More Intel leaks.. this one is not good though

[ItsMitch]

Just now, straight_stewie said:

Late reply to a popular thread, but he could quickly get in trouble for insider trading on that deal. Apparently manufacturers knew about Spectre and Meltdown for months before it was released to public knowledge. 

Wouldn't surprise me at all if he did.

[AresKrieger]

2 hours ago, DoctorWho1975 said:

AMD+Nvidia is fine

Well that often performs worse by a large margin regardless so........

 

1 hour ago, Zodiark1593 said:

As I read through the paper for the Spectre vulnerability, this particular sentiment has me quite concerned.

snip

Thoughts on this?

Basically you can make something perfectly secure but it will be essentially unusable from any other metric, realistically it doesn't bother me mostly due to the fact if something manages to get past the firewall on a computer you are generally already screwed so it doesn't matter to me, everything on my computer is non-sensitive as for sensitive data on more "secure" servers companies have already have lost said data so I think this issue is minimal compared to the plethora of unknowns.

[luigi90210]

16 minutes ago, SC2Mitch said:

Wouldn't surprise me at all if he did.

i remember back in 2014 hearing about the management engine vulnerabilities back in 2014 at REcon so i'm surprised it's come to light 3 years later

 

6 minutes ago, AresKrieger said:

Well that often performs worse by a large margin regardless so........

 

well it might not perform worse after the IME security update gets pushed through, in some tests on linux the 8700k is seeing a near 70% in performance reduction so...

[AresKrieger]

30 minutes ago, johnukguy said:

For those of you  still wondering  about AMD, AMD have apparently already produced a patch that 'fixes' the issue for their CPUs and has no performance hit:

That doesn't make any sense given the issue they can't patch it a kernal patch or numerous software patches by third parties would be required to fix said issue given how this can be program to program or kernal to memory

 

And given their claim that they were unaffected by the kernal issue there is nothing they could even provide input on to fix

[AresKrieger]

Just now, luigi90210 said:

well it might not perform worse after the IME security update gets pushed through, in some tests on linux the 8700k is seeing a near 70% in performance reduction so...

No test showed newer cpus with that drop, old ones possibly but it 30% was the max and the newer chips were the least affected

[luigi90210]

2 minutes ago, AresKrieger said:

No test showed newer cpus with that drop, old ones possibly but it 30% was the max and the newer chips were the least affected

https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

literally the first test shows a near 70% reduction in I/O output with a 8700k...

 

i posted this back on page 17 and apparently someone else posted that as well before me 

[AresKrieger]

5 minutes ago, luigi90210 said:

https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

literally the first test shows a near 70% reduction in I/O output with a 8700k...

1) that has nothing to do with nvidia + intel

2) I do not know what that software or site is so I couldn't care less until its verified, 30% maximum was verfied additionally there is no reason a 6700k would lose less than a 8700k unless something else changed given how older chips are more affected than newer ones.

 

Also by their own statement gaming performance doesn't appear affected.

[ItsMitch]

After digging more it appears Intel was told a lot sooner than November / December, according to Google's report and Amazon AWS is reporting this issue could affect Intel CPUs up to 20 years old

Sources:

https://aws.amazon.com/security/security-bulletins/AWS-2018-013/

 

https://googleprojectzero.blogspot.co.uk/2018/01/reading-privileged-memory-with-side.html?m=1

[Guest]

7 minutes ago, luigi90210 said:

https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

literally the first test shows a near 70% reduction in I/O output with a 8700k...

 

i posted this back on page 17 and apparently someone else posted that as well before me 

The first one shows a 43.7% reduction in speed while the second one shows a 53.9% reduction. While that is higher than 30%, that still isn't quiet near 70%

Edited January 4, 2018 by Guest
Second number off by ~3%

[Stefan Payne]

12 hours ago, iiNNeX said:

So this happens just as I am about to order my 8700k system.... hmmm. I ll wait for some benchies after the patches and see how this goes. I want ze frames!!!

Yeah, got time to wait and see what happens and think about what you gonna get again...

 

12 hours ago, Princess Cadence said:

I would very much appreciate if you'd quit the condescending tone when addressing me every time because I am reading the situation from the mainstream end user perspective only.

You are the one who is trying to downplay this issue and doesn't seem to understand it at all - or doesn't want to understand it. 

Its some serious shit that can get you to be a Member of a Botnet. And Botnets are used to commit crimes.

So Something to think about...

 

 

12 hours ago, Nicholatian said:

I'm sorry that you feel I'm being condescending, but given there's no tone of voice in text all I can say to you is that was the last thing in my mind while writing here. If I don't think somebody understands a concept, then it is the right thing for me to do to help them understand, because it leaves everyone more informed about what's going on.

 

With that in mind, I'd be happy to continue the discussion to answer any questions you have or anything you might be confused over, if I know. I'm not a security engineer, but I am a software engineer professionally and have learned a lot about computers, even studying to the level of assembly with ARM CPUs.

Naa, you are good, I apreciate your statements.

Seems like you know what you are saying - more than most people in this thread.

 

11 hours ago, Subtle Corruption said:

What actual intel CPU gens are hit by this?

All since Pentium Pro it seems.

So that means that this Bug is around for 20 years...

 

That's really really bad...

11 hours ago, Subtle Corruption said:

And i have a gut feeling intel knew of this already but pushed and pushed for market share and dominance for this length of time up until it will bite them on their ass, amassing enough to barely be hit by the problem hoping Windows will fix it for them.

Yeah, hard to believe that nobody notices this in 20 years.

 

10 hours ago, .spider. said:

Afaik Pentium Pro and later 

Yep, that's correct...

 

10 hours ago, Taf the Ghost said:

AMD's GPU side would be more of a concern with Raja gone, in theory. CPU side, Papermaster is still there. Keller wasn't responsible for Zen's design, though he was important for getting it through the design process. (He may be more responsible for moving AMD to a modular design approach, though, but he doesn't need to be there to continue with it.)

Raja wasn't that long there and couldn't have done much.

You need years to get into that shit to really make a difference. As most developments take years.

Whatever Raja did we won't see until 2020 or something like that.

 

 

10 hours ago, Princess Cadence said:

Technically the "speed" is the very same it has always been, your CPU is just doing extra work to achieve the same result it did before in order to be "safe".

Its NOT a speed issue.

Its a Security Issue. And Nocholatian alredy told you a good example for something that's easier to grasp for you.

And that's reality that the CPU can be exploited to get access for areas you should not have access to...

 

10 hours ago, Princess Cadence said:

Intel is not getting sued.

After carriying the bug around for 20 years?!

There had to be someone who noticed this happening...

 

9 hours ago, Taf the Ghost said:

https://www.phoronix.com/forums/forum/phoronix/latest-phoronix-articles/998707-initial-benchmarks-of-the-performance-impact-resulting-from-linux-s-x86-security-changes/page4

 

This would be on Linux, so it almost assuredly won't effect Windows. However, this is going to be just one more headache invovled.

Its not sure right now, we have to wait and see what happens...

9 hours ago, Arokhantos said:

So this only effects vm based performance ? dont use that anyway.

Nope, everything that uses system calls (like acess Filesystem)

 

7 hours ago, Misanthrope said:

Well bad news all around: Linux kernell will be patched for AMD processors too, even though they don't need the patch and will suffer the performance hit as well:

https://overclock3d.net/news/cpu_mainboard/both_intel_and_amd_cpus_are_being_reported_as_insecure_on_linux/1

I honestly deeply suspect intel chicanery is at hand.

If it would, that shows what "nice persons" the Linux guys are.

Because the AMD Patch was already accepted. But possibly they got a bit of money from Intel to activate it also on AMD even if it doesn't happen on AMD much.

 

And there has to be a reason why AMD doesn't have such extensive Out of Order Excecution like Intel does.

That can lead to the conclusion that someone at AMD thought that it might be a bad idea for whatever reason...

 

5 hours ago, M.Yurizaki said:

To everyone going "I'm going AMD because it's more secure than Intel!"

 

Zero publicly known exploits does not make a system any more secure than one with X amount of publicly known exploits. AMD is just as much of a black box as Intel is. This is like saying "Windows has 10 known security flaws while Linux has 20. Therefore, Windows is more secure."

How can you say that after everything that happened this year with Intel shit?!
 

Have you already forgotten the Managment Engine Exploit?!
ANd now that?!
And you still think that Intel is the best of the best and secure?! 

You really must be a believer in Intel...

 

 

5 hours ago, Princess Cadence said:

Actually it seems now they have issues just as much, I'll be renewing the thread with the new information soon.

No it doesn't.

Read the AMD Statement and don't force your wishful thinking on others.

 

AMD Is NOT affected by this stuff. 

 

 

Even if Intel said something that they aren't alone, doesn't mean that it's the truth. 

[ItsMitch]

8 hours ago, Stefan Payne said:

fucking sniperino. 

1

Where's the preach emoji when you need it

[luigi90210]

Just now, AresKrieger said:

1) that has nothing to do with nvidia + intel

2) I do not know what that software or site is so I couldn't care less until its verified, 30% maximum was verfied additionally there is no reason a 6700k would lose less than a 8700k unless something else changed given how older chips are more affected than newer ones.

1 yes it does, a reduction in performance is a reduction in performance and until someone can test this fix on a variety of hardware with a variety of games, anything is up in the air, even at the "confirmed" 30% hit thats enough to knock down intel to where ryzen is in terms of IPC and with the way nvidia's drivers use drawcalls, nvidia and intel will probably end up taking a hit in performance

 

2 exactly wait until its verified, saying intel and nvidia will perform better before this patch has dropped publically is spreading misinformation

6 minutes ago, tjcater said:

The first one shows a 43.7% reduction in speed while the second one shows a 54% reduction. While that is higher than 30%, that still isn't quiet near 70%

well whatever mr. literal, just looking at the graph it looks like a 70% drop, either way that 30% number isnt confirmed if there are other benchmarks showing a 54% reduction in performance which is significantly higher than what was told 

[AresKrieger]

2 minutes ago, luigi90210 said:

1 yes it does, a reduction in performance is a reduction in performance and until someone can test this fix on a variety of hardware with a variety of games,

Even with your article it says games are not affected, only highly multithreaded applications, mostly synthetic applications and encoding software.

[Stefan Payne]

Just now, AresKrieger said:

Even with your article games are not affected, only highly multithreaded applications, mostly synthetic applications and encoding software.

We don't know that right now and have to wait and see how it ends up with.

Also there can be issues with streaming right now because I/O Acess is also affected and that can cause stuttering.

 

I'd rather wait and see what people are saying about this shit before making any kind of statement in this regards. Everything is possible...

[AresKrieger]

5 minutes ago, Stefan Payne said:

We don't know that right now and have to wait and see how it ends up with.

Also there can be issues with streaming right now because I/O Acess is also affected and that can cause stuttering.

 

I'd rather wait and see what people are saying about this shit before making any kind of statement in this regards. Everything is possible...

All I'm saying is that the info presented to counter my point isn't directly related to what I'm saying and actually concludes differently than what said counter point was, essentially invalidating the argument unless new credible evidence is shown.

[luigi90210]

8 minutes ago, AresKrieger said:

Even with your article games are not affected, only highly multithreaded applications, mostly synthetic applications and encoding software.

do you understand how nvidia performs draw calls? i assume you dont given your response so i suggest educating yourself because right now until anything is confirmed it looks like intel/amd is gonna perform better than intel/nvidia and amd/nvidia will probably end up out performing intel/nvidia 

also so far a single game has been tested and that showed no difference but you can't draw any conclusions based on A SINGLE GAME's result thats why i said to wait and see when the patch goes live and to have people test these games, like linus, hardware unboxed, hardwarecanucks, ect. until independent testing is done across multiple generations of intel systems with a variety of hardware configurations you can't draw any real conclusions, only speculations 

[AresKrieger]

5 minutes ago, luigi90210 said:

do you understand how nvidia performs draw calls? i assume you dont given your response so i suggest educating yourself because right now until anything is confirmed it looks like intel/amd is gonna perform better than intel/nvidia and amd/nvidia will probably end up out performing intel/nvidia 

I do you failed to read the article you posted it specifically said otherwise which makes we question everything you say since clearly you don't actually read your own evidence

The link I underlined in your article shows comparisons in performance on games pre/post patch

[luigi90210]

but they havent tested any variety of games across NVIDIA GPUs(you know if you read the source you are claiming as fact), so far a vega card was tested and because of how AMD since GCN 1.1 have had a hardware scheduler, it makes sense that there wouldnt be a change

again educate yourself on how nvidia performs draw calls then comment

[AresKrieger]

6 minutes ago, Nicholatian said:

Intel shills are just as toxic as AMD shills, because they mystify the facts of the matter like it's partisan politics all of a sudden. Knowledge about the technicals, for better or worse, is only going to help everyone her

I just got annoyed when he tried to counter something I said with an article that countered what he said with in it ,contradictions simply piss me off

 

5 minutes ago, luigi90210 said:

but they havent tested any variety of games across NVIDIA GPUs(you know if you read the source you are claiming as fact), so far a vega card was tested and because of how AMD since GCN 1.1 have had a hardware scheduler, it makes sense that there wouldnt be a change

YOUR THE PERSON WHO POSTED THE LINK AS FACT!!!!

It's like you fail to realize the info I posted came from something you used as evidence in your post, we might as well throw the source in the garbage since clearly it's unreliable (which was what I was alluding to from the beginning)

 

[DeScruff]

Ok my main question would have to be: What about Windows 7 users?
Remember how Microsoft will not be issuing updates (including security updates) for machines running Ryzen, and Kabylake CPUs?

They have to make this an exception... right?

[johnukguy]

43 minutes ago, AresKrieger said:

That doesn't make any sense given the issue they can't patch it a kernal patch or numerous software patches by third parties would be required to fix said issue given how this can be program to program or kernal to memory

 

And given their claim that they were unaffected by the kernal issue there is nothing they could even provide input on to fix

The specific issue that affects Intel CPU's and requires the level of intervention that you mention, doesn't affect AMD CPUs. Of the other two vulnerabilities, one also doesn't affect AMD CPU's and the other is fixable by a software patch. At least according to AMD. The outlying issue seems to be that of the Bypass vulnerability, which I understand only has been found to affect AMD Pro and FX CPU's. Of course that can change, so this is a moving target at the moment, at least as far as most of us are concerned.

[johnukguy]

Just now, Sypran said:

Ok my main question would have to be: What about Windows 7 users?
Remember how Microsoft will not be issuing updates (including security updates) for machines running Ryzen, and Kabylake CPUs?

They have to make this an exception... right?

Microsoft are releasing an update next Tuesday for Windows 7 users.

[Stefan Payne]

2 minutes ago, Nicholatian said:

You would think I'm acting like a know-it-all for how repulsed the replies are, Christ... I'm not claiming to have personally engineered the city sewage network, but I'm enough of a plumber to have a thing or two to say when a water main breaks. It's anything but black-and-white.

No, I'm saying that I think that you know what you are talking about and aren't emulating it. 

 

I think you somehow misunderstood me...

 

1 minute ago, AresKrieger said:

All I'm saying is that based on the info presented to counter my point isn't directly related to what I'm saying and actually concludes differently than what said counter point was, essentially invalidating the argument unless new credible evidence is shown.

You don't know that. Depends on the Game.

Every communication between userspace and Kernelspace now cost really performance.

Ins ome games that do load textures on the fly that could cause Problems/Lags/Stutter...

And some games even read stuff from the Storage, loading some stuff of the Level...

[luigi90210]

6 minutes ago, AresKrieger said:

I just got annoyed when he tried to counter something I said with an article that countered what he said with in it ,contradictions simply piss me off

 

YOUR THE PERSON WHO POSTED THE LINK AS FACT!!!!

It's like you fail to realize the info I posted came from something you used as evidence in your post, we might as well throw then source in the garbage since clearly it's unreliable (which was what I was alluding to from the beginning)

 

YOU didnt read that article as i did, i NEVER CLAIMED nvidia and intel will performs better after this patch, unlike yourself

i used their source because it shows a I/O reduction which will DIRECTLY AFFECT HOW WELL NVIDIA'S DRAW CALLS ARE GONNA BE ON INTEL

 

again if you were educated you wouldn't be alluding to anything other than wait(which is what i have been saying for the last 3 posts), you couldnt even spend 5 seconds to click the article and see what GPU was tested before claiming that there is no performance loss in gaming, want to talk contradictions just read your posts and see

 

[DeScruff]

Just now, johnukguy said:

Microsoft are releasing an update next Tuesday for Windows 7 users.

Yeah, but remember that windows update is blocked for those CPUs? https://arstechnica.com/information-technology/2017/04/new-processors-are-now-blocked-from-receiving-updates-on-old-windows/
So is it for this one update going to be the exception?