Exploits for Intel CPUs (NOT AMD) Documented

don't participate in the class action.

sue independently in small claims.

Intel will not show up, and you'll get a thousand bucks for a new laptop.
 
AMD marketing will find a way to let this golden opportunity slip through it's fingers.

Most decisions based on opportunity are from Lisa Su directly, and her business acumen is better than many, her decisions have turned AMD from a billion dollar loss company to about that in profit in a year. AMD heads of the past perhaps would screw this up but Lisa has been getting AMD in the door even before this bombshell was dropped and you best believe she is not the type of person that will rest on laurels.
 
cab9430cb6a5aebe5b41a66d3cdb1c52b82d39b24a662baca26cc6e60d8812d1.jpg
 
Would be a bad idea to offline all of your critical data until all major services have this sewn up.
 
Spectre 1 applies to older amd cpu's not ryzen/TR/epyc from my understanding.

possible if they used the out of order kernel syscall method, not like many people use old AMD anymore, most FX users are likely using Ryzen now and FX actually seems fine, Athlon and Phenom meh I think people actually using them are near non existent, its to old and slow by modern standards.
 
This isn't going to effect gaming and desktop applications as much as it'll effect datacenter operations, such as database and large data set workload applications.

We're already in talks with Cisco about upgrading our blades, but they don't currently offer AMD based solutions. That's going to change very quickly. ;)
 
This isn't going to effect gaming and desktop applications as much as it'll effect datacenter operations, such as database and large data set workload applications.

We're already in talks with Cisco about upgrading our blades, but they don't currently offer AMD based solutions. That's going to change very quickly. ;)

It is a fantastic opportunity for AMD to earn some new customers and hopefully impress, pretty much what Lisa Su wants, product + satisfied client, this opens up new opportunities and eco systems for AMD down the line.
 
AMD better take this as an opportunity to step in and offer deals on enterprise data center solutions...the result could be Epyc for their bottom line.
 
AMD better take this as an opportunity to step in and offer deals on enterprise data center solutions...the result could be Epyc for their bottom line.

It would be like christmas in january/february, that and pinnacle ridge coming could make it a merry good start.
 
It is a nice system, I would probably not use those graphics cards but perhaps you have a reason for it and I am cool with that.

I do lots of adobe stuff at my job, and I crunch video for my students video projects. The Vega FE is a really nice card for that, I got the second one for well, just because....I've always had a soft spot in my heart for AMD. I've owned many Intel/Nvidia systems during my days though and will switch back when my wallet dictates. The 1950X was too good of a deal to pass up at the time.
 
I do lots of adobe stuff at my job, and I crunch video for my students video projects. The Vega FE is a really nice card for that, I got the second one for well, just because....I've always had a soft spot in my heart for AMD. I've owned many Intel/Nvidia systems during my days though and will switch back when my wallet dictates. The 1950X was too good of a deal to pass up at the time.

that seems like a perfectly understandable reason, serves a purpose. Like my CPU is not really conducive to the audio production software I have, TR would be an absolute breeze in this type of hobby.
 
https://spectreattack.com/spectre.pdf said:
1.3 Targeted Hardware and Current Status
Hardware.
We have empirically verified the vulnerability of several Intel processors to Spectre attacks, including Ivy Bridge, Haswell and Skylake based processors. We have also verified the attack’s applicability to AMD Ryzen CPUs. Finally, we have also successfully mounted Spectre attacks on several Samsung and Qualcomm processors (which use an ARM architecture) found in popular mobile phones.

From the Graz University of Technology team. Regarding Meltdown, they only say that it is unclear whether ARM and AMD are affected.

Amazingly, the "fix" for Meltdown was originally proposed to prevent ANOTHER security flaw on Intel processors: TSX timing attacks. So, good to know that exploit is also getting fixed?
 
Last edited:
This is still a little fuzzy to me. So what does the exploit do? does it dump random snippets from memory, the entire memory, or can it gain direct access to a specific targeted app's memory locations?
 
This is still a little fuzzy to me. So what does the exploit do? does it dump random snippets from memory, the entire memory, or can it gain direct access to a specific targeted app's memory locations?

Yes.
 
"Oh look, the hardware that runs the entire @#$% cloud has a major security flaw...but since my Crysis frame rate is unaffected, it must be no big deal."

Is this seriously a conversation we're having right now? :facepalm:

Linux users with their massive superiority ego made it such that a typical person gives no fucks about the finer nuances of computing systems.
 
ALL of the things you mentioned are enabled via these attacks.
Source on that?

Why would you want random memory fragments if you can target specific running apps? If the exploit enables targeted memory access the other two are meaningless.
 
I read that page and it doesn't mention any specifics like that. Or do you mean the actual white papers linked there? Did you read those?

Yeah, the actual white papers. I find the two-column format headache inducing but is better than the glossy overview.
 
This is still a little fuzzy to me. So what does the exploit do? does it dump random snippets from memory, the entire memory, or can it gain direct access to a specific targeted app's memory locations?

https://meltdownattack.com/meltdown.pdf said:
Meltdown allows an unprivileged process to read data mapped in the kernel address space, including the entire physical memory on Linux and OS X, and a large fraction of the physical memory on Windows. This may include physical memory of other processes, the kernel, and in case of kernel-sharing sandbox solutions (e.g., Docker, LXC) or Xen in paravirtualization mode, memory of the kernel (or hypervisor), and other co-located instances. While the performance heavily depends on the specific machine, e.g., processor speed, TLB and cache sizes, and DRAM speed, we can dump kernel and physical memory with up to 503 KB/s.

https://spectreattack.com/spectre.pdf said:
Spectre attacks involve inducing a victim to speculatively perform operations that would not occur during correct program execution and which leak the victim’s confidential information via a side channel to the adversary. This paper describes practical attacks that combine methodology from side channel attacks, fault attacks, and return-oriented programming that can read arbitrary memory from the victim’s process.

Spectre abuses branch prediction, and targets another process. Meltdown uses exceptions instead of branch prediction, and targets the kernel.
 
I read that page and it doesn't mention any specifics like that. Or do you mean the actual white papers linked there? Did you read those?

From the white paper:
... an attacker can leak the complete kernel memory if the virtual address of the kernel base is known. Since all major operating systems also map the entire physical memory into the kernel address space (cf. Section 2.2), all physical memory can also be read. Before kernel 4.12, kernel address space layout randomization (KASLR) was not active by default [30]. If KASLR is active, Meltdown can still be used to find the kernel by searching through the address space (cf. Section 5.2). An attacker can also simply de-randomize the direct-physical map by iterating through the virtual address space
 
Looks like computer engineers are going to have to learn a bit about criminal psychology, because this exploit functions exactly like social engineering. Social engineering relies on people performing predictable actions, and you just insert yourself into that predictable behavior long enough to get what you want from them.

Ive always known if you blow enough smoke up someone's ass you can get them to believe anything, didn't realize it worked the same way for computers. But that's way absolute order works, it only serves to shelter those who know how to exploit the order.
 
I like how the mufuka that be the intel ceo, keeps this bullshit quiet just long enough for him to dump his millions of bucks of shares before news leaked out and intel share value fell.

What a cnut.

Very possibly illegal

I'm not surprised at this.

Intel, as a corporate entity, has been criminal scum since the days when the Athlon was beating the Intel CPUs like rented mules, and Intel started using illegal anti-competitive business practices to suppress AMD sales by folks like Dell, etc.

That's why I have never bought an Intel CPU for any new build, nor will I in the future. I refuse to reward CRIMINALS for their actions.
 
Hope this doesnt blow up prices of Ryzen/Ryzen 2 and it's mind melded cousins.
 
Bunch of malarky to me. Disabling the particular workaround on my Win10 box and moving along without performance hits. I can accept this risk.
 
Ok maybe I'm not understanding this. But it looks to me like 1) someone has to get a malicious program running on my system; 2) has use this exploit to get a memory dumb; 3) has to somehow get that memory dump transmitted to them and 4) has to comb through to see if they got lucky to get a password (and know what that password applies to) or CC#. I'm not losing much sleep over this (but I'm not running a server for say Amazon I'm only talking about my home PC). And a program to do 1-3 seems like that's going to have a signature once known and can be virus scanned.


And for speed. This mostly is on I/O. I/O that's cached and can be predicted is that correct? So my most I/O intensive task is doing a make on a large project. But it's just reading straight from the SSD and then writing the complied code back to the SSD, so this isn't likely to be effected by the patch if I understand what's going on.

Yes, this is only a vulnerability. It's not a malware tool or package. The real issue is it needs to be fixed so it isn't exploited.

The fixes have bad side effects, or they have to replace hardware. That's much worse from a usability and business standpoint than most exploitable issues.

The headlines are because of what it takes to fix these. Most security issues are fixed by "take these pills, and all's well" these two are the equivolent to amputation, and brain transplant. The first is doable, but it's gonna have life changing effects. The second...
 
The cnn.Com article makes it sound like this is all intel, arm, and amd systems or pretty much all computers.

Didn't see a feedback link and had to leave for work.
 
The cnn.Com article makes it sound like this is all intel, arm, and amd systems or pretty much all computers.

Didn't see a feedback link and had to leave for work.
It's not AMD at all and only some arm processors which arm themselves have listed on their site.

As opposed to another chipmaker who is playing with fire.
 
Yes, this is only a vulnerability. It's not a malware tool or package.
Well, right now, but it seems like to be useful (for targeting your average home user) it will need a malware tool kit (again I'm only talking home, not some server that has tons of users of various levels). Anyway glad I'm not running DB servers, but those server farms, yikes!

Well of course there are those kids who might use it to steal their parents passwords (the ones who can't unlock their parents phones with their faces anyway) ;).
 
Bunch of malarky to me. Disabling the particular workaround on my Win10 box and moving along without performance hits. I can accept this risk.

You will regret that if your on the internet, very easy to get compromised with this particular vulnerability and likely wont know you were compromised. Kernel level access is all bad.
 
Who wants to bet Intel will have a new CPU line launched by end of month?
 
Back
Top