07-04-2020, 09:45 PM
I guess I'll throw in a few cents, as a relative newb to single board and to pine64 (getting myself a pinetab).
Over the next year or two a few really excellent and powerful ARM cores should hit the market for Laptop computing, both Microsoft and Apple are really pushing ARM on Laptop, both to break away from the sinking ship of Intel and to extend battery life. The SQ1 and 8xc aren't quite powerful enough to be daily drivers in workhorse settings. (at least with the current level of code optimization windows can give it.) But that line of investment and development should push the chipsets forward in a year or two, giving powerful, high spec workhorse processors to the ARM SOC world. Also what NUVIA is doing in servers could lead to more powerful and cost effective ARM workhorse chips. There's about to be an explosion of acceptably priced high powered ARM processors. There's also stuff like the A64FX which is a very interesting AI SoC. Lots of interesting diversification in the pipeline.
As far as I can tell, (as a newb in this space,) here's nothing obvious on the market to consolidate around for long term decade length PINE64 style build tho, (as far as I can tell from looking into this, there's a lot of potential chips in the next generation, but nothing in scale production yet)
It is going to be a few years before we're likely to have a low cost, long term, high power chipset to fit the mold of A64 and RockChip.
*So, I think this kind of workstation device is a bit premature unless we want to collectively jump into the Qualcomm world.*
That said, I do want the next gen high power ARM devices that Pine will be capable of releasing in the not to distant future!
Such a chipset could be transformative of the PINE64 line. I would love a powerfully spec'd PineBox & PineTab Pro, or a PinePhone Pro as a daily driver. I'm thinking that the RockChip just isn't powerful enough to significantly change the user experience on such devices compared to the A64. (at least its not competitive with an Intel desktop from even a decade ago.)
Really examining the release plans for chips manufactures in the next 18 months should reveal the optimal chip for a high spec line. A mid/high Range chip but priced better, something meant for use in 8xc/SQ1 inspired Next Gen Windows ultraportable laptop type devices.
*
Right now there's nothing but Flagship chips that are almost loss leaders and cutting edge sales/R&D chips, they aren't going to last for a long time, production will shift and evolve. The SQ1/8xc isn't quite hefty enough to impress Reviewers, so these chips won't last long. (Afaict the Cache just isn't big enough.) What comes after these chips should be more optimized for performance and have more heft, enough to drive a 4k convertible in a similar way to a laptop i5.
Such (Equivalent to a laptop i5) ARM SoC, say, roughly a "8xc2/SQ2 7nm gen 2, (ARM Cortex-X1/A78 based SoC)" but mass market & Manufactured by a third party is at least 5 years away from a Pine device at best.
The Allwinner A64 is a 40nm chip. RockChip Rk3399 is a 28NM process. The next Gen ARM Ultrabook processors in Windows devices will be 5nm in 2022.
*
Sorry this is a bit disjointed. I'm doing research and reasoning as I write, and so the direction of this message isn't fully coherent.
*
A Server ARM SoC with 16 Cores or such could be really interesting in a PineBoard Formfactor. There are other boards built around those chips for servers. But I don't know if I see a clear usecase for such a device in a PineBoard tinkerers mold.
But, if we want a PineWorksation/PineBox we could put together some kind of PineBox with a Clusterboard and Sopines. ~400$ for 28 1.2ghz cores and 14 gigs of ram. Could be a great little machine for Video processing and Home Ai. Put a Rockchip in there to run the UI and 4k video out, could really be a cool little device.
Of all the options, I think if we want a PineBox in the next few years it would be a Clusterboard/Sopine based device.
I don't see a really compelling alternative that's not prohibitively expensive or that isn't such a left turn that it isn't likely to flop.
Over the next year or two a few really excellent and powerful ARM cores should hit the market for Laptop computing, both Microsoft and Apple are really pushing ARM on Laptop, both to break away from the sinking ship of Intel and to extend battery life. The SQ1 and 8xc aren't quite powerful enough to be daily drivers in workhorse settings. (at least with the current level of code optimization windows can give it.) But that line of investment and development should push the chipsets forward in a year or two, giving powerful, high spec workhorse processors to the ARM SOC world. Also what NUVIA is doing in servers could lead to more powerful and cost effective ARM workhorse chips. There's about to be an explosion of acceptably priced high powered ARM processors. There's also stuff like the A64FX which is a very interesting AI SoC. Lots of interesting diversification in the pipeline.
As far as I can tell, (as a newb in this space,) here's nothing obvious on the market to consolidate around for long term decade length PINE64 style build tho, (as far as I can tell from looking into this, there's a lot of potential chips in the next generation, but nothing in scale production yet)
It is going to be a few years before we're likely to have a low cost, long term, high power chipset to fit the mold of A64 and RockChip.
*So, I think this kind of workstation device is a bit premature unless we want to collectively jump into the Qualcomm world.*
That said, I do want the next gen high power ARM devices that Pine will be capable of releasing in the not to distant future!
Such a chipset could be transformative of the PINE64 line. I would love a powerfully spec'd PineBox & PineTab Pro, or a PinePhone Pro as a daily driver. I'm thinking that the RockChip just isn't powerful enough to significantly change the user experience on such devices compared to the A64. (at least its not competitive with an Intel desktop from even a decade ago.)
Really examining the release plans for chips manufactures in the next 18 months should reveal the optimal chip for a high spec line. A mid/high Range chip but priced better, something meant for use in 8xc/SQ1 inspired Next Gen Windows ultraportable laptop type devices.
*
Right now there's nothing but Flagship chips that are almost loss leaders and cutting edge sales/R&D chips, they aren't going to last for a long time, production will shift and evolve. The SQ1/8xc isn't quite hefty enough to impress Reviewers, so these chips won't last long. (Afaict the Cache just isn't big enough.) What comes after these chips should be more optimized for performance and have more heft, enough to drive a 4k convertible in a similar way to a laptop i5.
Such (Equivalent to a laptop i5) ARM SoC, say, roughly a "8xc2/SQ2 7nm gen 2, (ARM Cortex-X1/A78 based SoC)" but mass market & Manufactured by a third party is at least 5 years away from a Pine device at best.
The Allwinner A64 is a 40nm chip. RockChip Rk3399 is a 28NM process. The next Gen ARM Ultrabook processors in Windows devices will be 5nm in 2022.
*
Sorry this is a bit disjointed. I'm doing research and reasoning as I write, and so the direction of this message isn't fully coherent.
*
A Server ARM SoC with 16 Cores or such could be really interesting in a PineBoard Formfactor. There are other boards built around those chips for servers. But I don't know if I see a clear usecase for such a device in a PineBoard tinkerers mold.
But, if we want a PineWorksation/PineBox we could put together some kind of PineBox with a Clusterboard and Sopines. ~400$ for 28 1.2ghz cores and 14 gigs of ram. Could be a great little machine for Video processing and Home Ai. Put a Rockchip in there to run the UI and 4k video out, could really be a cool little device.
Of all the options, I think if we want a PineBox in the next few years it would be a Clusterboard/Sopine based device.
I don't see a really compelling alternative that's not prohibitively expensive or that isn't such a left turn that it isn't likely to flop.