Velika greška u Intel procesorima, deonice padaju

4

Velika greška u Intel procesorima, deonice padaju

offline
  • Aco  Male
  • Moderator foruma
  • Aleksandar
  • Pridružio: 12 Maj 2006
  • Poruke: 16823
  • Gde živiš: /home/aco

Intel objavio listu procesora pogođenih Meltdown i Spectre ranjivostima



Intel je konačno objavio kompletnu listu svih svojih procesora, koji su pogođeni sigurnosnim ranjivostima poznatim pod nazivima Meltdown i Spectre. Jasno je da su praktično svi Intel procesori pod rizikom.

Kompanija je objavila da će uskoro ažurirati 90 procenata svojih procesora, koji bi trebalo da dobiju softversku ispravku, a ovo se očekuje tokom naredne nedelje.

Kompletna lista Intel procesora pogođenih Meltdown i Spectre ranjivostima uključuje:

Intel Core i3 procesor (45nm i 32nm)
Intel Core i5 processor (45nm i 32nm)
Intel Core i7 procesor (45nm i 32nm)
Intel Core M processor family (45nm i 32nm)
Druga generacija Intel Core procesora
Treća generacija Intel Core procesora
Četvrta generacija Intel Core procesora
Peta generacija Intel Core procesora
Šesta generacija Intel Core procesora
Sedma generacija Intel Core procesora
Osma generacija Intel Core procesora
Intel Core X-series Processor Family za Intel X99 platforme
Intel Core X-series Processor Family za Intel X299 platforme
Intel Xeon 3400 serija
Intel Xeon 3600 serija
Intel Xeon 5500 serija
Intel Xeon 5600 serija
Intel Xeon 6500 serija
Intel Xeon 7500 serija
Intel Xeon Processor E3 Family
Intel Xeon Processor E3 v2 Family
Intel Xeon Processor E3 v3 Family
Intel Xeon Processor E3 v4 Family
Intel Xeon Processor E3 v5 Family
Intel Xeon Processor E3 v6 Family
Intel Xeon Processor E5 Family
Intel Xeon Processor E5 v2 Family
Intel Xeon Processor E5 v3 Family
Intel Xeon Processor E5 v4 Family
Intel Xeon Processor E7 Family
Intel Xeon Processor E7 v2 Family
Intel Xeon Processor E7 v3 Family
Intel Xeon Processor E7 v4 Family
Intel Xeon Processor Scalable Family
Intel Xeon Phi Processor 3200, 5200, 7200 Series
Intel Atom Processor C Series
Intel Atom Processor E Series
Intel Atom Processor A Series
Intel Atom Processor x3 Series
Intel Atom Processor Z Series
Intel Celeron Processor J Series
Intel Celeron Processor N Series
Intel Pentium Processor J Series
Intel Pentium Processor N Series

https://www.benchmark.rs/vesti/intel_objavio_listu.....tima-74491

https://www.tweaktown.com/news/60411/heres-list-in.....index.html



Registruj se da bi učestvovao u diskusiji. Registrovanim korisnicima se NE prikazuju reklame unutar poruka.
offline
  • Pridružio: 05 Mar 2016
  • Poruke: 82

Meltdown and Spectre: Here’s what Intel, Apple, Microsoft, others are doing about it

По овоме чланку сви произвођачи процесора су погођени на неки начин.
Meltdown and Spectre ----- Intel, ARM, Apple
Spectre ----- AMD



offline
  • Pridružio: 14 Feb 2008
  • Poruke: 12340

Ne pratim situaciju ali sam primetio da se gejmeri žale na performanse u igrama koje dosta zavise od procesora, nakon primene te neke zakrpe.

http://steamcommunity.com/app/255710/discussions/0/1620600279671454810/

offline
  • Programer
  • Pridružio: 23 Maj 2012
  • Poruke: 4534

@Njubara, Apple koristi Intel procesore za svoje računare.

offline
  • Aco  Male
  • Moderator foruma
  • Aleksandar
  • Pridružio: 12 Maj 2006
  • Poruke: 16823
  • Gde živiš: /home/aco

Srki94 ::Ne pratim situaciju ali sam primetio da se gejmeri žale na performanse u igrama koje dosta zavise od procesora, nakon primene te neke zakrpe.

http://steamcommunity.com/app/255710/discussions/0/1620600279671454810/


Aco ::Imam razlike, kod mene je veća iskorištenost procesora. I što sam primetio najviše se odražava na igrama. Čak nakon ove zakrpe koje su izdali, igru koju imam instaliranu zna da zkoči pri igranju, moram na restart ili ubiti je task menagera. A sve do instalacije same zakrpe bilo je ok.

Oseti se da, zauzeće procesora je veće a slaboje preformanse, plus kod mene kočenje. Doom recimo nema šanse da nakon pokretanja ne zakuca u nekom trenutku.

offline
  • Pridružio: 14 Feb 2008
  • Poruke: 12340

Nisu samo igrači nadrljali već i developeri :



CPU Usage grafik pre i posle primene zakrpe na nekoj od server mašina koja pruža usluge za popularnu video igru Fortnite.

Više o tome :
https://www.epicgames.com/fortnite/forums/news/ann.....ity-update

offline
  • Pridružio: 05 Mar 2016
  • Poruke: 82

Napisano: 08 Jan 2018 16:15

@E.L.I.T.E
Јесте користе њихов процесор, него сам мислио на њихове чипове.
Citat:Apple's response took a little longer than others, only coming late on Thursday. Apple's position is a bit different; it designs its own chips and sells devices that contain them, and it also designs and develops its own operating system. Apple's reaction was high level, but fairly matter of fact and free of marketing fluff.

Dopuna: 08 Jan 2018 16:23

Исто из чланка који сам поставио, за оне који нису читали, Microsoft powershell скрипта.
Citat:Helpfully, Microsoft has also published a PowerShell script that describes the current system protection against Meltdown and Spectre.

offline
  • Pridružio: 26 Avg 2010
  • Poruke: 10621
  • Gde živiš: Hypnos Control Room, Tokyo Metropolitan Government Building

Nakon instaliranja zakrpe, zaštita protiv ovih napada neće biti automatski aktivirana već mora biti nakndano uključena. Biće da je što se igara tiče, MS na drugom mejstu napravio grešku.

offline
  • Pridružio: 12 Okt 2010
  • Poruke: 10699

Ovo je lepo objašnjeno:

Galaxitus ::This affect about 99% of the hardware in the world including the consoles, smartphone, tablet, PCs, etc.

To explain the Meltdown and Specter's effect, it's actually something quite simple to do:
You know the processor is basically the "task handler" of everything in your PC/Console/Tablet,etc. right? It runs those 0 & 1 at a super high speed.
The thing is that to run those 1 & 0, it has to, technically, store them in a physical way so that it can be passed through to what's next. As it's 0 & 1, it's not encrypted or protected. It's pure raw data. The encryption and protection are usually done after the data has passed through the processor... by a task handled by the processor (ironically). Now, what they have "found" (which is false. it's has been known since the 80's) is that it's possible to access this raw data by force feeding some 0 & 1 to the processor which can be hidden in anything and makes it start an hidden small software which, for example, could send a copy of the raw data through the web.

Since the 80's, it was said that it was relatively impossible to decrypt that raw data which pass through the processor as too much information is missing from their meaning. (Like if you're purchasing something through Paypal, the browser doesn't send the data (to be displayed) through the processor with a big "This is the key to this Paypal account". It's billions of 0 & 1 mixed together going in fast like actual lightning.

The most sensitive possible data that is easier to extract from this process are data like key-logs. Whenever you press any keys on your keyboards, the keyboard send a signal through the wire (or wireless signal) and the motherboard receive it one way or the other. It's a electrical and physical process that is easy to reverse engineer since it's identical regardless of the PC or console. Even consoles have this kind of feature in it, but it's just different. Same with tablet which has a digital keyboard. The processor, to work, does the same thing as an actual key-logger, but for the user and the OS instead of some illegal receiver. What do you think would happen if someone was able to add a small process in your PC so that the processor send all its data not only toward where it's supposed to go, but also to a third party?

As I wrote, the processor only uses raw data. This means that systems that shares data such as Cloud Services basically have this hole that allow anyone to access anything through the processor's real-short memory.
For something like a MMO, one example of use of this weakness in the hardware is that someone, through revert-engineering the data copied and send from the processor, could do anything on the data because he has a registry of everything that is going on in the cloud server.

The fix that is currently being deployed is something really basic and explains why the CPU usage raise through the roof by as much as 40% if not more.
As I explained, the processor doesn't run encrypted data, but instead you got raw data that is encrypted by another processor's task after the raw data passed. The "fix" is to encrypt the data before passing the data (so the processor doesn't have a "log" of the raw file, but instead a log encryption.) This is because steps are not logged... only results of each threads.

Example is if you ask the processor to do a calculation like 10+100*(10/2)-5, it will returns 505. It will log the 505 for a really short while (still enough to be recorded), but it will not record the 10+100*(10/2)-5 steps.
But since the data is encrypted in the processor first, then you got to include the decryption "process" in the calculation process so that what was done with the raw data can be done with the encrypted data.
As the example, that's like if you ask the processor to encrypt something like a password : ABCDE and the encrypted result is 18437439205578, if then you request the processor to manage the task of confirming if it's really 18437439205578, you got to repeat, every time, the decryption process within the confirmation so that anyone who gets his hand on the data only reads 18437439205578 and not ABCDE.
It's 1 additional much more complex task in every steps that, originally, was simple as pie.

This is why the "fix" in the Cloud, in this case, double the CPU usage. as displayed in the graph.
The fix is basically:
Before: (RAW_DATA) > Processor > ENCRYPT_DATA > Processor > STORE_DATA > Processor, then when requested, DECRYPT_DATA > Processor > CHECK_DATA > Processor > ERASE_DATA > Processor.
Now (fix): (RAW_DATA + ENCRYPT_DATA) > Processor > STORE_ENCRYPTED_DATA > Processor, then when requested, (DECRYPT_DATA + CHECK_DATA) > Processor > ERASE_DATA_FROM_RAM > Processor.

With the fix, the processor isn't used as much, but each time, the calculated data is bigger.
(If you don't get why it's slower... It's similar to how we think when we calculate. What's faster?
(200+25) / 5
or
(A*20) + (B * 20 /10) where A is 0.5 and B is 17.5 which is confirmed in the CHECK_DATA request and is not logged.

While in both cases, the answer is 45, but before, someone who has access to the raw data and can reverse-engineer the data will immediately know it's 45 while the one who gets (A*20) + (B * 20 /10) won't know what to do with it because he/she lacks the A or B which is only "known" during the CHECK_DATA and not logged as it's part of the process request and not the results as CHECK_DATA returns a True or False (1 or 0) only.


https://www.epicgames.com/fortnite/forums/news/ann.....ity-update

offline
  • Pridružio: 26 Avg 2010
  • Poruke: 10621
  • Gde živiš: Hypnos Control Room, Tokyo Metropolitan Government Building

https://www.bleepingcomputer.com/news/microsoft/mi.....tre-fixes/

Namjerno ili slučajno je pad performansi na 7 i 8.1 veći nego na 10, pitanje je sad.

Ko je trenutno na forumu
 

Ukupno su 679 korisnika na forumu :: 21 registrovanih, 2 sakrivenih i 656 gosta   ::   [ Administrator ] [ Supermoderator ] [ Moderator ] :: Detaljnije

Najviše korisnika na forumu ikad bilo je 3466 - dana 01 Jun 2021 17:07

Korisnici koji su trenutno na forumu:
Korisnici trenutno na forumu: Apok, Bahuss, Bane san, bankulen, BSD, CheefCoach, Deneb, DonRumataEstorski, dule clio, kuntalo, Milan A. Nikolic, Milos ZA, mračni čovek, novator, Ognjen D., Panonsky, sakota79, sovanova95, Srki98, Steeeefan, zogi036