Performance, Concerns, and Explanations: The security breach in processors that has become the hottest issue on the Web (updated) • HWzone
ComputersProcessorsFeatured articles

Performance, concerns, and explanations: The security breach in processors that has become the hottest issue on the Web (updated)

updating: The patch that fixes the security breaches for Windows 10 is available for manual download for all consumers - you can do this through This link.

Get updates from everyone in TelgramGet updates from us all at TelgramJoin the channel now


The original article: Will consumers need to choose between a slowdown in performance and adequate security that allows them to sleep at night? Is this a gap in my work? Only or all processor keys? It depends who you ask

Reported on the British site The Register He declared (Based on information received From a relatively anonymous Tumblr blog) On a fundamental flaw in the design of the modern labs of the company Which causes their core memory to 'trickle' out, as fixing the problem is likely to lead to a decline in overall performance - creating a massive snowball across the entire network, which continues to roll and evolve with each passing hour, it seems. Is there a reason for panic? Not everyone and not every situation, it seems.

Information that should be secure can be accessed through applications with basic privileges

However, it seems that this is a very specific order of operations that might allow simple applications with basic operating system access to access the kernel of the processor and view information that should be particularly secure And is available only for applications at the maximum access level. Designated malware can exploit this principle to gain access to user passwords, access data, and payment information without being aware that something is wrong.

Any processor of From the era of the, Which has been with us for close to a decade - exposed to some degree or other. For other key processors the situation is a little less clear at this point

It is worth noting that defects in the firmware of hardware products are not particularly rare. The major difference this time is the claim that a firmware update to the processors themselves can not solve the problem, and it needs to be addressed at the level of the various operating systems based on In the architecture of the-64 Common - which in turn can cause tens of percent damage to system performance (with greater damage to relatively old models). Certainly not something that would please the various consumers who paid hundreds or thousands of shekels for a certain level of ability when purchasing their computer system.

A headache for Intel only, And ARM?

The wave of rumors generated by the original report led to a lot of confusion, fear and anger among users, and required Intel to provide Official response Even before distributing the repairs to the various operating systems extensively and removing the embargo information about the problem itself - in its official publication, the chipmaker stated that it is working closely with its partners to repair the damage, which it defines as not exclusive to its products but is a defect in the way it is designed and built at all The modern processors, including those from home And even those based on the E- And not on -64, ostensibly.

The company promised to provide more detailed information about a breach once it opens and unleashes it through dedicated software updates to reduce the potential scope of consumer data - but made it clear that it is not an ability to inject new information, modify or delete it from the system kernel, but read its contents.

A relatively straightforward examination can tell later on whether a particular processor is at risk or not - although at this stage the information obtained seems to be necessarily inaccurate
Image source: phoronix.com

There is Signaling On the part of Its products are not exposed to the breach in question and to the performance that may result from its handling - while the company's factors They keep silent, it seems, hoping that they will not be dragged into the whole mess.

Injury of up to 30 percent in performance?

10 has been distributed to consumers who are part of the XNUMX program Insiders (but not the rest) in December, although the first to provide us with a performance comparison between the system at risk and the security system that received the required patch are actually the Linux experts of the Phoronix site, i7-8700K and the- i7-6800K and received results that could be a siren, at least to some consumers.

There is a decrease in performance due to the update of the operating system, but it is not certain that it is as sweeping and extreme as estimated

The performance of the compilers has not changed at all, and this gives hope that the performance of most of the features that characterize home users will not be affected while the server world, the information farm, and the HPC world will be Which are at an increased risk of significant damage to output. These are only preliminary conclusions that should be taken with limited discretion, but we are ready for any reason whatsoever to be encouraged.

The performance of the games show immunity to the patch - at least under Ubuntu

At this point it seems that simple home users do not necessarily have anything to do about this entire saga. Updates to the various operating systems that will try to block the breach will be distributed as mandatory updates that will be challenging to evade, and the entire process will be behind the scenes and almost transparent. We certainly look forward to hearing more detail about the defect and its significance, but as of now it seems that all we can do is wait for the exams More - and most importantly, hope that the performance impairment demon is not as bad as it has been claimed.


Tags

56 תגובות

  1. I have read that updating BIOS can also solve the problem. It is better to have this level of OS.

  2. This is probably a very deep problem if the MICROCODE update to the processor cannot resolve it. And what about future processors? Were they designed without this loophole in the first place?

  3. Is the mother of the bastards !!!
    It is inevitable to suspect that there is even a deliberate hand.

    It's amazing that all our processors are not vulnerable.
    The vast majority of AND processors.
    And I checked all our Intel processors that are from the 4 series that doesn't include the problem.

    Amazing that high-tech miracles are needed in 2018

  4. People, I recently bought a new book by Haim Shapira called "I Think I'm Wrong."
    Everyone who does - is wrong.
    So many human years are spent in modern processor soaring. And to put it another way, developing a modern processor produces many bugs.
    More than 70% of any hardware project development time is dedicated to debugging and repairing!
    And still, some bugs will always be avoided.
    Just as 99.99999% of the population did not receive 100 in all matriculation exams, and 800 in psychometric.
    We are human beings, for better or worse.

    Even if the companies extend the project 2, or 3 (which would be uneconomical, and then they would prefer not to develop products), * we will * never get to zero-bug mode.

    Ask any programmer, any engineer, or anyone involved in development. There are no perfect products.

    1. Don't you see a difference in the severity of the fault?
      Wouldn't there be a difference in response between a vehicle electric window malfunction and an airbag that throws metal pieces at you and kills you? Actually…

      1. I am a hardware engineer. I've already seen serious bugs, and even more serious ones. I saw chips that cost about a million dollars, and were completely dead due to a stupid mistake that no one understood how they hadn't discovered them before.

        By the way, my specific profession - is to look for bugs!
        (Anyone interested in how to do this in hardware development - welcome to the "specman" wave)

        And as someone who's been trying to find bugs all day, and signing a check on parts of the chips, I know my signature never says "no bugs." She says "within the time allocated to the project, the chances of bugs remaining are already low enough to move forward."

        I'm not saying it's not a joke - I'm saying that the end consumer hears so little joke.

        In conclusion, when making a new chip, they actually produce a stack of bugs. Then try to clean it. But some will always hide ...

        The question is not whether * a bug * was created, the question of whether * discovered the bug *.
        And so the severity of the bug doesn't excite me.
        It is precisely the serious bugs, which are quickly discovered, and you do not hear about them.
        And the fact that this bug took a lot of time to find out.

    2. Maybe you will buy / read a book "How Not to Be Innocent", you and anyone who thinks this particular "glitch" has "suddenly" been discovered.

  5. Quote of zvikaz

    People, I recently bought a new book by Haim Shapira called "I Think I'm Wrong." Everyone who does - is wrong. So many human years are spent in modern processor soaring. And to put it another way, developing a modern processor produces many bugs. More than 70% long ago Development Some hardware project is dedicated to debugging and debugging! Still, some bugs will always be avoided. Just as 99.99999% of the population did not receive 100 in all matriculation exams, and 800 in the psychometric exam. We are human beings, for better or for worse. Even if the companies extend the project according to 2, or 3 (which will be non-economic, then they prefer not to develop products), we will never get zero bugs. Ask any programmer, engineer, and anyone involved in development. There are no perfect products.

    That's not the problem. The problem is that we have reached a situation where x86 with all its extensions is so large, cumbersome, and complicated that it is almost impossible not to run into bugs. It contains so much nonsense and the main thing is to support backwards ...

  6. The "new discovery" that stirs the network has been posted a month ago:

According to the video they uploaded

Hardware Unboxed

It seems that the damage to the home computer after the update is not particularly large
  • Quote of se7en

    ^^^

    Yes.

    The response of the companies.

    Clarifications.

    Meanings.

    I wondered.

    Full new.

    And the fear of massive performance loss.

  • Good gentlemen.

    Do any of the experts here have verified news
    The extent of the performance decline (if any) depends on processors and operating systems
    And of course according to the applications - office, game, graphics 2D, video and three.

    Also, is AMD really not affected in any way by the update?
    That is, or Kay ,, AMD processors do not have the security breach (say) -
    but what ? The patches made in windows do not affect the performance of the processors at all?

    And in a broader way - will we get to the point that there will be windows for Intel and AMD separately?

  • Want or not want you have no choice. In the next cumulative amendment on Tuesday evening he will be installed!

    You messed up.

    I feel no difference and in the Benchmark of just the processor actually runs faster ……….

  • For most home users, this update is a placebo. I am currently running tests on an Optane 900P fast drive to see how much it affects the top end of the storage, the issue should be there most clearly

  • Why? The vague description of this problem is supposed to affect heavy processes that take advantage of the processor's command sets rather than the speed of access to external devices.

  • Lior, my life (sorry for the fascist language),

    Give gas clearly to the issue.

    From what I have read, it is indeed within Intel's obscure border

    Rendering and video eating it From updating the windows.

    There are three specifications stuck in the forum because of this matter.

  • something new?

    Quote of askme

    According to what I have read, indeed within the confines of the obscurity of Intel ) Precisely

    Rendering and video eating it From updating the windows.

    Q: If I have a very large file for Render in Premier, does that mean it's better to finish it before Tuesday night?

  • There is not too much information about how the performance will be affected, but for the home user it seems that there will be no noticeable impact.

    Some of the hawks that have already been published

    https://www.techspot.com/article/1554-meltdown-flaw-cpu-performance-windows/

    http://www.guru3d.com/articles-pages/windows-vulnerability-cpu-meltdown-patch-benchmarked,1.html

    Google also announced that they have installed patches for Linux for all their major systems and the impact on performance is negligible.

    We have to wait and see more data, but it seems to me that the whole performance issue came out a bit out of proportion.

    This does not mean that the problem does not exist, it exists (for example, on Linux servers with very degraded postgre DB).

    https://www.postgresql.org/message-id/[Email protected]) But she probably is not going to affect most people and systems.

    For me the main problem here is that there is a bug in the processor and the only solution they can give is at the code level of the operating system.

    Once hackers know about this weak point, it's a lost battle. It is only a matter of time before they find a way around the correction.

    1. AMD may not have caused the mess,
      But exploit and more how the "immunity"
      Supposedly and the different structure of their processors
      In the riot that ensued.
      It's hard to blame them, I have no doubt that the opposite would have happened,
      Probably in much larger dimensions.

  • How many of these loopholes allow people to steal keys to virtual money from personal computers and organizational computers / hired companies / manage them?

  • Quote of zvikaz

    People, I recently bought a new book by Haim Shapira called "I Think I'm Wrong." Everyone who does - is wrong. So many human years are spent in modern processor soaring. And to put it another way, developing a modern processor produces many bugs. More than 70% long ago Development Some hardware project is dedicated to debugging and debugging! Still, some bugs will always be avoided. Just as 99.99999% of the population did not receive 100 in all matriculation exams, and 800 in the psychometric exam. We are human beings, for better or for worse. Even if the companies extend the project according to 2, or 3 (which will be non-economic, then they prefer not to develop products), we will never get zero bugs. Ask any programmer, engineer, and anyone involved in development. There are no perfect products.

    Modern computing has been invented in an open environment, in research institutes and academies for internal use (such as the UNIX kernel from 1969, which is the basis of all operating systems), and therefore will always be vulnerable. Maybe only AI would be perfect.

  • An inside source whispered to me that Intel was celebrating with champagne. New generations of processors without the "malfunction" will replace the present processors. (And they will probably be slower but without fault….)

  • I don't know how up-to-date you are, but the bug exists in all processors including amd and even Qualcomm… right now it's a minor blow, not as serious as the headlines in the newspapers have done it ...

  • no he's not. There are 3 bugs, the most critical bug, known as Meltdown, exists only in Intel slaves.

    This is the bug that gets the fix that affects certain performance settings of the computer.

    And I really don't think it hits the wing. Intel has been selling processors for more than a year, knowing they have a bug and saying nothing to the public.

    More than that, they offered a new set of processors knowing they had a bug, didn't bother to fix it, and didn't announce it existed.

    I wouldn't be surprised if we see lawsuits against Intel soon for this whole story.

  • I would not attach too much importance to an article on technology in Calcalist, nor to any other article on a newspaper / other economic site.

    My mission may be to understand economics (even that is questionable) but they probably do not understand anything about technology.

  • Calcalist is just the stage.

    More important is who appears in it.

    But I brought it up more as a "color reporter," rather than as a rousing discovery. :)

  • Hello you guys.

    I read in your forum about a loophole for processors.

    How can I check if my CPU is hacked or not hacked? How do I test the processor for good health?

    By the way: I have 10 windows. And I did complete security updates. And I have an i7-6700 processor.

    Thanks in advance!

  • cool it.

    Still nothing broke!

    There is also a security breach in your processor that can be used

    To steal sensitive information more easily than other vulnerabilities.

    The next day, Tuesday, Patch will go out to the windows, which should close the aforementioned breach.

    All is well ! :)

  • It turns out that Microsoft's patch is also dependent on antivirus. In order for the patch to be installed, the antivirus manufacturer must first fix its software.

    The update checks (bargains) before it is installed that the antivirus is up-to-date and if not - the Microsoft update will not be installed.

    I do not know yet what and if the update will be reported if it is not installed because of the antivirus.

  • Quote of lompy

    Except for the b SSD , Where we see quite significant differences.

    Yes. But he hopes to have a solution.

    Most people probably don't remember, but 24 a year ago there was a problem with one of Intel's Pentium models, which was one thousandth of a percent in some calculations…

    Anyone who started out with computers back then may remember (the event was mentioned in the book of Prophet MICROSERFS and the inventor of the name 'Generation X', Douglas Copland. A book that is still worth checking out).

  • Remember well. The error was in the 386 math processor (there was one such thing). To date, this is brought to business administration courses as an example of "why not let engineers do PR".

  • Hi, good evening

    First of all, if you are interested in some of the major hardware sites, I saw a few minutes ago in Guru Benches and various news releases published by Intel and Microsoft.

    After the punch and before the punch if new processors (2016 and above as written) the difference is nil.

    Now the truth is that without relying on these sites, I personally have done some formatting in recent hours and cleaned up the computer and the filters a bit

    Etc. And I can say that after the BIOS update, reinstalling the 10 with the latest fix .. I feel no difference in PUBG and GTA online. The frames

    Same. Standard app performance and the operating system itself - no apparent difference. I did not do too many tests, nor did any tests

    Synthetic but in the real world it looks and feels exactly the same, I btw with i7 7700K rushed at moderate frequency.

  • It is not understood from the article, whether after the problem was discovered in the processors manufactured from here on, the solution will be at the hardware level and whether the state of the updates of the Vinodes knew that there is no need and will not slow down performance.

    Another question, has Microsoft implemented the latest Windows 1803 release for this update and can it be removed (what's the update tax?) In case a person is not afraid of hacking and prefers to maintain performance? (There are cases like this ... Not all of us have 6800k processors. There are many people who still have to use weak processors that barely carry Windows and more on this protection) For example, computers that work offline all year etc ...

  • It's a bit old, but basically no home user needs to worry about it, your computer is probably uninteresting and there are simpler ways to get information from it. In any case, the software patch requires a bit more access to memory and therefore has little impact on performance, and there is no noticeable difference in overall performance. According to Intel's latest release, their new architecture has slightly upgraded the processor memory access capabilities, probably favoring [or eliminating] The damage caused by the repairs.

  • Leave a Reply

    Back to top button
    Close
    Close