Thread 105822901

are there any actual use cases and benefits for non-x86 computers or is it just another jewish scam that the freetards got grifted into?

Yeah.

>>105822901 (OP)
Depending on your price point and power usage, you can scale from a tiny ARM core to 64 of them on rackmount servers. A lot more of the bootware is customizable and faster executing vs dealing with BIOSes etc. x86 is easier if you're also doing ARM, the reverse isn't true because of people like your picrel, except they do call themselves developers, engineers, and VPs.

>>105822901 (OP)
yes, dumbass.

>>105823219
> scale up ARM cores
> instead of just using a real computer to begin with
not sure why anybody would do this unless it's to save money. the argument made about operating systems in the orange reddit screenshot is a valid point. software compatibility and portability is still a big problem in the ARM microcontroller world of pure unadulterated faggotry. most shit i see on ARM is code originally made for other computers just being recompiled for ARM chips. why take the performance loss? better off just using x86.

>>105824508
>unless it's to save money
Are you retarded, or have you just never built actual products? You don't "save money", you make a fuckload more of it by not throwing it away for no reason. If you need a specific amount of computing power, you buy it and nothing more. Of courseyou recompile your code. Are you literally retarded? WTF

>>105822901 (OP)
non-x86 is just an attempt to lock down computers and kill open source OS

>>105822999
Checked.
>>105822901 (OP)
Other architechtures can have different performance characteristics which is definetly worth it, especially for example embedded.

>>105822901 (OP)
>freetards
anyone that uses this term is, in fact, a kike.

>>105822901 (OP)
It's the other way around. x86 can only be feasibly made by Intel and AMD due to patent bullshit. Technically the x86 patents already expired, but they add minor new features every now and then to get a new patent and everything starts using/requiring those new features rendering potential competitors non-viable whenever the previous patent expires (because the new features still have active patents). That's why everybody wants to ditch x86 even though the alternatives are worse.

>>105824772
That's just ARM. Sadly it's only desktop/laptop x86 competitor with any real traction.

ARM is as gay as Apple and such

>>105822901 (OP)
PowerPC was better.
ended up being relegated to mainframes mostly after the home workstation (porno surfer) has been dominated by intel (and now amd when they invented amd_64)

>>105822901 (OP)
there was a brief point in time when ibm made power 9 available , you can still buy them from raptor computing systems , buttrhey are insanely expensive ,initially you could get a motherboars and cpu for under 1000$ and they were reasonably performant, the main upside is that all of the software is open source, all the way down to the microcode, you can see it all, no back doors, a system you could actually mostly trust, and more importantly patchable by the user.

>>105822901 (OP)
>are there any actual use cases and benefits for non-x86 computers or is it just another jewish scam that the freetards got grifted into?
they arent even for consumers theyre bassicaly designed for embedded electronics like controllers for machinery in a factory and shit like that

>>105822901 (OP)
My MacBook Pro runs for like 16 hours on a single charge
It’s not Johny Srouji’s fault the rest of the ARM ecosystem is a shitshow

>>105824957
PowerPC was OK but my Intel Core 2 Duo mogged my G5 tower at a fraction of the wattage

>>105825062
Arm is unbeatable in efficiency.
I remember at one point, intel has just come up with their latest “Atom” processor after years, and comparable ARM processors were 25% the power consumption.
They kind of gave up after that.

>>105822901 (OP)
>SBC
>SBC
Retards who buy Chinese crap and expect it to just work deserve their fate.
None of these issues apply to the Raspberry Pi.
You simply download Ubuntu, flash it to an SD card, and enjoy. Just like any AMD64 computer.

>>105825102
> raspberry pi, ubuntu
Any alternatives? Like with less bloat.
I heard Windows has an arm version.

Will RISC-V ever actually go anywhere?

>>105825159
you can already get risc-v microcontrollers from china
as for actual risc-v PCs, i dont think so

>>105822901 (OP)
your gpu is a non x86 computer

>>105825100
>Arm is unbeatable in efficiency.
Fuck off retard this isnt 2020 anymore

>>105825311
I mean will it ever be widely adopted?

>>105825365
Depends entirely on what this fellow decides to do, m8
It also depends on whether America overplays its hand with sanctions and decides to block ARM from giving designs to China, if they do, then it's risc-v full steam ahead baby

>>105825138
Debian Sid
arch linux arm has a severe lack of packages

>>105825336
ARMtoddlers btfo

>>105825138
>alternatives
There are many RPi models, plenty of choice for every budget.
Other ARM boards are completely unsupported, even by their own manufacturers. You're lucky to get an old unofficial Ubuntu build with no further updates.
In terms of OS, apart from Ubuntu there's Raspberry Pi OS (the official Raspberry distro), RiscOS, and a few community Linux distros with varying levels of relevance.
Speaking of relevance, there are official FreeBSD images but this meme OS has no WiFi support, which even RiscOS has.

No Windows for Raspberry, Microsoft canned the port.

>>105826581
https://github.com/torvalds/linux/tree/master/arch/arm64/boot/dts

>>105822901 (OP)
the better to kill off troonix with my dear

>>105826834
Thank you for your contribution... I guess?

>>105825102
>BBC

>>105830724
Anon...

as far as I know, people only use ARM because its power efficient. x86 is for people who dont care about power restrictions.

>>105830740
BINGLE BOARD COMBUTER :DDDDDDDD

>>105825159
Yes https://tenstorrent.com/ip/risc-v-cpu

>>105825159
>>105825365
>>105825538
chinkshit will reach levels of chinkiness previously thought impossible as gooks add unratified extensions that overlap existing isa encodings, shit out random tarballs with custom patched gcc builds that adds the intrinsics (no we wont upload the source (no we wont tell you about the hardware bug workarounds))

>>105828046
this will never happen
the EU will regulate their asses if they try to do this

Threadly reminder.

>>105831396
As if the EU powers–that–be aren't creaming their panties at such a dystopian panopticon.

why would we need more than a handful of instructions?
>assign
>add
>comparison - branch if (not) equal
>???

There's working UEFI ARM64 firmware for the PI 4 (and Pi 3).
https://github.com/pftf/RPi4

This allows installing Windows 10 on the Pi3 and FreeBSD on the Pi4, although both suffer from a lack of WiFi and video drivers.

>>105831591
Because to do anything complex you'd need a bazillion of them.
And by the way, you can do almost everything in x86 with just one: https://github.com/xoreaxeaxeax/movfuscator

>>105825159
It's a RISC ISA in the age of CISCmaxxing. I think we all know the answer to your question.

>>105822901 (OP)
you can make your own RISC-V, Open POWER chips. You can't make your own x86 chip
>but muh too hard
get good

>>105830800
This, but there will be legit riscv cores that do things right. It will certainly be interesting.

>>105822901 (OP)
1. Not having to deal with Intel
2. Not having to deal with Intel
3. Not having to deal with Intel
4. Power efficiency

>>105824933
i'm a massive freetard, get the stick out of your ass.

>>105823219
>A lot more of the bootware is customizable and faster executing vs dealing with BIOSes etc.

i think this is a point that is often overlooked. device trees, TF-A, PSCI, etc. make shipping a new board easy, which is probably why RISC-V is copying this. i wouldn't even know where to begin implementing UEFI (Tianocore?) aside from handing Intel their loicense fees.

people want UEFI so they can just flash whatever kernel and go, but have you seen the fucking monstrosity of a spec it is? there's a VM with its own bytecode running in your BIOS to tell your kernel what hardware exists on the board https://uefi.org/specs/UEFI/2.10/22_EFI_Byte_Code_Virtual_Machine.html

i think if you forced ARM vendors to ship this, they would muster up a half-broken garbage implementation, which probably wouldn't boot outside their two "supported" kernels. device trees are simpler and better. it's too bad that they tend to ship with the kernel, which means that you can't boot a generic image.

>>105833935
>muster up a half-broken garbage implementation, which probably wouldn't boot outside their two "supported" kernels
This is why the people working on Snapdragon laptops just dump the onboard ACPI tables and convert them to device trees.

>>105830800
Based
Let the best instructions survive, not the ones some bullshit committee sitting in an ivory tower decide