Russian ARM processor (maybe for Pyra? :P)


Ziz

Advanced Member
Joined
Jan 15, 2006
Messages
3,584
For English speaking people: http://wccftech.com/arm-baikal-processor-8-cores-cleansweep-amd-intel-russia/

For German speaking people: http://www.heise.de/newsticker/meldung/Russland-entwickelt-Prozessoren-fuer-staatlich-eingesetzte-Systeme-2236338.html

Short: Russia will develop and produce (!) it's own ARM processor based on ARM Cortex A-57.

I think the reasons are obvious. ;)

  • CPU: Name: Baikal
  • Type: ARM
  • Architecture: Cortex A57 (64 Bit)
  • Clock Speed: 2 Ghz
  • Core Count: 8 Cores coming 2015 (16 Cores coming 2016)
  • Variants: Baikal M and Baikal M/S
How about a russian cpu for the Pyra? :p
 
Last edited by a moderator:
I doubt it's created for mobile usage, so it'll suck battery like hell.

But interesting development :D
 
yes, and a great occasion for this great video, thanks B-Zar...
 
I doubt it's created for mobile usage, so it'll suck battery like hell.

But interesting development :D
Yeah I know. And hopefully the Pyra arrives before this processor.
I think a big advantage, especially in China, is, that most of these attempts don't use Windows, but Linux. :D
 
This is pretty cool, but their reasons for doing it - paranoia that Intel and AMD CPUs enable backdoors to the NSA - are pretty dumb.
 
100% They will make a "mobile" version of it. They need to sell something after all.

But, there will be a difference between the SoCs for the Russian Government and the ones for the (global?) market.

They always do that with every single thing produced in Russia. :)
 
Russian government worried about US government spying on them so they create their own state designed and manufactured processors.  Are they planning on using these to spy on their own people, and maybe others if it becomes popular enough?  Maybe it will become Cold War II.
 
Why? There are rumors that Intels Hardware Random Number Generator may be not as random as you hope. https://www.schneier.com/blog/archives/2013/09/surreptitiously.html
Maybe I haven't thought this through, but it seems to me like putting a backdoor inside a CPU chip, even one with the level of other components integrated like a high-end Core series, isn't a very easy thing to do. At least not without having some other software, firmware, or hardware (like ethernet controller) that's similarly compromised - although if you have that you don't need a backdoor in the CPU anymore. I think the CPU would need some pretty intelligent code analysis to detect and modify something like an arbitrary ethernet driver, especially if it wants to do this without being detected, without affecting performance in a noticeable way, and without leaving big conspicuous blocks visible on the die. If someone has an idea on how it'd be plausible I'm curious to hear it.

But I think everyone, even a government that has already been exposed for lots of egregious violations of privacy, deserves to be held to a standard of innocence until proven guilty or at least held to suspicious based on some kind of evidence. I find the whole idea that both the government and Intel is presumed guilty because someone determined a way they could be sabotaging their hardware (like in the case of rdrand) to be pretty unfair, and I'm surprised that nowhere in the comments you linked does this really seem to be mentioned. Mind you, taking RNGs seriously is called for and these people know what they're talking about, I can understand wanting to only use RNGs where the implementation is fully known/trusted. But assuming CPUs have complex backdoors goes kind of beyond such a level of scrutiny. I even remember seeing people claim that Intel's CPUs implement backdoors through embedded radios that no one knows about, which is pretty crazy.
 
If they go on about how great it is for networking devices I can guarantee you the power consumption isn't attractive for mobile.

Now are we going to go through this same thing every time anyone announces a 64-bit ARM product with real SoC's first being available who knows how far into the future?
 
Now are we going to go through this same thing every time anyone announces a 64-bit ARM product with real SoC's first being available who knows how far into the future?
Yeah, I guess - I mean, the two months thing stuck, didn't it?
 
Now are we going to go through this same thing every time anyone announces a 64-bit ARM product with real SoC's first being available who knows how far into the future?
No... But Russians making a new chip is a quite interesting news for people on these board - even if the processor is not usable for Pandora or Pyra. ;)
 
Well, I wouldn't care too much about the processor itself... it's just as likely to posses some backdoor, but not from the NSA this time... just like a chineese supported processor... the interesting news is the whole ARM LINUX migration...
 
Why? There are rumors that Intels Hardware Random Number Generator may be not as random as you hope. https://www.schneier.com/blog/archives/2013/09/surreptitiously.html
Maybe I haven't thought this through, but it seems to me like putting a backdoor inside a CPU chip, even one with the level of other components integrated like a high-end Core series, isn't a very easy thing to do.
Unfortunately backdoors inside processors might be a reality. Some intel processors (starting with core vPro ones, but now in every core iX after Sandy Bridge) have a 3G interface used for administration and "anti-theft". This interface has access to the machine even when it is powered off, and it is rumored that has a backdoor. Right now you have not to worry unless you have 3G in your PC, but I would not feel comfortable with a 3G enabled laptop having one of these processors.

http://www.infowars.com/91497/

This also makes me wonder if this technology might work also with other ubicuous network interfaces (WiFi, Ethernet)...
 
while the whole 3g modem affirmation is dubious. why couldn't such a system be powered using the 5Vsb?
 
Back
Top