Is Odroid-N2 just fake?

Moderators: mdrjr, odroid

Re: Is Odroid-N2 just fake?

Unread postby rooted » Fri Oct 19, 2018 4:45 pm

I have used pull to refresh on the unread tab so many times I've worn a hole in my glass screen protector.

I wish I would have kept count.
User avatar
rooted
 
Posts: 5681
Joined: Fri Dec 19, 2014 9:12 am
Location: Gulf of Mexico, US
languages_spoken: english
ODROIDs: C1, C1+, C2
XU3 Lite, XU4
N1
VU7+
HiFi Shield 2
Smart Power (original)

Re: Is Odroid-N2 just fake?

Unread postby SadmirD » Fri Oct 19, 2018 4:51 pm

https://www.timeanddate.com/worldclock/ ... orea/seoul

4:50 PM, it's gonna be a last minute post before storming out of the office :D
SadmirD
 
Posts: 7
Joined: Mon Feb 12, 2018 8:04 pm
languages_spoken: english, bosnian, croatian, serbian

Re: Is Odroid-N2 just fake?

Unread postby nobe » Fri Oct 19, 2018 4:52 pm

it's safer using F5 on the keyboard :p
nobe
 
Posts: 117
Joined: Sun Feb 07, 2016 9:52 pm
languages_spoken: english, french
ODROIDs: Odroid-C2

Re: Is Odroid-N2 just fake?

Unread postby rooted » Fri Oct 19, 2018 4:54 pm

nobe wrote:it's safer using F5 on the keyboard :p
I wore it out first.
User avatar
rooted
 
Posts: 5681
Joined: Fri Dec 19, 2014 9:12 am
Location: Gulf of Mexico, US
languages_spoken: english
ODROIDs: C1, C1+, C2
XU3 Lite, XU4
N1
VU7+
HiFi Shield 2
Smart Power (original)

Re: Is Odroid-N2 just fake?

Unread postby tmihai20 » Fri Oct 19, 2018 4:59 pm

There are extensions that can do this on a webpage. You guys make me want to install one. Patience, people do not leave the office at 5 PM exactly.
Riddle me this, riddle me that
Who is afraid of the big, black bat?
I write (in Romanian mostly) on a blog (see my profile)
tmihai20
 
Posts: 118
Joined: Mon Nov 07, 2016 10:56 pm
Location: Romania
languages_spoken: english, french, italian, romanian
ODROIDs: XU4, Go

Re: Is Odroid-N2 just fake?

Unread postby rooted » Fri Oct 19, 2018 5:02 pm

I'm just joking, although I have refreshed way too much today.
User avatar
rooted
 
Posts: 5681
Joined: Fri Dec 19, 2014 9:12 am
Location: Gulf of Mexico, US
languages_spoken: english
ODROIDs: C1, C1+, C2
XU3 Lite, XU4
N1
VU7+
HiFi Shield 2
Smart Power (original)

Re: Is Odroid-N2 just fake?

Unread postby BadPritt » Fri Oct 19, 2018 5:18 pm

Will they make me a new favorite SBC? The C2 had that title for a long time. Since this month it`s shifted towards the NanoPi M4.
They have not produced a new sbc for a long time. Do they stil got it? What choices did they make? What hurdles did they take.
I hope to read it all here soon. Cheers.
(Please don`t make me wait untill the night)
NicoD
BadPritt
 
Posts: 62
Joined: Wed Jan 17, 2018 4:08 am
languages_spoken: english dutch french spanish german
ODROIDs: Odroid C2
Odroid XU4Q

Re: Is Odroid-N2 just fake?

Unread postby phaseshifter » Fri Oct 19, 2018 5:53 pm

i think i`ll just check it out during the week.. not sitting here watching the tide roll in and out
Build It And They Will Come...Be Bold And Mighty Forces Will Come To Your Aid..!!!
phaseshifter
 
Posts: 2970
Joined: Fri May 08, 2015 9:12 am
languages_spoken: english
ODROIDs: N-1..c1+ ..c-2..xu3 lite,xu4...u-3..
u-2...other odroid acc`s as well

Re: Is Odroid-N2 just fake?

Unread postby OverSun » Fri Oct 19, 2018 5:58 pm

My bet that the announcement is not going to blow our minds away, it's not going to be 64 core, 128 Gb RAM space-shuttle mars explorer all peripherals available at hand for 20$ board, as people expect for no reason, but a good solid well known chipset. This is how market works, guys, don't jump in with disappointment posts, Odroid proved to provide solid products, there is absolutely no indications the product is going to explore the unknown and provide something noone have.
User avatar
OverSun
 
Posts: 1386
Joined: Mon Apr 29, 2013 5:12 pm
languages_spoken: english

Re: Is Odroid-N2 just fake?

Unread postby back2future » Fri Oct 19, 2018 6:10 pm

odroid wrote:Give me several hours.


Seems he knows something and does not tell. So he's responsible to some amount for all future F5 problems here, too :lol:
A biological security routine saved another keyboard, we simply fell asleep in-between, but we were asking Cortana continiously about a new sbc board during NREM sleep phases and to K-complexes phases.
New sbc os should have automated page refresh for web browsers below the 10s rate :lol:
Last edited by back2future on Fri Oct 19, 2018 6:28 pm, edited 1 time in total.
naturally beYOnd spectrum
back2future
 
Posts: 128
Joined: Sun Jul 23, 2017 3:19 pm
languages_spoken: english

Re: Is Odroid-N2 just fake?

Unread postby odroid » Fri Oct 19, 2018 6:22 pm

User avatar
odroid
Site Admin
 
Posts: 28861
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English
ODROIDs: ODROID

Re: Is Odroid-N2 just fake?

Unread postby SadmirD » Fri Oct 19, 2018 6:25 pm

yay!
SadmirD
 
Posts: 7
Joined: Mon Feb 12, 2018 8:04 pm
languages_spoken: english, bosnian, croatian, serbian

Re: Is Odroid-N2 just fake?

Unread postby OverSun » Fri Oct 19, 2018 6:27 pm

very good, even more than I expected. good luck!
User avatar
OverSun
 
Posts: 1386
Joined: Mon Apr 29, 2013 5:12 pm
languages_spoken: english

Re: Is Odroid-N2 just fake?

Unread postby back2future » Fri Oct 19, 2018 6:30 pm

OK nice, we will read it tomorrow :lol: (probably, not really, no, now :lol:)
Now, never seen spike at www network demand?
naturally beYOnd spectrum
back2future
 
Posts: 128
Joined: Sun Jul 23, 2017 3:19 pm
languages_spoken: english

Re: Is Odroid-N2 just fake?

Unread postby rooted » Sat Oct 20, 2018 3:21 am

Turns out the N2 (announcement) was fake :) :) :)

Just kidding, can't wait to see what our next ARM board will be, in the meantime the H2 is a BIG welcome surprise.
User avatar
rooted
 
Posts: 5681
Joined: Fri Dec 19, 2014 9:12 am
Location: Gulf of Mexico, US
languages_spoken: english
ODROIDs: C1, C1+, C2
XU3 Lite, XU4
N1
VU7+
HiFi Shield 2
Smart Power (original)

Re: Is Odroid-N2 just fake?

Unread postby nobe » Wed Oct 24, 2018 7:32 pm

let's bring some confusion to those who wonder which soc will be used on odroid N2
today mediatek announced a new cortex a73 based soc : Helio P70
https://liliputing.com/2018/10/mediatek ... hones.html
nobe
 
Posts: 117
Joined: Sun Feb 07, 2016 9:52 pm
languages_spoken: english, french
ODROIDs: Odroid-C2

Re: Is Odroid-N2 just fake?

Unread postby mad_ady » Wed Oct 24, 2018 7:54 pm

How is mediatek's linux support? Compared to amlogic for example?
User avatar
mad_ady
 
Posts: 4711
Joined: Wed Jul 15, 2015 5:00 pm
Location: Bucharest, Romania
languages_spoken: english
ODROIDs: XU4, C1+, C2, N1

Re: Is Odroid-N2 just fake?

Unread postby nobe » Wed Oct 24, 2018 8:48 pm

i don't know much about mediatek to be honest
on the "pros" side, i know that they are directly involved in mainlining process for some of their socs
https://git.kernel.org/pub/scm/linux/ke ... iatek/dt64

but to be honest, i still believe amlogic s922 will be used on N2
i just mentionned this mediatek soc because it fits with the scarce information odroid told us so far (NDA + cortex a73)
nobe
 
Posts: 117
Joined: Sun Feb 07, 2016 9:52 pm
languages_spoken: english, french
ODROIDs: Odroid-C2

Re: Is Odroid-N2 just fake?

Unread postby AreaScout » Wed Oct 24, 2018 9:38 pm

mad_ady wrote:How is mediatek's linux support? Compared to amlogic for example?


I don't know a Linux box with MediaTek only cheap Android TV boxes and Helio P and X series Smartphones

https://en.wikipedia.org/wiki/MediaTek

And this year it seems a lot was done mainline if you have a look at https://github.com/torvalds/linux/commi ... m/mediatek

For the N2 only the MT6771 would came in mind if wikipedia is up-to-date
User avatar
AreaScout
 
Posts: 636
Joined: Sun Jul 07, 2013 3:05 am
languages_spoken: english, german
ODROIDs: X2, U3, XU3, C2, XU4, XU4Q, N1, Go, VU5A, Show2, CloudShell2

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Thu Oct 25, 2018 3:52 pm

AreaScout wrote:I don't know a Linux box with MediaTek


The 'famous' Banana Pi R2 and R64 are examples and as usual you find a bit of information in Armbian forum. I got the impression a few MTK engineers do a lot of this mainlining work in their spare time and they still struggle to convince managers to go the open source route. But maybe they had some success internally and things change.

For a SoC vendor doing it 'correctly' (upstreaming their driver code into mainline kernel/u-boot) looks like a lot more efforts since the mainlining process involves proper peer review and rejecting code all the time while their usual 'port and forget' coding style seems to be faster and less expensive (code quality doesn't matter any more, it's all about getting the ok from managers and a product somewhat working at around launch date)
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Thu Oct 25, 2018 4:07 pm

AreaScout wrote:For the N2 only the MT6771 would came in mind if wikipedia is up-to-date


Another requirement would be Hardkernel guys having changed their minds since now being fine with LPDDR4: "At that time, we heard a news about newer SoC will be available in September with faster CPU/GPU cores and native DDR4 support from other SoC vendor": viewtopic.php?f=149&t=31277#p226287

And all these mobile SoCs contain an expensive wireless radio as well...
Last edited by tkaiser on Thu Oct 25, 2018 6:44 pm, edited 1 time in total.
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby crashoverride » Thu Oct 25, 2018 5:11 pm

tkaiser wrote:For a SoC vendor doing it 'correctly'

This is your opinion, not a fact.

None of the ARM SoC makers (that I saw) mentioned target Linux. They target Android which means they answer to Google, not LKML. Google picks a kernel version for each Android OS version. SoC makers work within the limitations of that version. This means they may be forced to write code that will later be thrown away when Linux catches up and is not possible to mainline. HDMI2.0, AFBC, etc. did not even exist as concepts in LInux at the time of SoC releases that used them. So, SoC makers are indeed doing it "correctly". If a SoC maker had to wait for "mainline" approval, their products would be obsolete at launch.

Its Android that sells ARM SoC. Linux is barely a spec of dust on the financial balance sheets.
crashoverride
 
Posts: 4023
Joined: Tue Dec 30, 2014 8:42 pm
languages_spoken: english
ODROIDs: C1

Re: Is Odroid-N2 just fake?

Unread postby elatllat » Thu Oct 25, 2018 10:44 pm

crashoverride wrote:...kernel version...


Any decent developer is used to maintaining multiple branches so mainlining at the same time as targeting an Android LTS kernel/environment is only a little bit of extra work; not a mutually exclusive choice. The investment pays off when the company releases its next product on the next kernel and you already have a working branch. And it also pays off with customer satisfaction because no one wants to use low-quality proprietary/forked code, when they could have peer reviewed open source... one should always check the economic argument and well there might be short-term benefits in not mainlining it's clear that non-forked open source is dominating the market due to its long-term return on investment.
elatllat
 
Posts: 1066
Joined: Tue Sep 01, 2015 8:54 am
languages_spoken: english
ODROIDs: XU4, N1

Re: Is Odroid-N2 just fake?

Unread postby crashoverride » Thu Oct 25, 2018 11:57 pm

elatllat wrote:mainlining at the same time as targeting an Android LTS kernel/environment is only a little bit of extra work

This assertion can be easily tested by taking the C2's 3.14 kernel and attempting to compile its drivers into a mainline kernel.
https://github.com/hardkernel/linux/tree/odroidc2-3.14.y/drivers/amlogic

The actual "correct" way, contrary to what was claimed earlier, is for Linux to have a stable driver API like Windows or MacOS/iOS. When either of these OS's updates its kernel (within the same major revision) all the drivers continue to work and do not even require recompiling. The reasons Linux is different are political, not technical. The SoC makers have no control over this design choice, so I fail to see how they can credibly be blamed.
crashoverride
 
Posts: 4023
Joined: Tue Dec 30, 2014 8:42 pm
languages_spoken: english
ODROIDs: C1

Re: Is Odroid-N2 just fake?

Unread postby crashoverride » Fri Oct 26, 2018 12:15 am

My opinion is that recent events in the Linux development community foreshadow a major shake up. There is a developer "shortage" of those possessing the skills and talents to maintain its giant monolithic driver code base. This will likely mean an increasing amount of drivers will be orphaned with nobody to maintain them. Kernel changes that break these drivers will result in them being removed. Eventually, I expect the situation to become dire enough that there will be no other choice than to introduce some kind of driver API.

The "old days" of the Linux dictatorship are now sunsetings. New royalty and new governance is on its way in. Futures generations will laugh at the outdated and backward concept of "mainlining" a driver.

It would be fun to revisit this post in 5 years and see.
crashoverride
 
Posts: 4023
Joined: Tue Dec 30, 2014 8:42 pm
languages_spoken: english
ODROIDs: C1

Re: Is Odroid-N2 just fake?

Unread postby Vasant » Fri Oct 26, 2018 4:27 am

crashoverride wrote:The actual "correct" way, contrary to what was claimed earlier, is for Linux to have a stable driver API like Windows or MacOS/iOS.

I agree with you, but it is unlikely to happen. Also kernel maintains a very stable syscall interface so that server/console applications run without issues.
The GNU/Linux desktop user space on the other hand constantly breaks API -:( and is very fragmented. This pretty much guarantees the their usage statistics will
always be below 2%. Android one the other hand is very successful and ARM SOC vendors make sure that their SOC driver works well with whatever version of Googles Linux kernel that is being used.

Vasant
Vasant
 
Posts: 99
Joined: Mon Jan 19, 2015 3:44 pm
languages_spoken: english
ODROIDs: Odroid-c1

Re: Is Odroid-N2 just fake?

Unread postby elatllat » Fri Oct 26, 2018 5:01 am

crashoverride wrote:..stable driver API ...

LTS are 7 years, how long do you want them to last for?
For example I wish X11 (a 31 year old linux api) would die already, so I would not want improvements to be delayed by api holds for more than the current LTS length.
elatllat
 
Posts: 1066
Joined: Tue Sep 01, 2015 8:54 am
languages_spoken: english
ODROIDs: XU4, N1

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Fri Oct 26, 2018 5:39 am

elatllat wrote:LTS are 7 years


Depends. When I saw Amlogic currently fiddling around with a 4.9 kernel for their new SoCs I had a short laugh since my last update obviously was half a year ago (4.9 back then said to be EOL in January 2019). But fortunately situation has changed 4 months ago: https://web.archive.org/web/20180818095 ... eases.html vs. https://web.archive.org/web/20180825071 ... eases.html
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby crashoverride » Fri Oct 26, 2018 7:21 am

elatllat wrote:LTS are 7 years, how long do you want them to last for?

This is conflating a non-related concept. It becomes evident when applied to the original argument: "LTS are 7 years, why should anyone care about mainline?"

The intent is not for a driver to support a single LTS version of Linux as that is already the case. The intent is for a driver to work across multiple Linux releases. It would not matter if a driver is written for 4.4, it would still work unmodified on 4.20.

elatllat wrote: I would not want improvements to be delayed by api holds for more than the current LTS length.

This is neither implied nor required by a stable driver API. The API is a single abstraction that changes as required by Kernel versions. This effectively means only one driver (the API provider) changes when needed instead of all drivers. This is not a new or radical concept, it has been used successfully at global scale in Windows for decades.
crashoverride
 
Posts: 4023
Joined: Tue Dec 30, 2014 8:42 pm
languages_spoken: english
ODROIDs: C1

Re: Is Odroid-N2 just fake?

Unread postby AreaScout » Fri Oct 26, 2018 6:11 pm

tkaiser wrote:
Another requirement would be Hardkernel guys having changed their minds since now being fine with LPDDR4: "At that time, we heard a news about newer SoC will be available in September with faster CPU/GPU cores and native DDR4 support from other SoC vendor": viewtopic.php?f=149&t=31277#p226287

And all these mobile SoCs contain an expensive wireless radio as well...


Yes I know the memory socked design should be the same like on H2 they say but still it could be a sub model of that SoC type you never know.

This new mysterious N2 board starts to drive me crazy, I already have sleepless nights about that, can't wait until HK lift the curtain :D

RG
User avatar
AreaScout
 
Posts: 636
Joined: Sun Jul 07, 2013 3:05 am
languages_spoken: english, german
ODROIDs: X2, U3, XU3, C2, XU4, XU4Q, N1, Go, VU5A, Show2, CloudShell2

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Fri Oct 26, 2018 9:00 pm

AreaScout wrote:Yes I know the memory socked design should be the same like on H2 they say


Huh? Why should a TV box SoC have DIMM support? I would believe we're talking about soldered DDR4 on the upcoming N2.
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby mad_ady » Fri Oct 26, 2018 9:27 pm

I take it that a dimm is not just a hardware connector that has its data/address lines prolonged to the soc and there's more to it, right?
User avatar
mad_ady
 
Posts: 4711
Joined: Wed Jul 15, 2015 5:00 pm
Location: Bucharest, Romania
languages_spoken: english
ODROIDs: XU4, C1+, C2, N1

Re: Is Odroid-N2 just fake?

Unread postby ASword » Fri Oct 26, 2018 9:47 pm

mad_ady wrote:I take it that a dimm is not just a hardware connector that has its data/address lines prolonged to the soc and there's more to it, right?


At least for the Rockchip part it was a different memory standard.
ASword
 
Posts: 114
Joined: Fri Aug 04, 2017 12:48 pm
languages_spoken: english
ODROIDs: XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby AreaScout » Sat Oct 27, 2018 3:05 am

tkaiser wrote:
Huh? Why should a TV box SoC have DIMM support? I would believe we're talking about soldered DDR4 on the upcoming N2.


You may be right, only facts will tell and stop our general confusion :?:
User avatar
AreaScout
 
Posts: 636
Joined: Sun Jul 07, 2013 3:05 am
languages_spoken: english, german
ODROIDs: X2, U3, XU3, C2, XU4, XU4Q, N1, Go, VU5A, Show2, CloudShell2

Re: Is Odroid-N2 just fake?

Unread postby ShirleyMarquez » Sat Oct 27, 2018 4:31 am

ASword wrote:
AMD x64... I haven't been watching AMD in this space lately. Do they even have an x64 SoC that would fit this market?



AMD has the Embedded Ryzen chips, which are essentially a repurposing of the Mobile Ryzen. Embedded Ryzen has additional I/O capabilities like I2C, SPI, and multiple serial ports, which probably means that those same capabilities are lurking on the Mobile Ryzen and on the 2200G and 2400G desktop chips as they all use the same die. (That doesn't mean there is actually any way to get at those functions on the mobile and desktop chips; they might not even be brought out to pins.) Embedded Ryzen is well outside the class of SoC that would go into a hypothetical Odroid-N2. It's more in the realm of what might have gone into the recently announced Odroid-H2. But they decided to go with a Celeron instead, likely for cost reasons.

The UDOO Bolt is what the H2 might have been if Hardkernel had chosen to go with Ryzen, but it starts at $249 and it's not shipping yet. We don't yet know what the H2 will cost but it's certain to be less. On the other hand, the H2 appears to lack the low-level I/O ports that Embedded Ryzen has, which would have been useful in some applications. If you need that kind of capability with the H2 you'll probably have to pair it with a second board.
ShirleyMarquez
 
Posts: 3
Joined: Sat Oct 27, 2018 4:07 am
languages_spoken: english

Re: Is Odroid-N2 just fake?

Unread postby ASword » Sat Oct 27, 2018 12:27 pm

ShirleyMarquez wrote:It's more in the realm of what might have gone into the recently announced Odroid-H2. But they decided to go with a Celeron instead, likely for cost reasons.


Well the AMD offering is a fair bit more expensive and higher power than the Gemini Lake series. Not really the same market, although there’s an awfully fuzzy line between them.

The Gemini Lake SoC expects you’re going to connect your I/O through a bus of some kind — i2c, USB, PCIe, Ethernet. This is actually quite an efficient approach when you consider how ‘expensive’ the pins coming out of a SoC like this are. Keeping pin count down is really important, and GPIO pins offer a really minimal amount of functionality per pin (compared to the advanced protocols), so they’re the most obvious to remove. If you need that sort of pin, best to offload to another device. This also gives you more flexibility in where your ports are located — e.g. I use Ethernet and my GPIO ports are 10+ meters away from the CPU. Yes, it adds some cost but in many applications GPIO isn’t needed at all.
ASword
 
Posts: 114
Joined: Fri Aug 04, 2017 12:48 pm
languages_spoken: english
ODROIDs: XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Sun Oct 28, 2018 12:29 am

mad_ady wrote:I take it that a dimm is not just a hardware connector that has its data/address lines prolonged to the soc and there's more to it, right?


I know close to nothing wrt hardware so consider this just a guess. But I believe DIMM support means additional signals and SPD support: https://en.wikipedia.org/wiki/Serial_presence_detect

Also viewtopic.php?f=153&t=30336#p220677
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby FbS » Wed Oct 31, 2018 6:41 am

OK. So far, so good, but since the N1 came out, we had the odroid-go, then the H2.....all of which probably consumed lots of hardkernel ressources and finally delayed the N2 to the point it is too late to happen this year...

At least, if hardkernel team doesn't works on yet another project, may we have an update when the NDA will be ended, and when we can expect news of the N2, if the title of the thread hasn't (yet ?) come true ?
FbS
 
Posts: 17
Joined: Thu Feb 16, 2017 10:08 pm
languages_spoken: english

Re: Is Odroid-N2 just fake?

Unread postby wtarreau » Sun Nov 04, 2018 11:30 pm

AreaScout wrote:I don't know a Linux box with MediaTek only cheap Android TV boxes


The WiTi board and the EdgeRouter-X are both examples making use of MediaTek chips which are definitely not made for TV boxes.
wtarreau
 
Posts: 24
Joined: Thu Jan 21, 2016 1:22 am
languages_spoken: english, french
ODROIDs: C2

Re: Is Odroid-N2 just fake?

Unread postby elatllat » Mon Nov 05, 2018 7:53 pm

crashoverride wrote: ..."mainlining"...

There is also now mainline ARM SBC support for the Orange Pi Zero Plus2, Orange Pi One Plus, Pine64 LTS, Banana Pi M2+ H5, 64-bit Banana Pi M2+ H3, ASUS Tinker Board S, RockPro64, Rock960, and ROC-RK-3399-PC.

via phoronix
elatllat
 
Posts: 1066
Joined: Tue Sep 01, 2015 8:54 am
languages_spoken: english
ODROIDs: XU4, N1

Re: Is Odroid-N2 just fake?

Unread postby elatllat » Tue Nov 06, 2018 10:49 am

I notice the i.MX 8X has AES and ECC hardware, hopes for the N2...
elatllat
 
Posts: 1066
Joined: Tue Sep 01, 2015 8:54 am
languages_spoken: english
ODROIDs: XU4, N1

Re: Is Odroid-N2 just fake?

Unread postby meveric » Tue Nov 06, 2018 3:53 pm

elatllat wrote:I notice the i.MX 8X has AES and ECC hardware, hopes for the N2...

The N1 also has AES, I think AES is pretty much standard in aarch64 (at least modern version).
Donate to support my work on the ODROID GameStation Turbo Image for U2/U3 XU3/XU4 X2 X C1 as well as many other releases.
Check out the Games and Emulators section to find some of my work or check the files in my repository to find the software i build for ODROIDs.
If you want to add my repository to your image read my HOWTO integrate my repo into your image.
User avatar
meveric
 
Posts: 9313
Joined: Mon Feb 25, 2013 2:41 pm
languages_spoken: german, english
ODROIDs: X2, U2, U3, XU-Lite, XU3, XU3-Lite, C1, XU4, C2, C1+, XU4Q, HC1, N1, Go

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Tue Nov 06, 2018 4:08 pm

elatllat wrote:i.MX 8X ... hopes for the N2...


The 'affordable' but larger i.MX8 variants only contain a few A53 cores, just the expensive QuadPlus and QuadMax variants add one or two A72 cores (and would still be at or below RK3399 CPU performance then). The AES and ECC support only exists for the i.MX 8X variant (using 2 or 4 A35 at up to 1.2GHz -- this is ODROID C2 'performance' area).

The chip has been announced in 2013 and is available in products since a while. How would Hardkernel be able to talk about NDA drama here?
Last edited by tkaiser on Wed Nov 07, 2018 4:40 pm, edited 1 time in total.
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Tue Nov 06, 2018 4:25 pm

meveric wrote:
elatllat wrote:I notice the i.MX 8X has AES and ECC hardware, hopes for the N2...

The N1 also has AES, I think AES is pretty much standard in aarch64 (at least modern version).


Yeah, ARMv8 Crypto Extensions can be licensed from ARM and almost all 64-bit ARM SoCs have them today. Only known exceptions: the BroadCom SoC on RPi 3/3+ and the S905 on ODROID-C2. So good AES crypto performance is standard everywhere with 64-bit ARM except the two large SBC vendor communities.

But there's more, e.g. Marvell integrates an own '5Gbps Security Engine' in addition to ARMv8 Crypto Extensions on tiny SoCs like the one on cute EspressoBin and in this case with the i.MX 8X the mentioning of AES has a different meaning (it's about the specific requirements of the industrial/automotive market with commercial RTOS and stuff).

https://www.marvell.com/docs/embedded-p ... 016-01.pdf
https://www.nxp.com/docs/en/fact-sheet/IMX8XFAMFS.pdf
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby Tomas J. » Mon Nov 12, 2018 7:00 am

Found some leaked 922X Antutu7 scores - can somebody compare with N1?
https://www.ixbt.com/live/uploads/image ... 446c2c.png
Tomas J.
 
Posts: 21
Joined: Wed Aug 16, 2017 4:23 am
languages_spoken: english
ODROIDs: XU4

Re: Is Odroid-N2 just fake?

Unread postby ASword » Mon Nov 12, 2018 9:30 am

Tomas J. wrote:Found some leaked 922X Antutu7 scores - can somebody compare with N1?
https://www.ixbt.com/live/uploads/image ... 446c2c.png


That's pretty underwhelming. Let's hope it isn't accurate.
ASword
 
Posts: 114
Joined: Fri Aug 04, 2017 12:48 pm
languages_spoken: english
ODROIDs: XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby tkaiser » Tue Nov 13, 2018 10:21 pm

Tomas J. wrote:Found some leaked 922X Antutu7 scores


Geekbench available since 3 months already: [url]https://browser.geekbench.com/v4/cpu/search?utf8=✓&q=g12b_w400[/url]

Those 3 from 'Sep 05, 2018' are funny since someone (at Hardkernel?) figured out how to enable ARMv8 Crypto Extensions in the kernel config :)
tkaiser
 
Posts: 470
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1

Re: Is Odroid-N2 just fake?

Unread postby DarkBahamut » Thu Nov 15, 2018 9:12 am

The performance is expected. The A73 cores are from a time when ARM wasn't pushing performance. The front end shrank from A72 -> 73 to reduce die size and power so there was no performance increase at all with like for like clocks. Overall those Geekbench score are only ~10% faster ST (and slower MT) than my 4 year old phone. It will be fairly competitive against most other mainstream ARM SBC's given they all use 'old' designs.

It should run nice on passive cooling though assuming Amlogic haven't total made a mess of the implementation. That will be a plus over an XU4 at least :)
DarkBahamut
 
Posts: 320
Joined: Tue Jan 19, 2016 10:19 am
languages_spoken: english
ODROIDs: XU4, N1

Re: Is Odroid-N2 just fake?

Unread postby nobe » Thu Nov 15, 2018 3:16 pm

DarkBahamut wrote:The performance is expected. The A73 cores are from a time when ARM wasn't pushing performance. The front end shrank from A72 -> 73 to reduce die size and power so there was no performance increase at all with like for like clocks. Overall those Geekbench score are only ~10% faster ST (and slower MT) than my 4 year old phone. It will be fairly competitive against most other mainstream ARM SBC's given they all use 'old' designs.

It should run nice on passive cooling though assuming Amlogic haven't total made a mess of the implementation. That will be a plus over an XU4 at least :)


thanks for reminding me this

this anandtech article is very interesting to understand all this
https://www.anandtech.com/show/10347/ar ... unveiled/1

there are 3 differents R&D centers, which focus on different architecture family :
- a72 comes from Austin and is aimed at performance
- a73 comes from Sophia and is aimed at performance efficiency
- a53 comes from Cambridge and is aimed at low power consumption

in the 2nd page of the article, there are nice graphs to better understand the artichecture differences between a72 & a73
nobe
 
Posts: 117
Joined: Sun Feb 07, 2016 9:52 pm
languages_spoken: english, french
ODROIDs: Odroid-C2

Re: Is Odroid-N2 just fake?

Unread postby ASword » Fri Nov 16, 2018 2:52 pm

Anandtech writes a good article, as usual. A73 ought to perform better on most workloads given the memory subsystem and cache improvements, and the drop from 3-way to 2-way dispatch should only show up in microbenchmarks aimed specifically at exercising this. Practically speaking it should perform well if configured to the design maximum. Not all implementations are going to do that, however, and so just stating a SoC uses A73 isn't enough to predict its performance.
ASword
 
Posts: 114
Joined: Fri Aug 04, 2017 12:48 pm
languages_spoken: english
ODROIDs: XU4, HC1

PreviousNext

Return to General Chat

Who is online

Users browsing this forum: No registered users and 2 guests