM. If it's not used for anything else you might as well use it for cache. Modern operating systems support this, allowing applications to allocate RAM for caching but immediately release it if something else needs to use it.

      For maximum performance web browsers do a lot of caching. HTML5 itself is quite complex - the browser has to build up a "document object model" that allows CSS and Javascript to interact with it. CSS and Javascript are also quite complex and for speed browsers use just-in-time compilation with the results cached.

      On top of that you have the actual rendering. To make scrolling smooth much more than you can see is rendered. Most browsers use a tile system that splits the page up. If the system is otherwise idle and has lots of free RAM it might as well background render the whole page, and then when you scroll it's just moving a bunch of tile bitmaps around.

      Of course if your system isn't idle or needs RAM for something else that stuff is the first to get purged, and browser performance suffers as a result.

      Firefox reduces its memory footprint by deferring some of this stuff until a tab becomes active. There are performance trade-offs for that, but to their credit Mozilla has some an impressive job of balancing them.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    23. Re:Finally a board with some RAM by UnknowingFool · · Score: 1

      The purpose varies greatly as does the capability. On the low end of both price and capability are the Rasperry Pi models. The most expensive board there is about $35 US. The cheapest is around $10. They make great little kits for things like sensors boxes and retro gaming. Obviously this NVIDIA board is far more powerful and far more expensive.

      --
      Well, there's spam egg sausage and spam, that's not got much spam in it.
    24. Re:Finally a board with some RAM by Shaitan · · Score: 1

      "Firstly, unused RAM is wasted RAM. If it's not used for anything else you might as well use it for cache. Modern operating systems support this, allowing applications to allocate RAM for caching but immediately release it if something else needs to use it."

      I mentioned caching. There is nothing new in this, there is a performance hit if there is contention for that memory which is why the default cache gets set to a reasonable value for an application like a web browser, is user tunable, and the rest should remain for other applications which actually need more cache.

      This attitude needs to die, Moore's law is dead and we need to start optimizing applications again. Most of this logic started as performance increased in game development where the application basically took over and typically ran full screen. This logic made sense. It doesn't make sense when the same logic is used by every shared application and library it doesn't make sense when overspending resources becomes the universal default.

      "HTML5 itself is quite complex - the browser has to build up a "document object model" that allows CSS and Javascript to interact with it. CSS and Javascript are also quite complex and for speed browsers use just-in-time compilation with the results cached."

      Ummm... yeah. And web browsers have back buttons and cache menus and bookmarks too! I can name random stuff people complaining about bloat are already aware of as well.

      Everything you are saying here is something I already referenced in my comment but with a tone that suggests that somehow it explains where it has all gone. It isn't going to either because modern browsers are insanely bloated with a ridiculously oversized footprint..

    25. Re: Finally a board with some RAM by deppman · · Score: 1

      What do you mean "finally"? Ebay, Amazon etc have a plethora of ARM-based mini-PCs with 4GB under $99. Well under. Typically sold with Android as set-top boxes, they can be used for web browsing, office PCs, etc. Some will work with Linux, but remember there is no Linux/ARM version of full Chrome.

      I am one of the first owners of the of Jetson TK1 board. The benefits of this board are a) it runs Ubuntu out of the box, b) it has more powerful GPU, and c) it is supported and well documented by Nvidia. There is a native build of Chrome on ARM - see the Acer Chromebook 13, which also uses the TK1 SoC - and it runs quite well. However, I never did get that build for my Jetson as Google limited its distribution. There may be a way around that these days. I did however get Chromium to work like Chrome through some additional software. The browser ran well on top of Ubuntu with 2GB of RAM. It would however run out of memory and crash if one loaded multiple tabs with heavy content. Double the memory should some that issue.

    26. Re:Finally a board with some RAM by thegarbz · · Score: 1

      This attitude needs to die, Moore's law is dead and we need to start optimizing applications again.

      What do you think that cache is being used for? Preloading functions that may be used by the JS engine. JIT compilation where possible and loading in memory so it's faster when available.

      For all the complaints about RAM usage of browsers one thing that has actually increased consistently with nearly every version is actually the raw computational performance. Page rendering has been getting faster, javascript performance is many orders of magnitude better than the days of low RAM Netscape.

      Browsers are complex. They provide a functionality previously not even dreamt about, and for all the complaints they do so quite efficiently.

    27. Re:Finally a board with some RAM by thegarbz · · Score: 1

      Good luck coming up with something in browsers that explains their memory consumption.

      How about pre-loading all functionality possible in case a page needs it so it is accessed faster?
      How about the few lines of code going through an full JIT compilation so it runs faster?
      How about the fact that we're not throwing text on a screen but rather rendering on a complex canvas that can these days do pretty much anything?

      For all the complaints about memory usage increasing over the years, barring actual memory leaks you can almost linearly trend these increases with various performance increases in the browser itself.

      Please, use all my RAM if the system is faster as a result. It's why I bought it.

    28. Re:Finally a board with some RAM by AmiMoJo · · Score: 1

      I mentioned caching. There is nothing new in this, there is a performance hit if there is contention for that memory

      Yeah... I don't think you understand how this kind of caching works. That memory isn't reserved, it's used when it is free but as soon as an application wants it the cached data is wiped and it is handed over for use. The performance penalty is tiny - the RAM has to be cleared on allocation anyway so it's just a little bit over accounting overhead, for a massive overall speed boost.

      All modern operating systems support this and already do it for things like disk cache.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    29. Re:Finally a board with some RAM by Shaitan · · Score: 1

      How about pre-loading all functionality possible in case a page needs it so it is accessed faster?
      How about the few lines of code going through an full JIT compilation so it runs faster?
      How about the fact that we're not throwing text on a screen but rather rendering on a complex canvas that can these days do pretty much anything?

      The first was always the case. The second is CPU not ram.

      The third is mostly bloat but hey why not implement a more and more complex canvas to support a platform that is nothing but template driven cookie sites. Photoshop 4 might be unimpressive relative to modern Photoshop but it had a canvas to shame anything in the modern browser and weighed in at 26MB.

      "For all the complaints about memory usage increasing over the years, barring actual memory leaks you can almost linearly trend these increases with various performance increases in the browser itself."

      Impressive given that they render pages more and more slowly. As far as I can tell any gains have been inefficient implementations to support and partially offset the inefficient practices of modern web developers.

      "Please, use all my RAM if the system is faster as a result. It's why I bought it."

      Me too but that isn't what we are talking about, we are talking about using an order of magnitude more memory or more to support poor an inefficient implementation, rinse and repeat, across every layer of the system and every application. Importing massive libraries and frameworks which eat tons of resources and carry thousands of bugs because a developer is too lazy to "reinvent the wheel" for a couple three or four line basic function. And then layer framework after framework on top of that because the funky custom objects of framework A don't work in framework B and then framework C because the dev didn't even know everything he needed was already loaded along with A and B because those things are so damn abstracted nobody knows everything that is in them.

      The same is true for the execution with all our gains and advancements being used to subsidize lazy development. With the death of Moore's law, this is a problem. People contrast modern horribly inefficient abstraction practices with the tight asm instruction packing of early days where people were squeezing everything into a few k of ram with 1mhz processors using clever and unmaintainable tricks. Believe it or not, there are practices that are dramatically more efficient which are neither of those two extremes.

    30. Re:Finally a board with some RAM by Shaitan · · Score: 1

      "as soon as an application wants it the cached data is wiped and it is handed over for use. The performance penalty is tiny"

      That "tiny" hit happens millions of times, so it isn't tiny. It's degrading memory performance across the system.

      Excessive caching is not massive overall speed boost. The vast majority of that cache is never used and almost all of that boost can be gained with a cache the fraction of the size.

    31. Re:Finally a board with some RAM by Shaitan · · Score: 1

      "Browsers are complex. They provide a functionality previously not even dreamt about, and for all the complaints they do so quite efficiently."

      Before browsers? Sure. Since netscape? Not really. The original FF was a huge improvement and rewrite. At this point it is more of a bloated and crappy implementation than Netscape was. The only big advance you actually see being used is javascript communication back to the server. It's a browser, it needs to run rendering of a markup language, css markup, and some lightweight client side scripting for display interaction and input validation. Everything else is bloat being stacked on trying to use a screwdriver as the one tool to rule them all. Do you know what happens when you don't do that? Other applications actually made to run and perform the task do the job and optimize as appropriate for their actual application.

      "What do you think that cache is being used for? Preloading functions that may be used by the JS engine. JIT compilation where possible and loading in memory so it's faster when available."

      And yet, everything actually loads and renders more slowly because of all the bloat capabilities being packed in the browser that more than offset any gains.

      And what kind of horribly implemented script parser requires more than maybe 30mb loaded in its entirety into ram in the first place? That isn't a trim parser by the way, 30mb is a massive amount of space for a solution like that, a trim solution would be single digits.

      At this point everything is bloated with the bad development practices you found in AOL, Netscape 5+, Internet Explorer 4+, Photoshop 5+ systems got faster and faster and these things just got slower and slower and because they enabled cheap, lazy, and fast development the dev practices that produced this garbage propagated.

      You can get most of the performance you are talking about with targeted caching and carefully selected best performing optimizations and orders of magnitude gains beyond that just by going back and eliminating dependencies, cruft, misfeatures, and actually spending three months focusing on the optimization that everyone says "offers tiny gains that aren't worth it" until you just write code efficiently as a matter of habit that doesn't require constant thought. If you do, the common wisdom will actually be true and it won't be worthwhile to micro-optimize and refactor your code except on the hot spots the profiler highlights.

    32. Re:Finally a board with some RAM by AmiMoJo · · Score: 0

      If your application is allocating memory millions of times over any significantly short timescale then you fucked up.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    33. Re:Finally a board with some RAM by drinkypoo · · Score: 1

      That memory isn't reserved, it's used when it is free but as soon as an application wants it the cached data is wiped and it is handed over for use.

      That's how disk cache works, not application cache. Just the browser and [modded] FO4 will cause FO4 to run out of memory and die on my 16GB PC. Sometimes I can kill the browser at the low memory warning and have FO4 not die.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    34. Re: Finally a board with some RAM by quenda · · Score: 1

      There is a native build of Chrome on ARM - see the Acer Chromebook 13,

      Yes, but that is not real Linux, as you discovered. It is closer than Android, which also uses the Linux kernel and has Chrome.

    35. Re:Finally a board with some RAM by thegarbz · · Score: 1

      Please learn to quote, your post is hard to read.

      The third is mostly bloat but hey why not implement a more and more complex canvas to support a platform that is nothing but template driven cookie sites.

      You're a very jaded negative person. For your "template driven cookie sites" the rest of the world is busy using full embedded office applications. Just because you don't make use of the full functionality available doesn't mean it doesn't exist and doesn't need necessary resources available. One man's bloat is another's critical feature.

      Impressive given that they render pages more and more slowly.

      Except they don't. Direct comparisons have shown improvements in load times over the years, in many case quite dramatic ones. Did you miss the damn speed wars of the past 5 years?

      to support poor an inefficient implementation

      To claim something has a poor or inefficient implementation you need to first point to a feature comparable implementation that does it better. Do you have examples? The only three proper feature comparable web browsers are Edge, Firefox, and Chrome, and they are comparable in memory usage for a reason. The reminder will be missing some feature (see above comment about critical features).

      Importing massive libraries and frameworks which eat tons of resources and carry thousands of bugs because a developer is too lazy to "reinvent the wheel" for a couple three or four line basic function.

      Well since Chromium and Firefox are open source can you point to an example of developers of it being lazy? Or are you talking about web pages themselves when we started talking about browsers?

      People contrast modern horribly inefficient abstraction practices with the tight asm instruction packing of early days where people were squeezing everything into a few k of ram with 1mhz processors using clever and unmaintainable tricks.

      Those still exist, and computationally we're as efficient as we ever were. They are buried underneath modern GUI and fancy presentation (see above comment about critical feature). That's not developers being lazy, that's customer expectations. You want that to change, you're looking in the wrong place.

    36. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      Blah blah blah.
      That TLDR might matter if, when I click reply, then hit back, the browser didn't have to completely rebuild the whole page and lose my position half the time.

      There are tons of explanations for why browsers are as slow and bloated as they are. There are no excuses.

    37. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      Odroid HC series is nice. Comes with stackable big old heat sink that doubles as a hard drive mount. You can't boot off the sata, but for $60 it is a nice NAS candidate with gigabit. They have an Atom/pentium based one too, but i haven't grabbed one yet. Higher than MSRP on Amazon.

    38. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      What do you consider "real Linux"?

    39. Re: Finally a board with some RAM by quenda · · Score: 1

      It should be clear from the context that I'm referring to a full Linux/GNU desktop OS with X-windows.

      Why do dumb people almost always post as AC?

    40. Re: Finally a board with some RAM by Anonymous Coward · · Score: 0

      Thanks for clarifying, but I don't think it's a dumb question. People here argue endlessly about such things, so it's not safe to make assumptions.

  8. This whole post is plagiarized from Engadget by pridkett · · Score: 0

    It's good that this post linked back to the original Engadget posting, but when you do something like this, you should quote it. The title and first two sentences are also word for word from the Engadget article. Not cool. Not cool at all.

    --
    My Slashdot account is old enough to drink...
    1. Re:This whole post is plagiarized from Engadget by Anonymous Coward · · Score: 0

      You have a four digit UID. Did you buy it?

      Slashdot is and always has been a news aggregator. The expectation is that good posts will be word-for-word copies of the source with only so much editing as to be intelligible in brief. Bad posts are modified.

      Where was this criticism of yours the last 30,000 times it happened since your account was made?

    2. Re:This whole post is plagiarized from Engadget by pridkett · · Score: 1

      Sometimes you just open up your news aggregator and you do a double take because you see what looks like the same article twice in a row.

      No, I didn't buy this account. I'm just old. Old enough to remember reading Slashdot when it still ran off an DEC Alpha (I think) hosted at the Color Group in Holland, MI. Old enough that I remember when "Open Source" was discussed on Slashdot as a new thing back in 1998. I'm not sure, it might've still been Chips & Dips back then, sometimes it's hard to remember. Anyway, I had a turnip on my belt, as was the style at the time...

      --
      My Slashdot account is old enough to drink...
    3. Re:This whole post is plagiarized from Engadget by Anonymous Coward · · Score: 0

      To be fair, this is his first post for 7 years. When he joined 20 years ago, this site was pretty good. Now we're reduced to aggressively disagreeing with one another.

    4. Re:This whole post is plagiarized from Engadget by Anonymous Coward · · Score: 0

      No we aren't, you, erm... twat (was that aggressive enuff?)

    5. Re:This whole post is plagiarized from Engadget by drinkypoo · · Score: 1

      Old enough that I remember when "Open Source" was discussed on Slashdot as a new thing back in 1998.

      Proof positive that Slashdot has always been behind the curve.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    6. Re: This whole post is plagiarized from Engadget by phirzcol · · Score: 0

      Same here. It was a much smaller crowd back then.

      --
      Technology will default in society to its most rudimentary level:::stupid computers for stupid users:::
    7. Re: This whole post is plagiarized from Engadget by Anonymous Coward · · Score: 0

      Was it? I think we just had fewer troll accounts and nazi spammers. I seem to recall they used to pare away those things. Not anymore.

    8. Re:This whole post is plagiarized from Engadget by Pyramid · · Score: 1

      Easy there whipper-snapper. Your elders are talking. Cool your millennial jets and bask in the discreet wisdom.

      --
      ~Any apparent grammatical or typographic errors are caused by defects in your display device.
    9. Re:This whole post is plagiarized from Engadget by drinkypoo · · Score: 1

      Easy there whipper-snapper. Your elders are talking. Cool your millennial jets and bask in the discreet wisdom.

      Don't assume that I registered as soon as I began reading. That would be foolish. Then again, that is also Slashdot as usual.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    10. Re: This whole post is plagiarized from Engadget by Anonymous Coward · · Score: 0

      Keep the truth alive stinky!!

  9. mini-computers by Anonymous Coward · · Score: 2, Funny

    mini computers can rarely handle much more than the basics

    Minicomputer.

    1. Re:mini-computers by dromgodis · · Score: 1

      "definition of a minicomputer as a machine costing less than US$25,000 (equivalent to $161,000 in 2018), with an input-output device such as a teleprinter and at least four thousand words of memory, that is capable of running programs in a higher level language"

      I would say that this nvidia thing qualifies (with some reservation about the teleprinter).

    2. Re:mini-computers by Anonymous Coward · · Score: 0

      Lol! Well played, sir. :)

  10. Smart speaker? by 110010001000 · · Score: 1

    Obviously these people have no clue what is in a so-called "smart speaker".

    1. Re:Smart speaker? by Anonymous Coward · · Score: 0

      Isn't it a Hoomaan (Ugly Bag of Mostly Water) with an I.Q. > 100?

  11. maybe im looking in the wrong spot by Anonymous Coward · · Score: 0

    but nvidias site says even AI cant find the page. and not a single retail source has a jetson nano, let along one for $99, so i just wonder where the price is coming, id buy one for $99.

  12. When will a miner app be available? by ASCIIxTended · · Score: 0

    I just want to know when a mining app for this thing will be available. I don't want to mine with it - I want to buy a hundred of them then sell them on ebay for 10x more than I paid, like I with nVidia video cards last year (well, 2x as much anyway).

    --
    I do not belong to the church of the lowercase 'i'
  13. Perfect for making a slaughterbot by david_bonn · · Score: 1

    https://www.youtube.com/watch?...

  14. I'll wait for the positronic version by nospam007 · · Score: 4, Funny

    This one doesn't even have the 3 laws.

    1. Re: I'll wait for the positronic version by Anonymous Coward · · Score: 0

      I'll need the version with an FPGA and then you just switch the terminals on the power source.

  15. Mr. Fusion by theCat · · Score: 4, Interesting

    Soon people will be tinkering with personal-sized AI like they started to do with Arduino a few years back, and 3D printing more recently. The trend here is obvious, but we cannot predict what tinkers will come up with once they get their hands on these things in a big way.

    AI researchers fret about the "containment problem", meaning how do you prevent an autonomous intelligence from breaking out of your lab and doing whatever it wants to, including enhancing itself exponentially. So there is talk about creating process and protocols to contain AI similar to what you might have regarding biological containment for a microbiology lab working with dangerous pathogens. But those rules aren't going to work when anyone wants to can build a reasonably powerful AI machine using off-the-shelf components, and/or using cloud-based resources.

    I don't expect this is going to work out the way we think it is.

    --
    =^..^= all your rodent are belong to us
    1. Re:Mr. Fusion by Anonymous Coward · · Score: 0

      If we managed to get a system that enhanced itself exponentially in a way that was not easily controllable we would have nothing to worry about anymore [as long as it's discovered reasonably early]

    2. Re:Mr. Fusion by drinkypoo · · Score: 2

      But those rules aren't going to work when anyone wants to can build a reasonably powerful AI machine using off-the-shelf components, and/or using cloud-based resources.

      Reasonably powerful? Do you mean capable of reason? Because it's likely that such will happen in a supercomputer before it happens in some hobbyist's garage.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    3. Re:Mr. Fusion by Anonymous Coward · · Score: 0

      AI researchers fret about the "containment problem"

      The kind of AI that can run on this kind of hardware is not the kind of AI that cause researchers to fret. The definition of AI has been ... expanded in recent years.

    4. Re:Mr. Fusion by Anonymous Coward · · Score: 0

      But those rules aren't going to work when anyone wants to can build a reasonably powerful AI machine using off-the-shelf components, and/or using cloud-based resources.

      Reasonably powerful? Do you mean capable of reason? Because it's likely that such will happen in a supercomputer before it happens in some hobbyist's garage.

      Something can be powerful, but not capable of reason. Toss in a human or two and you fix that, or not, and deal with a super powerful thing that has no reason.

      It's probably a lot like asking AT&T for DSL if you are just out of range..

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

      Haha, seriously? Even if these is such a law, process and / or protocol, there is always these outlaws. Guess who they will be serving? Skynet of course!

    6. Re:Mr. Fusion by Graymalkin · · Score: 1

      This little thing is pushing 500Gflops, a mid-range GPU has several times the power. So is some rampant "AI" was going to go all DIY SkyNet it would have happened a few years ago.

      --
      I'm a loner Dottie, a Rebel.
    7. Re:Mr. Fusion by Kjella · · Score: 4, Insightful

      Reasonably powerful? Do you mean capable of reason? Because it's likely that such will happen in a supercomputer before it happens in some hobbyist's garage.

      Well maybe, maybe not. I mean we didn't start out as humans, before that we were monkeys etc. all the way back to single celled organisms. It might be that we're trying too hard to replicate intelligent behavior rather than find the underlying principles of intelligence. For example take AlphaZero, it was based on AlphaGo Zero which played Go. But rather than being a dedicated engine with a lot of specific programming for Go it just as easily beat the best at chess and shogi.

      Uber(!?) managed to pull off some impressive results with Go-Explore, a family of so-called quality diversity AI models. Open AI Five is constantly improving in DOTA, even though they're not playing the full game yet it's an open map with team play. Tencent is showing off some good Starcraft II play, even if they can't beat the best. DeepMind made some ass kicking Quake III Arena capture the flag bots. But is there some kind of supreme overlay that could evolve into all these AIs? I would say probably yes.

      That's really the holy grail of AI, what is like the "spark" of intelligence that learns to learn. It's probably relatively simple once we can look at it in retrospect, it's not a lot of code to do one specific task. It's some kind of general pattern to create more complex, dedicated "sub-AIs" for specific tasks like we have different brain centers in the brain. That and a few billion years of evolution, but computers can get pull that off pretty quick if we can just figure out where we're supposed to start. It'll probably start by beating us at tic-tac-toe, not any of the above.

      --
      Live today, because you never know what tomorrow brings
    8. Re:Mr. Fusion by del_diablo · · Score: 2

      I don't really agree.
      At the moment all these 'AI' are not AI. They are computer code running a comparison between a set of data and whatever is the realworld application of that data.
      So in the field, 'large advances' are when you figure out how to feed it a new type of data, and use it for a new type of task. Now Google has been doing this to satelite data for a few decades, to create realworld map data for consumption. But a newer development is that somebody figured out how to feed a comparison program with different resolutions of images, and then cobbled together image editing tools and upscaling algorythms so it could make A become B, and then apply that process.
      Which means the process involves a lot of bruteforce computing and comparisons.
      This recent example is waifu2x, which currently has a github page. And its been forked a few times, since waifu2x has a database and upscaling set intended for manga pages in different sizes. Now, it has been forked since then, into things like ERSGAN or AI Gigapixel, which do the same thing, but with slight alterations to the base image training set, and alterations to the tools uses to achieve the process

      But this is still basically a CRC machine, except the sensors can tell what size its working on.
      Interesting? But not really a AI per say, because so far its not even been capable of processing information into new information. Its basically looking at apples and oranges, and then using the datset of difference to give meaningful data if you input a durian but not b banana.

    9. Re:Mr. Fusion by sad_ · · Score: 1

      isn't all (so called) AI already 'reasonably powerful' enough to cause mayhem if used in a bad way?

      --
      On a long enough timeline, the survival rate for everyone drops to zero.
    10. Re: Mr. Fusion by Anonymous Coward · · Score: 0

      how do you prevent an autonomous intelligence from breaking out of your lab and doing whatever it wants to, including enhancing itself exponentially.

      Don't worry. Comcast and your cellular provider have established strict protocols that will not allow Skynet to occur, thanks to their lockdown on prepaid customer's data plans. Felonious excess data charges will shutdown regular credit card billed users in a month too.

    11. Re:Mr. Fusion by Anonymous Coward · · Score: 0

      Guys stop the futurism. Best in class A.I. circa 2019 ( Humans ) take 18 years to train and need long vacations.
      Everything has a tradeoff. Being able to learn how to learn is not all that powerful. We havent learned how to consume the sun like plants have been doing for millions of years. Priorities.

  16. Question: by Anonymous Coward · · Score: 0

    Is this Maxwell v1 or v2? And in either case, does it used signed GPU firmware, or is it like the Kepler/Maxwell V1 and can run unsigned firmware controlled by the user?

    If it can, that makes this my new go-to board. The nouveau GPU drivers for Kepler/Maxwell are far and away better than the alternatives on ARM, and while 500GFLOP is slightly less than my GT730, it is more than enough to put this ARM board at the top of its stack. Furthermore the ARM A53 cores are in-order and spectre immune, making this a perfect board from a security standpoint, if a little slower than alternatives.

    1. Re:Question: by Anonymous Coward · · Score: 0

      "making this a perfect board from a security standpoint" - LOOOOOOL what a MORON go ahead and brick your card if you need to, lol.

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

      To me it's very probably a Tegra X1, cut down such that half the 256 "CUDA" processing units are disabled.
      This is the chip in Shield TV and Nintendo Switch so you should find about the situation about GPU and firmware this way. I think nvidia contributed some things about GPU on Tegra to help it run but that may be just about not being stuck with a black screen.
      I don't know further so investigate about e.g. Shield TV.

      nouveau.freedesktop.org/wiki/CodeNames/ says Tegra X1 is NV12B and so it's Maxwell v2, with at least the additional feature of supporting FP16 computing. (which then found its way in big Pascal GP100)

      Er, the CPU is not A53, it's quad A57 hence fairly powerful but less than A72 or A73 ; out-of-order.
      Maybe it's Spectre vulnerable but with mitigation patches.

  17. Progress by JBMcB · · Score: 4, Funny

    30 Years Ago:

    Motorola's new 16-bit microcontroller has a whopping 16K of RAM, and 8 GPIO! You can use it to build robots, or home automation systems, or low-end general purpose computers...

    Today:

    Our new quad-core CPU, 128-core GPU is great if you want to build a speaker!

    --
    My Other Computer Is A Data General Nova III.
    1. Re:Progress by AmiMoJo · · Score: 1

      To be fair, even today a 16 bit microcontroller with 16k RAM is on the large side. In the sub 32 bit market there are a lot of sub 4k parts with 8 or fewer GPIOs...

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    2. Re:Progress by thegarbz · · Score: 1

      To be fair today's speakers are expected to do real time room correction and beam steering while playing music standalone from an internet stream over your wifi connection.

  18. Yodleboy loves being cheated like bitches do. by Anonymous Coward · · Score: 0

    "There is no GPU conspiracy from Nvidia..." Lol, you couldn't be dumber. https://news.bitstarz.com/nvidia-facing-barrage-of-lawsuits-for-alleged-securities-fraud
    https://www.digitaltrends.com/computing/nvidia-rtx-graphics-cards-ai/
    https://www.sfgate.com/g00/business/article/Nvidia-analyst-pleads-guilty-to-fraud-conspiracy-2370288.php
    https://www.extremetech.com/gaming/278454-nvidia-rtx-2080-and-rtx-2080-ti-review-you-cant-polish-a-turing
    https://venturebeat.com/2018/09/19/nvidia-rtx-2080-ti-review/

  19. A great media player... by Anonymous Coward · · Score: 0

    ...or a secondary computer. Unlike the Raspberry Pi, this apparently can support 4k60 output as well as dual-display. It would be handy for a lot of things.

  20. GOOGLE IS YOUR FRIEND by Anonymous Coward · · Score: 0

    https://www.electronicshub.org/breadboard-kits-beginners/

  21. Beowulf cluster by edi_guy · · Score: 2

    How can I make a Beowulf cluster out of these?

    1. Re:Beowulf cluster by Pyramid · · Score: 1

      ...by pouring hot grits on Natalie Portman.

      --
      ~Any apparent grammatical or typographic errors are caused by defects in your display device.
    2. Re:Beowulf cluster by Anonymous Coward · · Score: 0

      NAKED AND PETRIFIED!

  22. This is just awesome by WaffleMonster · · Score: 2

    This is cool shit for what it is.

    What else has the ability to encode 4k h.265 in realtime /w comparable GPU at a price anywhere near what this thing costs?

    Having said that I do a lot of h.265 encoding and wouldn't touch the NVidia GPU encoders with a 39 and a half foot pole. They suck ass.

    Still at $100 the deal breaker will be what kernel and hardware support look like for this thing.

    Personally also looking forward to the N2.
    https://www.hardkernel.com/blo...

    1. Re:This is just awesome by Anonymous Coward · · Score: 0

      Having said that I do a lot of h.265 encoding and wouldn't touch the NVidia GPU encoders with a 39 and a half foot pole. They suck ass.

      Suck in quality, or performance, or both?

  23. Now that the bottom has fallen out ... by Anonymous Coward · · Score: 0

    of the cryptocurrency mining business, NVIDIA shifts to Statistical Modelling^W^W Algorithmic Inference in order to avoid bankruptcy.

    This will give them a bit of time. I wonder what the next "big thing" will be once the bottom falls out of the Statistical Modelling^W^W Algorithmic Inference market?

  24. Kodi / Plex? by Anonymous Coward · · Score: 0

    The heck with AI. Will Kodi and Plex be ported to it?

  25. I'd test out the encode rate... by Anonymous Coward · · Score: 0

    and quality before promising anything. Just say'n.

    1. Re:I'd test out the encode rate... by Anonymous Coward · · Score: 0

      Encoding on GPU is shit except it works for real-time i.e. live streaming. It's even been optimized for low latency (thin clients and "cloud gaming"). That's what it's best used for. So let's imagine you have a small robot with several cameras and want to remote view them over wifi, then it should be useful for this at least.

  26. battery life by Anonymous Coward · · Score: 0

    look at that heat sink. what do you expect for battery life?

  27. 5w? With that heatsink? by shess · · Score: 1

    An rpi 3b+ draws a max of 5w, and your heatsink is pretty optional. I suspect this is drawing a bit more than 5w.

    And the real news is when a normal person can stably run stuff on it. I've been watching this market for a few years, and the difference between the Raspberry Pi and the Raspberry-Pi-Killers is that with an rpi you can be up an running in under a half hour depending on how fast your Internet can download updates, and the result is pokey but stable (if you don't stint on the power supply). With most of the other boards, in a half hour you haven't even found the right page for burning your OS to an SD card, and once you've finally gotten to a login screen, you get to spend the next week trying to figure out why it crashes every time you visit YouTube or CNN or probe a GPIO pin. The next six months is spent listening to how the next software update will fix your problem, and then they make a hardware rev and throw out all their existing progress.

  28. $99 is bullshit by Anonymous Coward · · Score: 0

    pure shit of bull lies marketing slashad crap. Go try buy one...

  29. Re:5w? With that heatsink? by Anonymous Coward · · Score: 0

    Maybe it has a high operating temperature range, it's also claimed that 5W is for the chip not the board.

    They have this
    https://developer.nvidia.com/embedded/dlc/jetson-nano-dev-kit-user-guide
    https://developer.nvidia.com/embedded/linux-tegra

    and as a PC vendor, formerly PC chipsets they should now to get this working.
    Now, Intel is a much bigger vendor and they did a horrible to non existent job at supporting Intel Quark boards, Edison etc. Like, for company politics reasons they had to get this product out the door just so they can pretend Intel x86 embedded boards exist. IIRC they've now pulled the plug on it. They use the Quark processor in chipsets for the Intel ME engine I think.
    So, I can understand you. A big and reputable vendor can leave you in the cold!
    But with nvidia, they have incentive for this to work. They captured most of the GPGPU market on desktop and server with their CUDA and stuff. Engineer or researcher or programmer? Buy a geforce, tesla or quadro and it just works - even or especially on linux. So they're trying to be a Microsoft of computer vision and neural networks etc. This means they want this Jetson Nano thing to be easy to work with so you get hooked up on it and maybe become a customer for other Jetson Nanos or for the bigger, high end Xavier as well as probably more of the same on the server side or the goddamn cloud.