ARM fights against RISC-V ?
Kuemmel (439) 384 posts |
This reminds me exactly on how Microsoft was fighting against Linux back in time Quite a shame that now ARM needs to go along with that style…or may be it shows that RISC-V is really something to look into ;-) ? |
Jeffrey Lee (213) 6048 posts |
Wow, they’re really clutching at straws there.
|
Jon Abbott (1421) 2651 posts |
“Fragmentation risk” – LOL, they should look at their chips more closely! No two ARM’s implement the same instruction set, even ones that use the same instruction set version. |
Glen Walker (2585) 469 posts |
Nasty sneaky tactics. Boo to ARM. I have played around with RISC-V but never got any hardware in the end. Its an interesting product but certainly isn’t going to threaten ARM dominance today, tomorrow or even the day after. Definitely a shame that ARM feel the need to spout marketing bullcrap like that and even more of a shame is that there are hundreds in middle-management who will be buying into it. |
Steffen Huber (91) 1953 posts |
Interesting that the über-fragmenters from ARM (26bit vs. 32bit, F[U/E] oh no we do VFPvWhatever, SIMD in ARMv6 vs. NEON, Jazelle, Thumb/Thumb2, AArch32 vs AArch 64, Cortex-R vs. Cortex-A vs. Cortex-M…) say fragmentation is something bad… And “Security”? Really? After the Spectre and Meltdown disaster? |
Rick Murray (539) 13850 posts |
I think that’s the first time I’ve seen that meme taken up to eleven. WIN! |
Patrick M (2888) 126 posts |
I’m disappointed to see ARM engaging in FUD like this, much like Microsoft did with Linux back in the day, but I guess it’s to be expected as ARM is just another big corporation, same as the other big corporations. I hope RISC-V will be a success. What I’d really like to see in the future is a computer system that’s 100% open, down to the hardware level, with absolutely no proprietary firmware or drivers required, but also affordable like the raspberry pi is. |
Glen Walker (2585) 469 posts |
Likewise! Its why I got into RISC-V in the first place…quite a mammoth task though. I guess it would have to run some kind of Linux/BSD though? RISC OS is legally conjoined to ARM isn’t it? edit: GNU Hurd maybe…? Whatever happened to that…? |
Chris Mahoney (1684) 2165 posts |
Did the project ever even “start” or did they just talk about it for a while and not actually build anything useful? Edit: It seems that something was done, at least. |
David Boddie (1934) 222 posts |
There have always been people quietly working on the Hurd, and the pace has picked up again recently. You can run it in qemu or i386-compatible hardware. I guess that ARM are the new Intel now. I know people have a soft spot for them because they remember the origins of the company and still think of them as the underdog. It’s always disappointing to see a company or person get into a dominant position only to start abusing their power, especially if you had higher expectations of them. I’d given up on ARM in that respect a long time ago. |
James Wheeler (3283) 344 posts |
Pretty sure RISCV was immune to spectre attacks, and meltdown is Intel-only, I thought.
Hear, hear! I think if you drop the PowerVR GPU support and thereby loose OpenGL ES, the BeagleBone Black could possibly be binary blob free. If the more powerful RISCV boards come down in price, I’d probably adopt it. |
Glen Walker (2585) 469 posts |
Well if you have a big wadge of cash hiding under your sofa why not splurge on one of these: https://www.sifive.com/products/hifive-unleashed/ Or if you fancy something simpler you could try this: http://www.xsltpro.co.uk/content/files/515r3.pdf But you’ll have to build it and program it yourself! Actually I was hoping to port USB-V to it and turn it into a keyboard controller but never really got started… |
James Wheeler (3283) 344 posts |
A nice board, but too expensive for me. Might be a bit too simple for me xD |
Rick Murray (539) 13850 posts |
I think they were smart enough to realise when a lone bored geek handed them their ass on a plate. Talk all you want about the benefits of this kind of kernel versus that kind of kernel…all the idealogy vanishes when you are confronted with one that works and one…….
Remember that the ARM that we remember is not the ARM of today… in fact, who the hell owns ARM right now? Japanese? Chinese? Dubai? amanfromars1?
It’s to do with shortcuts taken when accessing protected memory through speculative execution. As such, it seems that Intel was hit hard when confronted with the specific known attacks, but fundamentally similar attacks worked on a greater range of processors because, well, there’s the same fundamental problem of how to deal with switching into privileged mode. Do you stall the speculative execution unit until the mode actually switches into a privileged mode (speed hit) or execute the code regardless and throw away the result (security hit)? It’s actually a tiny side effect of that which led to the big problem that everybody panicked over, just to demonstrate that you can have speed or you can have security.
I grew up with ARM32 so I’ll stick to that. Actually, at the age of 44, I ought to practice for being an old fart, so: I grew up with ARM32 so I’ll stick to that come hell or high water. :-) |
Jeffrey Lee (213) 6048 posts |
Ha! |
David Boddie (1934) 222 posts |
Well, if you’ve ever worked somewhere that made you wonder, “What have they done now?” when someone tells you that your employer is in the news then you’ll know how the ARM engineers felt on Monday. |
Kuemmel (439) 384 posts |
Funny…the site is down already…at least it gave same publicity to Risc-V ;-) Shit like that happens when some “clever” management dudes listen to some “clever” PR guys or as usual some “clever” externally hired consulting companies to tell you what to do and even believe them…same story within all big companies…I like to see the responsible people fired. That was so useless… First benchmarks of the Risc-V dev board show 10 time less speed than ARM…what a threat :-) |
Rick Murray (539) 13850 posts |
It’s all part of the master plan. Bring Risc-V to awareness, spread some FUD, let the media trash it (like with the benchmarks). Then as the outfit is scrambling to get a handle on what happened, buy them out, grab the IP, then sink the project. No Risc-V, no competition. That is how you do “evil”. |
Rick Murray (539) 13850 posts |
https://web.archive.org/web/20180710130206/https://riscv-basics.com/ Wayback was watching, b**ch. :-p |
Steve Pampling (1551) 8172 posts |
Actually, decades of watching with a techie cynical eye suggests the most likely outcome is a promotion for the idiot that did it. |