Hi,
sorry to interrupt - does that look like we're not getting our EOMAs this year again, if ever? How people are moving forward with their computing needs? What's the 'next best thing' to invest here?
George
On Monday, July 20, 2020, George Sokolsky sokolgeo@posteo.net wrote:
Hi,
sorry to interrupt
no feel free, george, this list is an open one.
- does that look like we're not getting our EOMAs this year again, if
ever?
it happens when it happens. i would love to have not encountered so many technical issues.
How people are moving forward with their computing needs? What's the 'next best thing' to invest here?
really good questions.
l.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Monday, July 20, 2020, George Sokolsky sokolgeo@posteo.net wrote:
How people are moving forward with their computing needs?
I'm not receiving an EOMA68, rather I'm building my own laptop from a firefly RK3399 SBC. It's more powerful then the EOMA68, but it's lack of FLOSS SW support is problematic. For example, the firmware flashing didn't flash firmware and did mess up my kernel enough to require a reboot. Grrr! I'm working on and off on the SW and HW.
The RK3399 is a real winner of a processor, being sold on a lot of SBCs for as little as $50. It's getting better and better SW support as devs reverse engineer it's internal mali GPU and associated interconnect and power controls. AMDs laptop offerings are also very temping and I'm recommending them to normal people who want to purchase a laptop. Linux support is in great shape for the 4000 series AFAIK[1]. Supposedly, AMD is coming out with an APU even more amazing before the end of the year.
Again, personally, I'll bang my head against this SBC until it does what I want! No PSP[2] for me!
What's the 'next best thing' to invest here?
Off the top of my head since the inception of the EOMA68...
Correct me mercilessly if I'm wrong, luke. @ @ V
The RISC-V folks are being unkind to luke and other devs (they will not let them participate), so that's out of the question. The Epiphany processor was looking like it's silicon form would really be a powerful CPU, but then the guys in charge said that there were not enough people who wanted one to manufacture them on silicon. MIPS was looking promising until you read it's "open source" license. Currently, the folks of the POWER arch are releasing version 9 as open-source (still hacking out the terms AFAIK).
So, if you're looking for Linux kernel rate OSHW releases you're bound to be disappointed. Otherwise I'd wait for OpenPOWER and invest in products from that endeavor -- assuming it doesn't go South too.
Other then that, you can look at OSHW crowdsourced projects and continue to invest your money to what you can use there.
Sincerely, David
[1]: https://www.phoronix.com/scan.php?page=news_item&px=Ryzen-7-4700U-Daily-... https://www.phoronix.com/scan.php?page=article&item=141-benchmarks-lapto...
[2]: Platform Security Processor. Less scary then Intel's ME (Management Engine) but still noteworthy. Talks on all three: https://youtu.be/bKH5nGLgi08 https://youtu.be/0o8Co1ekemU https://youtu.be/3CQUNd3oKBM
On Tuesday, July 21, 2020, David Niklas doark@mail.com wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Monday, July 20, 2020, George Sokolsky sokolgeo@posteo.net wrote:
How people are moving forward with their computing needs?
I'm not receiving an EOMA68, rather I'm building my own laptop from a firefly RK3399 SBC. It's more powerful then the EOMA68,
correction: EOMA68 is a protocol standard. an SBC cannot be compared to a standard.
you must mean "an EOMA68-A20 computer card".
if in the future an EOMA68-RK3399 existed, it would compare directly with and have exactly the same speed, capactity and performance as the RK3399 Firefly.
but its lack of FLOSS SW support is problematic. For example, the firmware flashing didn't flash firmware and did mess up my kernel enough to require a reboot. Grrr! I'm working on and off on the SW and HW.
The RK3399 is a real winner of a processor, being sold on a lot of SBCs for as little as $50. It's getting better and better SW support as devs reverse engineer its internal mali GPU and associated interconnect and power controls.
this is one thing that really concerns me. the RK3399 came out nearly 3 years ago yet you're saying that *only now* is support for it trickling through to mainstream? what gives?
AMDs laptop offerings are also very temping and I'm recommending them to
normal people who want to purchase a laptop. Linux support is in great shape for the 4000 series AFAIK[1]. Supposedly, AMD is coming out with an APU even more amazing before the end of the year.
they're just doing so much better than intel. you saw the news about Apple reporting more QA issues in Skylake during its development than Intel engineers?
Again, personally, I'll bang my head against this SBC until it does what I want! No PSP[2] for me!
What's the 'next best thing' to invest here?
Off the top of my head since the inception of the EOMA68...
Correct me mercilessly if I'm wrong, luke. @ @ V
:)
The RISC-V folks are being unkind to luke and other devs
anyone who will not join the Foundation, basically, is unwelcome.
(they
will not let them participate),
oh you can use the Standards: you just can't participate in their enhancement without joining rge Foundation.
which sounds reasonable until you read the fine print and find that there are secrecy clauses that are completely incompatible with transparency and openness requirements of, for example, the Charitable Foundation that is funding my work on Libre-SOC.
so that's out of the question. The Epiphany processor was looking like its silicon form would really be a powerful CPU, but then the guys in charge said that there were not enough people who wanted one to manufacture them on silicon.
sigh, bless them, they didn't properly "read" their customers' needs. my brother understande this extremely well. it's not the product itself that sells itself, it's whether the team can explain to the potential customers that they can be trusted to help them fulfil their needs.
MIPS was looking promising until you read it's "open source" license.
i did try contacting them. the open website was closed for business. oops
Currently, the folks of the POWER arch are releasing version 9 as
open-source (still hacking out the terms AFAIK).
that's done.
long story, here. a group has been working for a LONG time (like 10 years) to create the OpenPOWER Foundation. right in rhe middle of that, RISCV started up :)
the EULA was out in... january i think. i got "helloworld" running a couple of weeks ago in simulation, in the Libre-SOC core.
if you want to know more then do join one if rhe virtual coffee calls
https://openpowerfoundation.org/openpower-virtual-coffee-calls/
So, if you're looking for Linux kernel rate OSHW releases you're bound to be disappointed. Otherwise I'd wait for OpenPOWER and invest in products from that endeavor -- assuming it doesn't go South too.
it won't. IBM is not going away.
l.
On Tue, 21 Jul 2020 04:31:43 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Tuesday, July 21, 2020, David Niklas doark@mail.com wrote:
On Monday, July 20, 2020, George Sokolsky sokolgeo@posteo.net wrote:
How people are moving forward with their computing needs?
I'm not receiving an EOMA68, rather I'm building my own laptop from a firefly RK3399 SBC. It's more powerful then the EOMA68,
correction: EOMA68 is a protocol standard. an SBC cannot be compared to a standard.
you must mean "an EOMA68-A20 computer card".
Yes. My bad.
if in the future an EOMA68-RK3399 existed, it would compare directly with and have exactly the same speed, capactity and performance as the RK3399 Firefly.
but its lack of FLOSS SW support is problematic. For example, the firmware flashing didn't flash firmware and did mess up my kernel enough to require a reboot. Grrr! I'm working on and off on the SW and HW.
The RK3399 is a real winner of a processor, being sold on a lot of SBCs for as little as $50. It's getting better and better SW support as devs reverse engineer its internal mali GPU and associated interconnect and power controls.
this is one thing that really concerns me. the RK3399 came out nearly 3 years ago yet you're saying that *only now* is support for it trickling through to mainstream? what gives?
You're the one to tell us how open-HW is best and you're asking what gives? Surely, this is just another example of how much better FLOSS is!
AMDs laptop offerings are also very temping and I'm recommending them to
normal people who want to purchase a laptop. Linux support is in great shape for the 4000 series AFAIK[1]. Supposedly, AMD is coming out with an APU even more amazing before the end of the year.
they're just doing so much better than intel. you saw the news about Apple reporting more QA issues in Skylake during its development than Intel engineers?
Yes, but forgot to save the article.
"Acer Swift 3 Narrow Bezel Laptop, 14" IPS Full HD, Ryzen 7 4700U 8-Core up to 4.10 GHz (beats i7-1065G7), 8GB RAM, 256GB SSD, USB-C/DP, Wi-Fi 6, FP Reader, Backlit, Mytrix Ethernet USB Hub, Win 10" ^^^^^^^^^^^^^^^^^ It's such a bad PR situation that sellers are literally spelling it out in their advertisements.
so that's out of the question. The Epiphany processor was looking like its silicon form would really be a powerful CPU, but then the guys in charge said that there were not enough people who wanted one to manufacture them on silicon.
sigh, bless them, they didn't properly "read" their customers' needs. my brother understande this extremely well. it's not the product itself that sells itself, it's whether the team can explain to the potential customers that they can be trusted to help them fulfil their needs.
If you have a link or more to say, please continue. I thought it was a wonderful project but wanted to wait until they did the silicon vs. the Xylinx FPGA CPU. My choice, no doubt also the same as others, made the initial project appear less successful then it really was.
MIPS was looking promising until you read it's "open source" license.
i did try contacting them. the open website was closed for business. oops
lol
Currently, the folks of the POWER arch are releasing version 9 as
open-source (still hacking out the terms AFAIK).
that's done.
long story, here. a group has been working for a LONG time (like 10 years) to create the OpenPOWER Foundation. right in rhe middle of that, RISCV started up :)
the EULA was out in... january i think. i got "helloworld" running a couple of weeks ago in simulation, in the Libre-SOC core.
if you want to know more then do join one if rhe virtual coffee calls
https://openpowerfoundation.org/openpower-virtual-coffee-calls/
:cry: Another Zoom conference. Proprietary executable. Known spying company. Why luke, why do people always have to choose the worst of proprietary stuff?
https://nypost.com/2020/04/10/chinese-spies-are-trying-to-snoop-on-zoom-chat... https://medium.com/swlh/zoom-are-you-spying-me-7c07feebf85
We have linphone, ekiga, not to mention the myriad of stuff on shareware sites ( Here's a good one I know from my windowz days https://www.freewarefiles.com/ ), if it must be proprietary. We have IRC, Jabber, Argh! So much good stuff! Conferencing stuff (not all FLOSS, but still a lot there): https://www.ubuntupit.com/top-20-best-linux-video-conferencing-software/
So, if you're looking for Linux kernel rate OSHW releases you're bound to be disappointed. Otherwise I'd wait for OpenPOWER and invest in products from that endeavor -- assuming it doesn't go South too.
it won't. IBM is not going away.
l.
All 3 projects I listed as failed/failing have problems that are not technical. I'd love to be excited, but I'm feeling cautious and pensive.
Thanks for your thoughts, luke, David
On Tue, Jul 21, 2020 at 8:29 PM David Niklas doark@mail.com wrote:
this is one thing that really concerns me. the RK3399 came out nearly 3 years ago yet you're saying that *only now* is support for it trickling through to mainstream? what gives?
You're the one to tell us how open-HW is best and you're asking what gives? Surely, this is just another example of how much better FLOSS is!
it's more another example of the shocking waste of talent and resources of proprietary companies spongeing off of software libre resources, expecting them to "pick up after them".
the Libre-SOC ASIC is a hard lesson learned from that: i am making sure that *everything* is in place even before the first silicon is out the door.
if you want to know more then do join one if rhe virtual coffee calls
https://openpowerfoundation.org/openpower-virtual-coffee-calls/
:cry: Another Zoom conference. Proprietary executable. Known spying company. Why luke, why do people always have to choose the worst of proprietary stuff?
i'm not going to try to explain that one to them. it's "open" calls, there's nothing confidential discussed. and there do exist telephone numbers.
sigh.
anyway, they are really wonderful people. the conversations are fascinating, and funny, and often bugger-all to do with PowerISA :)
l.
On Tue, 21 Jul 2020 20:37:15 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Tue, Jul 21, 2020 at 8:29 PM David Niklas doark@mail.com wrote:
<snip>
if you want to know more then do join one if rhe virtual coffee calls
https://openpowerfoundation.org/openpower-virtual-coffee-calls/
:cry: Another Zoom conference. Proprietary executable. Known spying company. Why luke, why do people always have to choose the worst of proprietary stuff?
i'm not going to try to explain that one to them. it's "open" calls, there's nothing confidential discussed. and there do exist telephone numbers.
sigh.
anyway, they are really wonderful people. the conversations are fascinating, and funny, and often bugger-all to do with PowerISA :)
l.
Can I use the general email to contact them and try to convince them myself? Or would you advise I not convince them or talk to someone particular?
Thanks, David
On Tue, Jul 21, 2020 at 9:08 PM David Niklas doark@mail.com wrote:
Can I use the general email to contact them and try to convince them myself? Or would you advise I not convince them or talk to someone particular?
well... it's hugh (personally) who is running them. he's been in the open source community for... over 25 years? so he will know the story. i'd leave it. he knows already and will have good reasons.
l.
On Jul 21, 2020, at 13:29, David Niklas doark@mail.com wrote:
On Tue, 21 Jul 2020 04:31:43 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
[…]
if you want to know more then do join one if rhe virtual coffee calls
https://openpowerfoundation.org/openpower-virtual-coffee-calls/
:cry: Another Zoom conference. Proprietary executable. Known spying company. Why luke, why do people always have to choose the worst of proprietary stuff?
https://nypost.com/2020/04/10/chinese-spies-are-trying-to-snoop-on-zoom-chat... https://medium.com/swlh/zoom-are-you-spying-me-7c07feebf85
We have linphone, ekiga, not to mention the myriad of stuff on shareware sites ( Here's a good one I know from my windowz days https://www.freewarefiles.com/ ), if it must be proprietary. We have IRC, Jabber, Argh! So much good stuff! Conferencing stuff (not all FLOSS, but still a lot there): https://www.ubuntupit.com/top-20-best-linux-video-conferencing-software/
What about Jitsi Meet? https://jitsi.org/jitsi-meet/
Free Software Encrypted Anonymous Free Service
On Friday, July 24, 2020, Richard Wilbur richard.wilbur@gmail.com wrote:
What about Jitsi Meet? https://jitsi.org/jitsi-meet/
really nice as long as no participants use firefox, at which point everybody's browsers crash.
l.
On Fri, 2020-07-24 at 02:29 +0100, Luke Kenneth Casson Leighton wrote:
On Friday, July 24, 2020, Richard Wilbur richard.wilbur@gmail.com wrote:
What about Jitsi Meet? https://jitsi.org/jitsi-meet/
really nice as long as no participants use firefox, at which point everybody's browsers crash.
This seems to have been fixed sometime late May:
https://github.com/jitsi/jitsi-meet/issues/5439#issuecomment-635623174
Cheers, `~Eric
On Tuesday, 21 July 2020 03:02:14 CEST David Niklas wrote:
On Monday, July 20, 2020, George Sokolsky sokolgeo@posteo.net wrote:
How people are moving forward with their computing needs?
[...]
The RK3399 is a real winner of a processor, being sold on a lot of SBCs for as little as $50. It's getting better and better SW support as devs reverse engineer it's internal mali GPU and associated interconnect and power controls.
These Rockchip SoCs reportedly run pretty hot, don't they? That said, I noticed that the Raspberry Pi 4 has fan accessories available for it. So much for keeping up the pretense of Cambridge continuity with Acorn Computers unless the ultimate aim is to replicate products like the (unreleased) Acorn Business Computer with, apparently, a very noisy fan.
AMDs laptop offerings are also very temping and I'm recommending them to normal people who want to purchase a laptop. Linux support is in great shape for the 4000 series AFAIK[1]. Supposedly, AMD is coming out with an APU even more amazing before the end of the year.
I just upgraded my system (after fifteen years!) to one using a Ryzen 3400G APU (if they still call them that). However, Linux kernel compatibility is such that I'm running a kernel from the Debian backports repository because system reliability with these parts was a long time coming. And you have to use proprietary firmware blobs because all these companies need to have their secret sauce.
So, upcoming AMD products could be decent, but I would be wary about stability for a while after their release. Following the Linux kernel bug tracker can be informative:
Searching for "amdgpu" is probably what you want to do. For quite some time, people were having problems with the 3200G, and that made me worried about the 3400G, but it seems that even within product families there might be some parts that are better supported than others.
[...]
MIPS was looking promising until you read it's "open source" license.
Since I've experimented with MIPS things for a while, I have to say that the custodianship of MIPS has been generally disappointing.
First of all, Imagination Technologies seemed to want to either compete head- on with ARM by having a processor architecture and graphics technologies, or (when their Apple partnership collapsed) to broaden their offerings and to use MIPS as a vehicle into IoT and such. At that time, they had their academic programme to try and get people to experiment with the architecture. But RISC- V was already on the runway at that point, so it was too little too late.
Then, when ImgTec was acquired, MIPS got acquired and quickly passed on to Wave Technologies (presumably making some people some quick and easy money). But the MIPS business seems to be in maintenance mode, at least if you look at their Web assets. And, of course, all the MIPS Creator stuff just fell off the desk. Any announcements about opening the architecture might well be perceived as just making some noise and keeping existing licensees on board.
Indeed, casually following Microchip over the last couple of years or so, I saw it said that Microchip's acquisition of Atmel would mean that Microchip would probably double down on ARM and abandon new MIPS product development. Various pundits/punters claimed that Microchip wouldn't look twice at RISC-V. But then I saw this:
"A low-cost dev kit for Microchip's PolarFire SoC, a low-power FPGA integrated with a hardened quad core 64-bit RISC-V microprocessor subsystem"
https://www.crowdsupply.com/microchip/polarfire-soc-icicle-kit
That is from the Microsemi part of Microchip's business, however, meaning that it is another recent acquisition: Microchip presumably buying in new technology to remain competitive.
Paul
So basically, Ryzen is okay-ish, RISC-V and OpenPOWER are worth watching, MIPS is dying and Intel are in deep shit? Good times.
On Wed, Jul 22, 2020 at 12:36 AM Paul Boddie paul@boddie.org.uk wrote:
On Tuesday, 21 July 2020 03:02:14 CEST David Niklas wrote:
On Monday, July 20, 2020, George Sokolsky sokolgeo@posteo.net wrote:
How people are moving forward with their computing needs?
[...]
The RK3399 is a real winner of a processor, being sold on a lot of SBCs for as little as $50. It's getting better and better SW support as devs reverse engineer it's internal mali GPU and associated interconnect and power controls.
These Rockchip SoCs reportedly run pretty hot, don't they? That said, I noticed that the Raspberry Pi 4 has fan accessories available for it. So much for keeping up the pretense of Cambridge continuity with Acorn Computers unless the ultimate aim is to replicate products like the (unreleased) Acorn Business Computer with, apparently, a very noisy fan.
AMDs laptop offerings are also very temping and I'm recommending them to normal people who want to purchase a laptop. Linux support is in great shape for the 4000 series AFAIK[1]. Supposedly, AMD is coming out with an APU even more amazing before the end of the year.
I just upgraded my system (after fifteen years!) to one using a Ryzen 3400G APU (if they still call them that). However, Linux kernel compatibility is such that I'm running a kernel from the Debian backports repository because system reliability with these parts was a long time coming. And you have to use proprietary firmware blobs because all these companies need to have their secret sauce.
So, upcoming AMD products could be decent, but I would be wary about stability for a while after their release. Following the Linux kernel bug tracker can be informative:
Searching for "amdgpu" is probably what you want to do. For quite some time, people were having problems with the 3200G, and that made me worried about the 3400G, but it seems that even within product families there might be some parts that are better supported than others.
[...]
MIPS was looking promising until you read it's "open source" license.
Since I've experimented with MIPS things for a while, I have to say that the custodianship of MIPS has been generally disappointing.
First of all, Imagination Technologies seemed to want to either compete head- on with ARM by having a processor architecture and graphics technologies, or (when their Apple partnership collapsed) to broaden their offerings and to use MIPS as a vehicle into IoT and such. At that time, they had their academic programme to try and get people to experiment with the architecture. But RISC- V was already on the runway at that point, so it was too little too late.
Then, when ImgTec was acquired, MIPS got acquired and quickly passed on to Wave Technologies (presumably making some people some quick and easy money). But the MIPS business seems to be in maintenance mode, at least if you look at their Web assets. And, of course, all the MIPS Creator stuff just fell off the desk. Any announcements about opening the architecture might well be perceived as just making some noise and keeping existing licensees on board.
Indeed, casually following Microchip over the last couple of years or so, I saw it said that Microchip's acquisition of Atmel would mean that Microchip would probably double down on ARM and abandon new MIPS product development. Various pundits/punters claimed that Microchip wouldn't look twice at RISC-V. But then I saw this:
"A low-cost dev kit for Microchip's PolarFire SoC, a low-power FPGA integrated with a hardened quad core 64-bit RISC-V microprocessor subsystem"
https://www.crowdsupply.com/microchip/polarfire-soc-icicle-kit
That is from the Microsemi part of Microchip's business, however, meaning that it is another recent acquisition: Microchip presumably buying in new technology to remain competitive.
Paul
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Wed, Jul 22, 2020 at 3:14 AM Allan Mwenda allanitomwesh@gmail.com wrote:
So basically, Ryzen is okay-ish, RISC-V and OpenPOWER are worth watching, MIPS is dying and Intel are in deep shit?
embarrassingly large amounts of doo-doo, up to their eyeballs, yes.
hilariously, despite that, i now have an 8-core i9 dual hyper-threaded 5 ghz dell laptop with *64* GB of DDR4 RAM, and am really happy with it.
l.
i9's are power hungry as sin, and completely overshadowed by the new 4000 series mobile Ryzens which are both faster AND cheaper. In other news,this looks like something you would be interested in Luke. https://www.techpowerup.com/270207/globalfoundries-partners-with-synopsys-me...
On Wed, Jul 22, 2020 at 12:56 PM Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Wed, Jul 22, 2020 at 3:14 AM Allan Mwenda allanitomwesh@gmail.com wrote:
So basically, Ryzen is okay-ish, RISC-V and OpenPOWER are worth watching, MIPS is dying and Intel are in deep shit?
embarrassingly large amounts of doo-doo, up to their eyeballs, yes.
hilariously, despite that, i now have an 8-core i9 dual hyper-threaded 5 ghz dell laptop with *64* GB of DDR4 RAM, and am really happy with it.
l.
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Thursday, July 23, 2020, Allan Mwenda allanitomwesh@gmail.com wrote:
cheaper. In other news,this looks like something you would be interested in Luke. https://www.techpowerup.com/270207/globalfoundries- partners-with-synopsys-mentor-and-keysight-on-interoperable- process-design-kit
ah what they mean by that is, "open and shared between themselves once you have signed the ipen iPDK NDA".
*face-palm*...
There was another one by Google but at like 100nm, maybe that one has no caveats?
https://antmicro.com/blog/2020/06/skywater-open-source-pdk/
On Thu, Jul 23, 2020 at 11:50 AM Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Thursday, July 23, 2020, Allan Mwenda allanitomwesh@gmail.com wrote:
cheaper. In other news,this looks like something you would be interested in Luke. https://www.techpowerup.com/270207/globalfoundries- partners-with-synopsys-mentor-and-keysight-on-interoperable- process-design-kit
ah what they mean by that is, "open and shared between themselves once you have signed the ipen iPDK NDA".
*face-palm*...
--
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 _______________________________________________ arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Sunday, July 26, 2020, Allan Mwenda allanitomwesh@gmail.com wrote:
There was another one by Google but at like 100nm,
130nm.
maybe that one has no caveats?
caveat: designs must be libre / open.
(yay!)
get in there and test some RISC-V and OpenPOWER designs amirite?!
On Sun, Jul 26, 2020 at 10:00 PM Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Sunday, July 26, 2020, Allan Mwenda allanitomwesh@gmail.com wrote:
There was another one by Google but at like 100nm,
130nm.
maybe that one has no caveats?
caveat: designs must be libre / open.
(yay!)
--
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 _______________________________________________ arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Tuesday, 21 July 2020 23:36:28 CET Paul Boddie wrote:
So, upcoming AMD products could be decent, but I would be wary about stability for a while after their release. Following the Linux kernel bug tracker can be informative:
Searching for "amdgpu" is probably what you want to do. For quite some time, people were having problems with the 3200G, and that made me worried about the 3400G, but it seems that even within product families there might be some parts that are better supported than others.
Since then, my Ryzen system has worked better than my work laptop that has an Intel Core 7 CPU with integrated graphics that aren't used because there is also an Nvidia video card which occasionally does some very weird display memory "picture interference" thing.
(I thought that might be the HDMI output playing up, having had to mess with Synopsys HDMI on the Ingenic JZ4780 and seeing the many weird things that have to be set up to enable the peripheral, but is probably some dodgy interaction between the binary firmware blobs and the baroque "Linux plus GNOME plus whatever is in-between" graphics stack.)
The only weird thing I have seen with the Ryzen 3400G is this kind of message (with dmesg prefixes removed):
Corrected error, no action required. CPU:7 (17:18:1) MC3_STATUS[-|CE|MiscV|-|-|-|SyndV|-|-|-]: 0x9820000000000150 IPID: 0x000300b000000000, Syndrome: 0x000000002a000503 Decode Unit Ext. Error Code: 0, Micro-op cache tag parity error. cache level: RESV, tx: INSN, mem-tx: IRD
I have no idea whether this is really bad or not.
[...]
"A low-cost dev kit for Microchip's PolarFire SoC, a low-power FPGA integrated with a hardened quad core 64-bit RISC-V microprocessor subsystem"
https://www.crowdsupply.com/microchip/polarfire-soc-icicle-kit
That is from the Microsemi part of Microchip's business, however, meaning that it is another recent acquisition: Microchip presumably buying in new technology to remain competitive.
Since that campaign, they've started another:
"PolarBerry is a System-on-Module (SoM) SBC utilizing the Microchip PolarFire SoC, which integrates a low-power FPGA with a highly secure, four-application- core, 64-bit, Linux-capable RISC-V subsystem."
https://www.crowdsupply.com/sundance-dsp/polarberry
They go on about how it has "defense-level security". And again, it relies on Microsemi's proprietary toolchain.
Meanwhile, this one slipped below the radar given that SiFive already crowdfunded a similar kind of board successfully a while back:
"Powered by the SiFive Freedom U740 RISC-V SoC and targeted for creating RISC- V applications, the platform features 8 GB of 64-bit DDR4 memory operating at 2400 MT/s, high-speed interconnects with PCIe Gen 3 x8 operating at 7.8 GB/s, Gigabit Ethernet, and USB 3.2 Gen 1."
https://www.crowdsupply.com/sifive/hifive-unmatched
However, I think this one might appeal to Luke a bit more:
"Elbrus is a mini-ITX security-oriented motherboard for mobile/embedded usage based on the MCST Elbrus-8CB 8-core @ 1.5 GHz VLIW CPU on ELBRUS architecture."
https://www.crowdsupply.com/sra-centr8/icepeakitx-elbrus-8cb
In case OpenPOWER isn't exciting enough.
Paul
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sat, Nov 28, 2020 at 10:23 PM Paul Boddie paul@boddie.org.uk wrote:
On Tuesday, 21 July 2020 23:36:28 CET Paul Boddie wrote:
So, upcoming AMD products could be decent, but I would be wary about stability for a while after their release. Following the Linux kernel bug tracker can be informative:
Searching for "amdgpu" is probably what you want to do. For quite some time, people were having problems with the 3200G, and that made me worried about the 3400G, but it seems that even within product families there might be some parts that are better supported than others.
Since then, my Ryzen system has worked better than my work laptop that has an Intel Core 7 CPU with integrated graphics that aren't used because there is also an Nvidia video card which occasionally does some very weird display memory "picture interference" thing.
(I thought that might be the HDMI output playing up, having had to mess with Synopsys HDMI on the Ingenic JZ4780 and seeing the many weird things that have to be set up to enable the peripheral, but is probably some dodgy interaction between the binary firmware blobs and the baroque "Linux plus GNOME plus whatever is in-between" graphics stack.)
The only weird thing I have seen with the Ryzen 3400G is this kind of message (with dmesg prefixes removed):
Corrected error, no action required. CPU:7 (17:18:1) MC3_STATUS[-|CE|MiscV|-|-|-|SyndV|-|-|-]: 0x9820000000000150 IPID: 0x000300b000000000, Syndrome: 0x000000002a000503 Decode Unit Ext. Error Code: 0, Micro-op cache tag parity error. cache level: RESV, tx: INSN, mem-tx: IRD
my gaming laptop (also i7) i bought 2-3 years ago now had immediate problems which i "resolved" by cranking down the PCIe bus speed to level "2". this stopped devices disappearing off of the PCIe bus, such as "the entire USB subsystem" or "the NVMe SSD".
i suspect that the Reference Design (supplied by intel to laptop manufacturers) simply failed to provide sufficient power, and consequently would cause devices to go unstable and drop off the PCIe bus if they ran above a certain speed, draining a certain amount of power.
here were the options i ended up using in grub.cfg:
ro apparmor=0 nouveau.blacklist=1 nvme_core.default_ps_max_latency_us=0 acpi_backlight=vendor acpi_os_name="Windows 2015" acpi_osi=! noapic pcie_aspm=off mem_sleep_default=deep
there may have been more.
l.
/etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="apparmor=0 nouveau.blacklist=1 nvme_core.default_ps_max_latency_us=0 acpi_backlight=vendor acpi_os_name="Windows 2015" acpi_osi=! noapic pcie_aspm=off mem_sleep_default=deep"
On Mon, Jul 20, 2020, 2:57 AM George Sokolsky sokolgeo@posteo.net wrote:
How people are moving forward with their computing needs? What's the 'next best thing' to invest here?
I've long since held that the universal computer is the one you build yourself. It's far simpler and easier than the average person is willing to admit to themselves, and it has a wide range of benefits, from customization and the options for it, through to financial benefits.
For perspective... I have Asperger's, just strong enough so to keep me on the unemployment line. I'm weird enough, just baseline naturally, to creep out any prospective employers for positions where I might actually be of use. My Disability "check" (as we call it here in the USA -- those in Europe would likely consider it a form of pension, but here that's really something only for retired people) is 1095 USD a month.
I don't moonlight, so that 1095 USD has to cover *everything* -- rent, electricity, water/sewer, Internet service, mobile/cell phone service, food, medication, and (since I don't drive) public transportation... and I'm able to set aside 100-150 USD after that, roughly, for "disposable income", which covers the fun stuff -- computer tinkering, art supplies, decorations, etc.
I'm building my seventh computer in the cyberdeck form factor that I've come to love... and I'm planning out #8, slowly, as I do so.
#7 aka "SPACE CAADET" (after the famous MIT Lisp Machine -- or would that be "MIT Lithp Maschthine", the way Sylvester Cat would say it on Looney Tunes? -- keyboard), yes I named it, I always do that, I'm silly ;) and I'm not shouting, it's a (b)ac(k)ronym... it's going to be my first to run off battery power (unfortunately, it can't run and charge at the same time) and my first with a Celeron N4000 SoC -- prior to this, everything I've used pretty well topped out at an Atom x5-8300.
Follow along here, if you'd like (warning to those still on limited/90s-kind-of-slow connections -- it's *extremely* picture heavy!) -- https://hackaday.io/project/173780-space-caadet
I intend, if I can ever get myself together enough, to someday put together a book, which I'll also have done up as a series of online videos if I can, that teaches people how to make their own.
That's my answer :)
arm-netbook@lists.phcomp.co.uk
-
Allan Mwenda -
Christopher Havel -
David Niklas -
Eric Bavier -
George Sokolsky -
Luke Kenneth Casson Leighton -
Paul Boddie -
Richard Wilbur