Return to “Everything & Anything”

Post

Re: Design flaw in intel processors...

#16
Dinosawer wrote:
Thu Jan 04, 2018 1:01 am
Graf wrote:
Wed Jan 03, 2018 2:36 pm
I spent a good deal of money on a nice processor and now it's going to get nerfed by at least 20%? I guess it's good they found the bug before it became common knowledge.
Ehm, no. The 20-30 percent number is only for tests that were made to have a lot of the affected operations, and unless you're running virtual machines you're not going to have anything near that.
For real life applications and gaming the difference is more 0 - 4 percent...
which still sucks, but less so.

https://www.hardwareluxx.de/index.php/n ... esign.html
https://www.computerbase.de/2018-01/int ... itsluecke/
https://www.phoronix.com/scan.php?page= ... ming-Tests
Yeah, 80% of the systems we have at work run on virtual machines on intel cpus...
Work is going to be a freaking "joy" when I get back on monday...

Urggg...
°˖◝(ಠ‸ಠ)◜˖°
WebGL Spaceships and Trails
<Cuisinart8> apparently without the demon driving him around Silver has the intelligence of a botched lobotomy patient ~ Mar 04 2020
console.log(`What's all ${this} ${Date.now()}`);
Post

Re: Design flaw in intel processors...

#21
I think I will hold onto my actual CPU for as much as I can.
Since these flaws appear to be transversal to manufacturers there just is no good choice to make for the time being :-/

Of course the very next set of processors is going to come out with bells, whistles, and every special turned ON, all screaming "unexploitable!!!!!". Of course I am not going to believe it.

-fox
Post

Re: Design flaw in intel processors...

#23
Talvieno wrote:
Thu Jan 04, 2018 2:49 pm
Well, this is bad news for me, but at least the impact is minimal. :? Surprised it took so long for someone to discover it, though.
Dude, the Bash bug has a FAR bigger attack surface and was undetected for 20+ years.
That thing, in it's base form, allowed a remote attacker full rights on any system that had bash running (read everything non-windows) and was so incredibly trivial to exploit that you could do it with about a dozen lines of code to fully take over a target system.
°˖◝(ಠ‸ಠ)◜˖°
WebGL Spaceships and Trails
<Cuisinart8> apparently without the demon driving him around Silver has the intelligence of a botched lobotomy patient ~ Mar 04 2020
console.log(`What's all ${this} ${Date.now()}`);
Post

Re: Design flaw in intel processors...

#24
I've been reading up on this as well. https://thehackernews.com/2018/01/meltd ... ility.html

I've been using the Brave browser created by a team led by the guy who created JavaScript. With that background and Brave's focus on security, I'd expect that browser to provide the best defense possible from at least the Spectre exploit via JavaScript... but they haven't said anything about it yet.

So I asked them.

If they reply usefully, I'll try to note it here.

:x https://www.youtube.com/watch?v=RfiQYRn7fBg
Post

Re: Design flaw in intel processors...

#25
On the premise that I am no expert... I understand that all CPUs work -fundamentally- in the same way, even though manufacturers end up using different architectures.
(correct me where I am wrong, of course)

So, what I do *not* understand is HOW can different CPUs (built by different companies that have no desire to share their engineering secrets with one another) present the same exploits.

Can anyone shed some light? I am not finding explanations. Plenty of articles parrot the same news, but none goes into detail of what the problem is and how it came to be.

-fox
Post

Re: Design flaw in intel processors...

#26
Thank god the CEO sold half of his shares in Intel 6 weeks ago, keeping only the minimum he was required to under his contract as CEO. He might not have been able to buy a 7th house if he had waited. :roll:
Image
Challenging your assumptions is good for your health, good for your business, and good for your future. Stay skeptical but never undervalue the importance of a new and unfamiliar perspective.
Imagination Fertilizer
Beauty may not save the world, but it's the only thing that can
Post

Re: Design flaw in intel processors...

#28
fox wrote:
Fri Jan 05, 2018 11:53 am
On the premise that I am no expert... I understand that all CPUs work -fundamentally- in the same way, even though manufacturers end up using different architectures.
(correct me where I am wrong, of course)

So, what I do *not* understand is HOW can different CPUs (built by different companies that have no desire to share their engineering secrets with one another) present the same exploits.

Can anyone shed some light? I am not finding explanations. Plenty of articles parrot the same news, but none goes into detail of what the problem is and how it came to be.

-fox
The simple answer is industry standards/best practices. (in the following context)

Most of it boils down to decisions made to help increase IPC in a world where IPC increases are increasingly hard to come by.
Post

Re: Design flaw in intel processors...

#29
From what I've read, the reason modern processors from Intel, AMD and ARM all suffer from the same problem is that they all use the same optimization method of "speculative execution."

That is, they all use a similar technique of guessing what execution path some code will use (and rolling back if they guessed wrong). There's nothing evil about that; it's just a clever performance optimization technique that happens to expose a CPU's kernel to processes that shouldn't have that level of privilege. Oops.

If anything, I'm sort of surprised that the speculative execution thing wasn't patented by one of the big three chip makers, limiting its use by the other two. (Or was it?)

Online Now

Users browsing this forum: No registered users and 11 guests

cron