The officially official Devuan Forum!

You are not logged in.

#26 2020-09-13 18:34:32

tuxd3v
Member
Registered: 2019-11-14
Posts: 125  

Re: mnt reform support?

Head_on_a_Stick wrote:

PowerPC (PPC64?) is old and slow and doesn't really compare with the current 7nm x86_64 AMD processors.

Well, actually PowerPC is one of the newest arch's around, created initially in the 90's, and its a open arch, even NXP gave permission to fully open documentation to the project www.powerpc-notebook.org.
It is part of the Open Power Community..

Yes, this processor in cause, is not as powerful as 7nm x86_64 launched recently, but it is the best performance/Open/Cost option available at the moment by far..
It has all the connectivity modern hardware has, and its a lot more performant than a ARM cortex a53, by far..


Best Regards,
tux

Offline

#27 2020-09-15 07:27:13

zapper
Member
Registered: 2017-05-29
Posts: 95  

Re: mnt reform support?

Head_on_a_Stick wrote:
tuxd3v wrote:

Other project, that is not ARM based, but Instead PowerPC based is the www.powerpc-notebook.org

PowerPC (PPC64?) is old and slow and doesn't really compare with the current 7nm x86_64 AMD processors.

The RISC-V laptop from Balthazar looks interesting but all RISC-V implementations are focused on power efficiency rather than performance so the same would apply to those.

POWER9 (pp64el) and perhaps POWER10 both compete on performance, if not price, and Raptor offer fully free (as in speech) desktops: https://www.raptorcs.com/content/BK1B01/intro.html

But with a TDP in excess of 90W (!) I don't think we'll see any laptops based on that ISA...

You see, before you why I do not trust openpower for the foreseeable future.  Its too damn heavy wattage wise.

If I cannot use it on a laptop, and if its that heavy, no use for me.

Servers it might be good for, but all else... no way...

ARM and Risc-V are the future.  Well... actually arm is more now related, but Risc-V is the future period. Until someone else develops something more secure... wink

Offline

#28 2020-09-15 07:31:42

zapper
Member
Registered: 2017-05-29
Posts: 95  

Re: mnt reform support?

tuxd3v wrote:
Head_on_a_Stick wrote:

PowerPC (PPC64?) is old and slow and doesn't really compare with the current 7nm x86_64 AMD processors.

Well, actually PowerPC is one of the newest arch's around, created initially in the 90's, and its a open arch, even NXP gave permission to fully open documentation to the project www.powerpc-notebook.org.
It is part of the Open Power Community..

Yes, this processor in cause, is not as powerful as 7nm x86_64 launched recently, but it is the best performance/Open/Cost option available at the moment by far..
It has all the connectivity modern hardware has, and its a lot more performant than a ARM cortex a53, by far..

I wonder if any processors based on A76 or A77  will be liberated in maybe 5-10 years of their blobs... or perhaps under 5 years...

Why? Well no meltdown spectre vulnerabilities for them. smile

We can hope right?     

Till Risc-V comes out with something as fast as intel's p8600 or faster.

Anyways, I am pondering stuff right now.   I will have time to ponder long after this. Thanks for particpating in this thread.

I wonder if I should wait on balthazar.space'  idea of a laptop.

Offline

#29 2020-09-22 03:50:48

tuxd3v
Member
Registered: 2019-11-14
Posts: 125  

Re: mnt reform support?

zapper wrote:

I wonder if any processors based on A76 or A77  will be liberated in maybe 5-10 years of their blobs... or perhaps under 5 years...
Why? Well no meltdown spectre vulnerabilities for them. smile
We can hope right?     

Till Risc-V comes out with something as fast as intel's p8600 or faster.

Anyways, I am pondering stuff right now.   I will have time to ponder long after this. Thanks for particpating in this thread.

I wonder if I should wait on balthazar.space'  idea of a laptop.

Well the next devices to expect are like the Rockchip RK3588, 8nm, 4xA76 + 4xA55, its around 30-40% better than the rk3399, which should be ok, for a decent laptop..
But probably we will not see then tile 2021 or 2022 sadly..

This ARM processors are affected by Spectre and/or Meltdown, as they employ the technologies that makes the exploit possible, that already happen in rk3399 for example( 2xA72 )

Everything above Cortex a55, and you have lots of vulnerabilities.. sadly.


Best Regards,
tux

Offline

#30 2020-09-26 07:54:21

zapper
Member
Registered: 2017-05-29
Posts: 95  

Re: mnt reform support?

tuxd3v wrote:
zapper wrote:

I wonder if any processors based on A76 or A77  will be liberated in maybe 5-10 years of their blobs... or perhaps under 5 years...
Why? Well no meltdown spectre vulnerabilities for them. smile
We can hope right?     

Till Risc-V comes out with something as fast as intel's p8600 or faster.

Anyways, I am pondering stuff right now.   I will have time to ponder long after this. Thanks for particpating in this thread.

I wonder if I should wait on balthazar.space'  idea of a laptop.

Well the next devices to expect are like the Rockchip RK3588, 8nm, 4xA76 + 4xA55, its around 30-40% better than the rk3399, which should be ok, for a decent laptop..
But probably we will not see then tile 2021 or 2022 sadly..

This ARM processors are affected by Spectre and/or Meltdown, as they employ the technologies that makes the exploit possible, that already happen in rk3399 for example( 2xA72 )

Everything above Cortex a55, and you have lots of vulnerabilities.. sadly.

Good point, also... nvidia just aquired arm, so we'll have to see what happens because of that. 

Imx8m  might be the way to go for me then... till risc-v is useful for laptops.

Offline

#31 2020-10-18 02:13:25

zapper
Member
Registered: 2017-05-29
Posts: 95  

Re: mnt reform support?

tuxd3v wrote:
zapper wrote:

If it can handle an emulator like desmume and palemoon or two emulators and palemoon, than it would probably be good enough for me till they get a better processor.

I wonder what the most recent, fastest processor type that you can get without having any remote backdoors/blobs is with 4gbs of ram or more obviously and if possible LPDDR4, wink

Anyways thanks for the thoughts.

You welcome,
Indeed maybe mnt reform is the most open of them all( the price seems to go up when you talk of openess.. ) smile

balthazar.space
it seems they will attempt something with RISCV in mind, but they don't rule out a ARM cortex-a7x version..  smile
The project seems ambitious indeed..

Other project, that is not ARM based, but Instead PowerPC based is the www.powerpc-notebook.org
They seem each time closer to get a pcb design, for the Slimbook Eclipse laptop chassis smile
But its a laptop around 1500€( also a bit expensive.. )

Its also a 4 core cpu( SMT 2 - 8 threads ) at 2Ghz, but with some serious amount of Ram, I believe it goes till 32 or 64 GB Ram..
The problem with PowerPC(  to a less extent than RISCV ofcourse,.. because support for RISCV right now almost doesn't exist.... ), is the environment around with supported applications..

At same time Debian has a port for PowerPC 64 bits, and its growing fast..
At the moment, this is my dream of a laptop, I confess! smile

A powerful AMD eGPU( Open Source Drivers ), mechanical keyboard with RGB, tons of memory Ram( even tough its LPDDR3 ) and its dual chanel, and a cpu with 8 threads at 2Ghz, with nvme disks, everything..
But it costs around some 1500€( its a estimated price range, they have done.. )

It could be that "I loose my head" and advance for this one smile

Sorry to bug anyone with this thread again, but I wondered, i had a netbook once called the acer aspire one, it had the n450 atom intel processor in it, would the mnt reform be 2x or 4x faster then that laptop and would it be able to handle 2x or 4x more applications then  n450? keep in mind, the n450 i had only had one gigabyte of ram vs mnt reform's 4gb.

I also could only load one firefox application. would the mnt reform be able to handle 2-4x more applications.


So, it is this:

https://ark.intel.com/content/www/us/en … 6-ghz.html
with 1gb of ram in acer aspire one


vs mnt reform

with its 1mx8m 4 arm a53 cortex processors.

How much faster do you think it would be?

Or should I have started a new thread?

let me know.

Offline

Board footer