[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/g/ - Technology



Thread archived.
You cannot reply anymore.



File: emergency.jpg (15 KB, 350x200)
15 KB
15 KB JPG
UPDATED: this is way bigger than previously thought. Apparently there's more than 1 vulnerability. Intel is the most severely affected and KPTI (performance-degrading fix) is only for them, but there are other security bugs which affects everyone (including AMD).

1) The #IntelChip is only one piece of a much bigger security problem. There are two critical security flaws in microprocessors. The first, called Meltdown, affects virtually all Intel microprocessors.
2) The second, dubbed Spectre, is a design flaw that affects virtually all modern microprocessors (not just Intel, but AMD and ARM) and has NO PATCH.
3) Meltdown presents an urgent crisis and administrators need to implement the KAISER patch ASAP (even though it will slow performance speeds by as much as 30%), but the Spectre flaw is far more pervasive and will require a complete rearchitecture of virtually all microprocessors
4) Spectre is harder to exploit, but has no easy fix, and is far more pervasive. Researchers say it is highly likely the threat from Spectre will be with us for the decade to come.
5) The basic issue is age old security dilemma: Speed versus Security. For the past decade, processors were designed to gain every performance advantage. In the process, chipmakers failed to ask questions about the security fundamentals of their design.
6) Now, Meltdown and Spectre, show that it is possible for attackers to exploit these design flaws to access the entire memory contents of a machine. The most visceral attack scenario is an attacker who rents 5 minutes of time from an Amazon/Google/Microsoft cloud server and...
7) ...is able to access the programs and data off every other Amazon/Google/Microsoft customer renting space on that cloud. That means logins/passwords/SSL keys/files/the motherlode.

[continues in following post]
>>
8) The business/economic implications are not clear, since eventually the only way to eradicate the threat posed by Spectre is to swap out hardware. It's not clear there is even manufacturing capacity for this. It is also unclear who will bear the cost (chipmakers or customers).
9) The industry will be dealing with this for a long time to come. END.

Sources:
- https://twitter.com/nicoleperlroth
- https://twitter.com/tomwarren/status/948674110795800577
>>
1st if true
>>
So will there be lots of cheap second hand cpus flooding the market?
>>
File: INTEL.png (77 KB, 953x535)
77 KB
77 KB PNG
https://security.googleblog.com

There are three risks
>>
Google blog post: https://security.googleblog.com/
>>
>h-how do we implicate everyone in this
>we need everyone to help! it's not just us!
>t-the processors are working to spec anyway
>>
LELELEL

I just wanna watch the CPU's burn. I cancelled my order for a new PC. I'm going to wait this one out untill it's fixed.
>>
>>64130805
gonna have to wait a while bud
>>
>>64130690
You can tell they knew about Meltdown all this time and prepared the announcement of Spectre together with it just to make Intel sound less worse.
>>
>side channel attacks have existed for a decade or so
NSA was working with them in 1955
>>
>it's not an issue with our product

YOU FUCKING LYING CUNT
>>
>>64130780
>Intel getting BTFO with the kernel exploit
>Drops news of second "flaw" it has been saving for this exact moment
>Instead of just them burning now everyone is in the same ship
>Original kernel exploit no longer a big worry
>>
>>64130805
Smart move
>>
>>64130690
Power8 master race
>>
File: zFcVeCT.gif (80 KB, 665x669)
80 KB
80 KB GIF
ahh sheeeeeez.
>>
>>64130837
Who are you quoting?
>>
File: 1504450504567.jpg (120 KB, 540x786)
120 KB
120 KB JPG
Time to install temple os
>>
>>64130690
CPUfags on suicide watch!
>>
>>64130690

>(including AMD)

Incuck lies.
>>
>>64130853
Intel exec on the investor pacification call

Pajeet openly asked them whether it affects AMD or not and they bluffed for 5 minutes and then outright lied when he followed up
>>
File: 1502326078595.png (391 KB, 800x733)
391 KB
391 KB PNG
>>64130690
I for one am glad intel dragged amd down with them :)
>>
Daily Reminder that silicon fabrication and production is organised years in advance. Intel's next 2 entire gens will still have this issue.

The CEO bailed. When the pilot is ejecting out of the roof with his briefcase and parachute, you know what's going down.

We're about to see a repeat of Seagates spectacular fall from supremacy to Western Digital circa 2008
>>
>>64130858
Are the glow in the dark CIA niggers behind all this?
>>
https://security.googleblog.com/2018/01/todays-cpu-vulnerability-what-you-need.html

>These vulnerabilities affect many CPUs, including those from AMD, ARM, and Intel, as well as the devices and operating systems running them.

AYYMD IS FINISHED & BANKRUPT

AYYMDPOORFAGS CONFIRMED ON SUICIDE WATCH
>>
LOL GayMD is fucking finished for fucking lieing about this!!!
>>
>>64130690

go away brain
>>
>>64130690
>2) The second, dubbed Spectre, is a design flaw that affects virtually all modern microprocessors (not just Intel, but AMD and ARM) and has NO PATCH.
>dubbed Spectre
>Spectre
As in its a ghost, it doesn't exist, Intel just threw this in to cover their ass.
>>
File: Everyone Was Harmed.gif (1.88 MB, 500x370)
1.88 MB
1.88 MB GIF
>as we see it we don't expect any financial impact
>I wouldn't expect any change in acceptance of our products
>>
>>64130748
Yeah, absolutely tons of secondhand server hardware that's insecure for cheap.

Holy fuck you could build server-class home NAS for the price of an NAS today
>>
http://www.jpost.com/Israel-News/Intel-CEO-We-think-of-ourselves-as-an-Israeli-company-as-much-as-a-US-company-484209

Oy vey
>>
File: 1511709667364.jpg (66 KB, 803x711)
66 KB
66 KB JPG
>>64130805
>until it's fixed.
>>
>>64130823
probably true.
But it actually doesn't evem matter. it's ogre now.
>>
What does this mean for me an average poweruser who only uses his desktop rig a few hours a day for gaming and windows optimizations?
>>
>>64130690
Does that mean the apple chips in the ipads are good to go? Will Apple have an advantage?
>>
>>64130891
CIA niggers are incompetent hacks. You mean NSA niggers.
>>
>>64130939
literally nothing, don't update
>>
>>64130805
Enjoy your nocomputer for the next 5 years. You can't fix this shit, they've been cheating all this time by having this insane flaw that allowed them to stay faster than the competition

This is tearing down the tower and rebuilding from scratch, not a quick remodel
>>
>>64130939
If you connect to the internet you will have your PC host to a Pajeet coinmining operation
>>
>>64130939
nothing. but the porn site you go to and companies that run your online games are about to get assraped.
>>
>>64130960
Oh fuck
>>
>>64130690
And this is why I only use Allwinner CPUs. With the Chinese CPU we are all winners.
>>
File: Popura_Tantrum.gif (851 KB, 400x225)
851 KB
851 KB GIF
>>64130838
crap crap crap.
they've outsmarted us.
>>
Spectre is already getting patched.
>>
>>64130958
What about for laptop users?
>>
This will result in many newcomers competing and innovating while the price lowers and performance soars right?
>>
>>64130971
Cucks will have to downgrade from 4k to 720p blacked.com vids?
>>
what the fuck? i just read how amd said all 3 variants don't affect them?? who do i trust?

https://twitter.com/rhhackett/status/948676213505232897
>>
>To take advantage of this vulnerability, an attacker first must be able to run malicious code on the targeted system.
>>
> locally executed
> still can't modify memory

really makes one ponder
>>
>>64130882
you couldn't handle one night and most of a single day of butthurt, could you?
>>
>>64130983
As they said many many MANY times, "average users" won't see any effect locally.

But the websites you use are high-impact and will either become creakingly slow or have to reduce service or quality significantly to cope
>>
THE IMPORTANT part is that the Meltdown fix incurs a huge performance loss, who the fuck cares about this Spectre meme?
>>
>>64130819

It's going to be my very first top of the line gaiming PC. I don't mind waiting for an extra year for some i7 9700k (fixed) and GTX 1180/2080

>>64130842

I hope so
>>
>>64130950
>ARM
>>
>>64130897
>>64130899
pathetic inshills
>>
Where is the list of Intel CPU's it effects? I see "Modern" but how far back are we talking? Skylake? Devil's Canyon? Sandy Bridge? Core 2 Quad days?
>>
Why doesn't IBM take advantage of this?
>>
>>64131002
>lol they can't change your passwords, just read them
>>
>>64130995
Are you stupid or just retarded?

Intel GPUs have HEVC Main10, VP9 10bit and H.264 hardware decoding up to 8K res from Kaby Lake onwards
>>
>>64130690
Should I run out and buy as many PowerPC as I can?
>>
>>64131001
man if they're lying about AMD... well I wouldn't put it past them
>>
>>64131019
Pentium 2 onwards

Early 90s
>>
>>64131019
1995
>>
>>64130919
> absolutely tons of secondhand server hardware that's insecure for cheap
sell BTC
>>
>>64131019
pentium pro
>>
>>64131019
All x86 since the 90's.
>>
>>64130939
>only uses his desktop rig a few hours a day for gaming
>poweruser
Pick one.
>>
>>64130995
yes. or increased ads or cost to use.
>>64131030
are fucked? we are talking cdn.
>>
>>64131006
No
>>
>>64131008
What if they just lowered their dividends?
>>
>>64130823
>You can tell they knew about Meltdown all this time

They truly aint shit.
>>
>>64130909
These numbered points didn't come from the Intel conference call, which has just finished, it's from some cunt who writes for the Verge and quotes no source. No indication where the details, let alone the names came from. Chances are the exploits don't even have names adn he's made them up himself, like those cunts who hyped up a minor issue a year or so ago, called it a scary name like badlock and started a viral campaign to raise awareness of it unnecessarily.
>>
>>64131073
"no"
>>
>>64131077
They're decent names though.
>>
>>64131032
How do we know PowerPC isn't affected?
It may just be untested due to lack of machines to test on.
>>
https://meltdownattack.com/meltdown.pdf
https://spectreattack.com/spectre.pdf
>>
>>64131077
They have names, but the names are
>Bounds Check Bypass
>Branch Target Injection
>Rogue Data Load

https://s21.q4cdn.com/600692695/files/doc_presentations/2018/Side-Channel-Analysis-Security.pdf
>>
File: sugoi.jpg (39 KB, 960x540)
39 KB
39 KB JPG
>>64131001
THANK YOU ANON
YOU JUST SAVED ME.
THANK YOU.
>>
I love how this is a blatant attempt at damage control on Intel's part.
>oh fuck this security flaw is really an issue
>oh fuck the patch just makes it worse
>what do we do?
>I got it, we make up some bullshit that effects EVERYONE so it can be a team effort to fix everything and not just a failure on our part.

Nonsense.
>>
what about Power Processors?
>>
We've been fine for 40 years without this problem, how bad can it be. Plus it only affects the CPU it's just billions of ones and zeroes why would anyone want to hack them?
>>
>>64130974
True, with that chip you already KNOW it has a state-controlled backdoor in it. No need to wonder.
>>
>>64130690 So we will just go back to good ols dedicates servers and shared hosting. It's no the end of the world. Dumb normies would pay all the extra expenses anyway.
>>
>>64130989
I'm sure some burgers believe this.
>>
>>64131086
I'll give you that, spectre is cool. Meltdown is too obvious in a shit made for TV hacker movie kinda way
>>
https://www.phoronix.com/scan.php?page=news_item&px=Google-CPU-Disclosure

>The issue is now being called "Meltdown and Spectre" with the bug description up at SpectreAttack.com.

AYYMD IS FINISHED & BANKRUPT

AYYMDPOORFAGS CONFIRMED ON SUICIDE WATCH
>>
>>64130690
I'm glad to see the end of computing. Time to get out my Itanium Server.
>>
The funniest part is how we will be seeing servers being compromised by this shit for decades because no one ever updates until it's too late, just look at the wannacry meme that only affected ancient WinXP systems.
>>
YEAR OF THE RISC-V PLUS GANOO LEENUCKS DESKTOP
>>
>>64130959

If it takes them 5 years to fix it, they might as well call it a day and go bankrupt. They will work like crazy to fix this ASAP
>>
What the fuck is this I just take a day to enjoy seeing someone still use Windows XP and there's all this shit going on

So effectively if you're using a computer or a smartphone you're buttfucked or what now? Have we come to the point where, maybe by 2025, people will stop using technology or something? I don't know what the fuck is going on anymore...
>>
pic related.
what a wild ride.
>>
Time to dust out my trusty 22mhz computer from 1990
>>
>>64131023
>>64131032
>>64131103
Speculative execution has been a feature of pretty much any architecture worth a shit for decades. The only reason POWER wasn't mentioned is because nobody gives a shit about it.
>>
>>64131001
PLS BE TRUE

I DON'T WANT COMPUTERS TO DIE
>>
>>64131124
They're patching, but that will fuck performance.

Replacing the chips they've been using for 20 years and that have been then only thing keeping them ahead of the competition isn't quick or easy
>>
>>64131092
... which he clearly thought weren't dramatic enough. Rogue data load is the only which probably sounds a little menacing to the layman due to the word rogue.
>>
>>64131023
They don't have the capability anymore. They don't make anything. IBM really isn't relevant anymore. They're just "idea"-guys now.

They do have the new Power chips I remember reading about some time ago. Sadly, they've already abandoned the field to Intel, much like AMD did.
>>
File: 1510031718253.jpg (290 KB, 1012x1324)
290 KB
290 KB JPG
Terry was right when he said Intel were CIA niggers trying to shove as much bloat on their silicon as possible.
>>
File: 1492308494997.gif (2.12 MB, 480x270)
2.12 MB
2.12 MB GIF
>>64131096
and still, their back up plan backfired.
>>
>>64131143

Then they'll call it a day and accept their huge losses. Time for another player (AMD) to be the champ.
>>
>>64131148
Your mom enjoys rogue data loads every night.
>>
>>64131127
>Have we come to the point where, maybe by 2025, people will stop using technology or something?
Exactly what the people at (((The Top))) want to see.
>>
I fucking hate all the noise from corporate cocksuckers. So much mis-information from people spamming their favorite company won (they do it for free btw) is causing confusion for people who really need to understand this issue is tied to many modern CPUs from all ISAs

If you believe everything a corporate entry says on this matter for the time being then you need to remove the cock from your mouth before you pass out.
>>
>>64131117
>>64131001
>>
>>64131106
Yeah, just like intel
>>
>>64131158
Can't abandon a field you never played on, mainline POWER chips were never made to compete with Intel offerings on the consumer market and IBM never wanted them to.
>>
>>64131117
I was, for like 5 seconds until I checked the sources.
but you? you're still fucked.
>>
>>64131148
>https://www.phoronix.com/scan.php?page=news_item&px=Google-CPU-Disclosure
I take this back, I see that the guys who claimed the bounty on it named them that. I guess they are entitled to do so.

https://spectreattack.com/
>>
A hivemind civilization wouldn't have this problem.
>>
File: Clipboard01.png (491 KB, 918x612)
491 KB
491 KB PNG
None of this would have happened if there were more women in technology.
>>
> 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.
>>
>>64131184
You hate differing opinions? Are you mad that you made a mistake?
>>
File: golly gosh.jpg (190 KB, 461x692)
190 KB
190 KB JPG
>>64130780
>>
>>64131202
Even the men look feminine what is this shit
>>
>>64131188
>Yeah I-I may still be a bootlicker but so are you!
ARMlets do it again.
>>
>>64131170
That you Dan? If so, no-one cares.
>>
AYYMD is lying as usual

Lied about Barcelona performance & TLB bug

Lied about Vega performance and power efficiency

They are lying about the vulnerability not affecting them, you can't trust these liars
>>
intel is dying before our very eyes
>>
>AMD prices begin to climb

oh god no
>>
So this has been a thing for years and nothing has been affected.
>>
>>64131167
Exactly. You know for a FACT that AMD probably has everyone running on all cylinders looking for any sort of flaw in Ryzen. Anything that can come and bite them in the ass down the road. Especially since all of their future chips are built on gen 1 Ryzen. Threadripper, EPYC, Ryzen+, Ryzen 2, etc.
>>
>>64131184
All Intels are fucked by Meltdown, fix reduces performance up to 63% depending on workload and model.
AMD doesn't have this vulnerability.
That's all you need to know.
>>
Whats spectere and how can they fix it. Will it also come with a profirmance hit
>>
>>64131202
>Hey Prash, we're taking pictures for the website could you come over here?
>Yeah gimme one sec to put on my sunglasses, you mind taking my picture while I'm screaming?
>>
>>64131243
t. amd fanboi
>>
This is what you cucks get for about buying an macbook
>>
>>64131238
There has been thousands of hacks that was probably caused by this

Now that everyone knows about it there will be even more
>>
>>64131226
Amd shills are delusional. They think amd is their friend and they can trust them
>>
>>64131223
Bottom left isn't so bad
He's fat but I'll take a fatty over some little twink soyboy
>>
What is the difference between Spectre and Meltdown?
>>
>>64131244
It's some irrelevant noise brought up by Intel to distract people from Meltdown.
>>
>>64130690
>AMD statement:
"None of these attacks work on our chips."

Well. Time to go back to AMD.
>>
>>64131243
where did 63% come from
>>
Join us now and free the software
You will be free hacker you will be free
...
>>
>>64131243
Fucking wrong retard it is fucking 5% at best gaymd shill.
>>
>>64131260
Understand that EVERYTHING is fucked from grandma's CompaQ to macbooks, to servers, to gaming powerstations
>>
Both Intel and AMD are trying to damage control the fuck out of this shit. Do not trust either or become nothing but a low intelligence person who shouldn't be allowed to reproduce.
>>
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
>>
Recording of the Intel call for anyone who missed.
https://my.mixtape.moe/cqnhdh.ogg
>>
>>64131001

>trusting intel

Not even ONCE!
>>
>>64130690
>>64122607
>>64131020
>>64129452
>>64125824
>>64129994
Ya'll fuckers need to check the catalog
>>
>>64131300
And the slides they refer to

https://s21.q4cdn.com/600692695/files/doc_presentations/2018/Side-Channel-Analysis-Security.pdf
>>
>>64131293
>>64131289
>>64126171
>>
>>64131289
He was lying, it's closer to 80% in I/O tests.
>>
>>64130851

Please, upload the collection. :3
>>
File: 0f42dda4ed.png (273 KB, 618x568)
273 KB
273 KB PNG
>>64131001
more convenient format.
this will be reposted a lot I reckon.
>>
>>64131307
This is a happening, anon, this isn't some fucking chink shit general
>>
>>64131243
>63%

First it was 30% now 63%? I'm smelling bullshit.
>>
>>64131296
>AMD literally excluding the fix from their CPUs since it doesn't affect them at all
>damage control
>>
The truth is:

>some ARM64 designs are affected by this same flaw
>AMD is one source manufacturing these ARM designed cores
>intel implies that AMD is affected by this bug

The security flaw DOESN'T affect AMD x64.
>>
>>64131331
The percentage absolutely depends on what is done.
>>
File: mfw.jpg (27 KB, 534x504)
27 KB
27 KB JPG
>mfw Socket 7 Pentium (the newest CPU unaffected by all of this) skyrockets in price
>happen to own five MMX 200 chips for no discernable reason
>>
>>64131314
Literally fake garbage paid for by gaymd
>>
>>64131200
https://meltdownattack.com/

>Has Meltdown or Spectre been abused in the wild?

>We don't know.
>>
>>64131347
7 Pentium is affected
>>
>>
File: 1514979384044.png (887 KB, 550x550)
887 KB
887 KB PNG
La Productividad Extintor...
>>
>>64131367
Best meme of week
>>
on the bright side, I'm so glad I'm not a sysadmin for any company right now.
>>
>>64130690
isn't it possible for them to implement a more efficient patch later on which speeds up the processor again?
>>
>>64130690
>>64130690
>>64130690
>>64130690
PATCH FUCKING WHEN???
>>
I'm willing to bet money they knew all along what this exploit is/was. Maybe even intentional. How do you think FBI/CIA/NSA gets in? You don't think those agencies actually make their own exploits and backdoors do you?
>>
I will not update
>>
>>64131019
All computing devices back to the abacus
>>
>>64131273
Intel shills are delusional. They think Intel is their friend and they can trust them
>>
>>64131226
>>64131273
.10c has been deposited into your account.
>>
>>64130690
>https://meltdownattack.com/meltdown.pdf
>We also tried to reproduce the Meltdown bug on several ARM and AMD CPUs.
>However, we did not manage to successfully leak kernel memory with the attack described in Section 5, neither on ARM nor on AMD.
>neither on ARM nor on AMD.

INTEL BTFO
>>
honestly lost all interest now that I know kiketel isn't the only one getting rekt. I only cared about this because I thought it would mean kiketel would finally burn to the ground.
>>
>>64131120
Wannacry affected unpatched Windows 7 and Server 2012 systems as well.
>>
File: mfw2.jpg (32 KB, 523x409)
32 KB
32 KB JPG
>>64131364
Aw shit
>>
>>64131391
There's no efficiency to be had here, the fix literally disables the functions that are vulnerable and offloads them to a slower but safe system, there's no way to actually fix it through software.
>>
>>64131396
We knew they been knew.
>>
>>64131342
Opterons are affected?
>>
File: intelplsgo.png (212 KB, 793x307)
212 KB
212 KB PNG
this is straight from google's project zero

stop making shit up intel babbies
>>
I listened to the conference call:

1. They tried real hard to make it seem like this is just a software attack and that the CPUs are operating AS DESIGNED. Fucking assholes: If the design allows for the attack to happen, then the design is FLAWED by definition.

They're trying really hard to avoid being labeled as "flawed".

2. They completely fucking sidestepped the datacenter workload question. "We don't classify workloads by datacenter or user". They then said "average user workloads won't see a significant impact from the software fix".

Again: kiketel jewing the english language to avoid addressing the real problem: all datacenter servers running intel chips are gonna get impacted.

3. Pajeet asks if they know whether or this is an "industry issue" or an "intel issue". Senior kike officer gish gallops the questioner for five minutes with intel bullshit and FINALLY responds saying "industry issue" after pressed.


In short: Intel is going to become the catalyst for a performance holocaust the likes of which cpu cycles have never seen.
>>
>>64131423
AMD isn't affected.
>>
So I should just change to AMD, right?
I don't have any preference, I think, can I still use Nvidia GPUs with it or do I have to change the whole thing?
>>
>>64131396

Someone is going to end up in pieces in a suitcase. They gonna say it's a suicide LMAO
>>
>>64131423
>implying "Spectre" performance hit even exists
>implying "Spectre" even exists
>>
is my gaymin gonna be hit by this performance loss or what
>>
>>64131469
They made it unsafe by design to make it artificially faster. :^)
>>
>>64131396
Just wait until someone discovers all the UEFI backdoors
>>
>>64131486
What the fuck do I know, I'm just a dumb fucking normie who hates kiketel
>>
File: 1512101882943.jpg (8 KB, 272x252)
8 KB
8 KB JPG
>>64131434
I think the last is a regular pentium, even pentium pro is compromised
>released 1993
There are literal adults on /g/ younger than that, just let that sink in
>>
>>64131469
>In short: Intel is going to become the catalyst for a performance holocaust the likes of which cpu cycles have never seen.
So you're saying it's another shoa?
>>
>>64131487
Expect lag spikes if you do multiplayer.
>>
At least i will now get the chance to enjoy every single frame of my games
>>
>>64131487
Around 3% based on a small set of benchmarks. If you do play multiplayer games then that may be a bigger issue for the servers.
>>
>>64131052
Most x64 AMD processors aren't affected by one of the security flaws
>>
>>64131001
>AMD is not susceptible to all three variants
This means they're susceptible to at least one.
>>
>>64131479
nv gpus should be fine as long as you have an amd cpu
>>
>>64131473
Yes it fucking is retarded shill
>>
>>64131447

ARM64 server chips are branded opteron.
>>
>>64131479
Sure, Ryzen + GeForce master race is the only true race.
>>
>>64131457
INTEL FAGGOTS

B T F O
T
F
O

SELL YOUR INTEL STOCK ASAP
>>
Even AMD is affected by this?
>>
>>64131423
see >>64131457

AMD is unaffected.
>>64131469
my cock is hard thinking about how hard they're going to hit the ground. I wouldn't feel this way if Intel wasn't guilty of Jewish trick after Jewish trick. With their shady as fuck business practices and gimped products, I hope they go down.
>>
>>64131364
P5 doesn't do speculative execution though?
>>
>>64131510
>at least one
Intel?
>>
>>64131226
fucking this. I'd take honest company over a fucking lying leaf any day
>>
>>64130690
Can someone explain to me what is going on? I have been out of it since the 26th of December due to appendicitis and a complication that put me under heavy drugs till this morning.

I'm sysadmin on paid leave right now and I opened my laptop to HUGE amounts of emails, my phones to texts, from other sysadmins and my bosses.

What the fuck is going on?
>>
WHOOPS
>>
File: 1515020500540.png (254 KB, 1424x634)
254 KB
254 KB PNG
>>64131355
>fake garbage
>intel ceo lied twice already on live TV today
>he dumped all of his stock
>they failed to drag AMD down with them
>it backfired beautifully, it didn't even last one minute.
>>
>>64131500
Fuck, I'm about to hit platinum in Street Fighter V. Should I really stop?
>>
Isn't this the most serious thing to happen in a very long time?
>>
>>64131512
>>64131518
Ok then, I need some upgrades anyway. Might as well get a new gpu and a new ssd

I hope at least amd doesn't force me to use W10 like intel does.
>>
Jesus christ when did hiroshimoot let these blatant marketers in
>>
File: intel.png (573 KB, 597x767)
573 KB
573 KB PNG
ALL YOU HAD TO DO WAS LISTEN
>>
>>64131457
Link? Would be useful to counter the pankajtel shills
>>
File: secure system.png (215 KB, 800x600)
215 KB
215 KB PNG
>>64131498
>>64131347
Feels good being safe.
>>
>>64131342
If what you say is true, I swear, I will NEVER buy any Intel product again. This is the scummiest of scum tactics.
>>
>>64131534
Someone finally figured out that the thing that lets Intel processors be better than everyone else is a huge backdoor to access your entire computer via the kernel memory

Intel investor pacification call
https://my.mixtape.moe/cqnhdh.ogg

Accompanying slides
https://s21.q4cdn.com/600692695/files/doc_presentations/2018/Side-Channel-Analysis-Security.pdf
>>
>>64131543
It will but there are workarounds.
>>
File: 1468950650321.jpg (22 KB, 600x507)
22 KB
22 KB JPG
>inb4 AMD turn all smug, market their shit as super secure and price Zen+ shit double
don't do it fags
>>
>>64131552
>http
>safe
>>
>>64130690
ARM64 is effected, not ARM32
>>
>>64131560
I figure the workaround are the same as the ones for intel
>>
>>64131524
Yep, their literally finished, fix drops performance over 50% LOL!
>>
>>64131534
>I'm sysadmin on paid leave right now and I opened my laptop to HUGE amounts of emails, my phones to texts, from other sysadmins and my bosses.

Condolences.
>>
>>64131534
Tell yer boss to buy AMD next time.
>>
>>64130765
>>64130780
> I'm not dirty you are dirty

Intel mudslinging in progress
>>
>>64131534
tl:dr;

huge exploitable flaw in Intel's architecture going back to every single chip they've made for the last 10 years
>>
>>64131549


This makes sense now
>>
>>64131542
People seemed to have gotten over heartbleed quickly
>>
File: 1497140200987.gif (282 KB, 499x499)
282 KB
282 KB GIF
Thank God I have a Ryzen™
>>
>>64131513
see
>>64131457

and the official report on the Meltdown exploit. The researchers specifically say that the attack didn't work on ARM nor AMD
>>
>>64130690
Is the bug happening in long mode only?
>>
File: aaaaaaa.png (118 KB, 327x333)
118 KB
118 KB PNG
>mfw consolefags win again
ITS NOT FAIR
>>
You all know this is a ploy to buy new computers, right? There's no way they couldn't have found this in my ten year old core2quad this far in the game and for damn sure the slowdown fix is suspect.
>>
>>64131542
Yup. This worse than shellshock or heartbleed. It's a huge fucking disaster.
>>
>>64131457

%rip in piece, Intel
>>
>>64131524
AMD is confirmed unaffected by all three issues, according to both AMD and the team who revealed them.
>>
>>64131567
Irrelevant if i don't transmit anything and don't enable JS. No idea why 4chan didn't use https by default, that browser does support it
>>
File: 43242323.png (83 KB, 614x524)
83 KB
83 KB PNG
We won bois!
https://twitter.com/bryanlunduke/status/948430797266042880
>>
>>64131539
The game is trash anyway, most people are waiting until AE comes out
>>
>>64131536
Spectre affects pretty much any CPU which implements spectualative execution.
>>
>>64131606
Of course they knew, but nobody else did. That's the point.
>>
>https://archive.4plebs.org/pol/thread/117886401/

Why didn't we listen?
>>
>>64131603
Apparently not, since everything since Pentium pro is supposedly affected.
But PCID requires 64bit mode
>>
>>64131606
>10 year old
The exploit is more than 20 years old, it's ogre.
>>
>>64131483
Came here to say this, but was too skeered to mention it first.
>>
>>64131562
AMD won't do shit because Intel never lowers prices under any circumstances
>>
File: 54.png (17 KB, 109x65)
17 KB
17 KB PNG
>AMD rebukes Intel, says flaw poses 'near-zero risk' to its chips
>"To be clear, the security research team identified three variants targeting speculative execution. The threat and the response to the three variants differ by microprocessor company, and AMD is not susceptible to all three variants. Due to differences in AMD's architecture, we believe there is a near zero risk to AMD processors at this time."
https://www.cnbc.com/2018/01/03/amd-rebukes-intel-says-flaw-poses-near-zero-risk-to-its-chips.html
>>
>>64131549
>ids
so intel insider posted on /pol/ and they didnt listen?
>>
File: gotem.jpg (23 KB, 596x554)
23 KB
23 KB JPG
>>64131576
>Shintel processors literally drop to laptop tier hardware
>>
>>64131551
https://googleprojectzero.blogspot.com

data taken from Google's Project Zero
>>
odds on cia/nsa being involved with this?
>>
Just got an Intel 8400 how fked am I bois?
>>
SAVE THE ANIME AND DISCONNECT FROM THE INTERNET
>>
>>64131641
>/pol/ was right again

what the fuck
>>
>>64131659
they probably were

it helps facilitate the killing of browns
>>
>>64131641
Intel ME is unrelated
>>
>>64131641
Why do we never listen
>>
>>64131605
>implying consoles aren't also affected
>>
File: 1514936278873.png (162 KB, 633x900)
162 KB
162 KB PNG
NOOOO! THIS WAS OUR INDUSTRY TO RULE FOREVER! WE HAD A PLAN! IT WASN'T SUPPOSED TO BE THIS WAY!
>>
>>64131568
I'm getting mixed signals here, some say they don't others say they do, any link to prove your statement
>>
>>64131627
You're so damn late.
>>
>>64131660
you're about as fucked as a pornstar in a gangbang scene
>>
>>64131285
more like "None of our chips work"
lmao
>>
https://newsroom.intel.com/news/intel-responds-to-security-research-findings/
Is Intel actually lying now? Wtf I don't know what to believe anymore
>>
>>64131594
Because the fix was trivial. This flaw requires gimping performances on Intel for meltdown and fundamentally changing CPU architectures for everyone for Spectre.
>>
>>64131524
Google says AMD is definitely unaffected by 2 of three varients and almost certainly unaffected by the third
>>
File: 1515011589961.jpg (58 KB, 700x700)
58 KB
58 KB JPG
>>64131684
>>
AMD suffers from spectre only. NOT MELTDOWN.
>>
>>64131694
More like
Intel official statement:
"our processors are literally laptop tier now"
>>
>>64131692
Damn it man I just wanted to play games
>>
>>64131655
I just really hope nothing is discovered in AMD's architecture. I hope they're not sitting on their laurels waiting. That AMD is actively auditing their own shit.
>>
>>64131524
We don't know about SPECTRE, but the other 2 don't affect AMD
>>
>https://meltdownattack.com/
Patches will roll out next Tuesday and it looks like your games and everyday activity on your desktop won't be affected.
>>
>>64131666
>>64131680
/pol/ is a fucking hypocrite for supporting drumpf the kike lover.
>>
Damn Intel is crashing horribly with no survivors
>>
>>64131659
They've probably known this for years
>>
>>64131696
>now
Their entire business in built on lies, you only noticed now because you're paying attention.
>>
>>64131689
See >>64131655
None of bugs affect the AMD CPUs
>>
Senior Microsoft software engineer here. Guys, I am trying SO hard to make sure the performance hit affects AMD too, please give me your support to keep the hardware market fair.
>>
>>64131696
Intel has been lying ever since their CEO sold all his stock in December in a totally unrelated event that definitely wasn't because he was told he'd be resigning in early January
>>
Fuck off AMD shills.
I bet you this is 100% exaggerated and doesn't affect anyone.
>>
>>64131656

He got a lot of (You)s so I think people were listening, especially since he made it sound like the vulnerability was used to spy on Trump
>>
File: virus dos.jpg (109 KB, 1920x1080)
109 KB
109 KB JPG
>not running everything in ring0 anyway
plebs!
>>
>>64131720
save up for a 1200 or 1600, it's basically an unlocked i3/i5 with a different chipset
>>
File: 1436852241141.png (123 KB, 644x635)
123 KB
123 KB PNG
>>64131641
>guy drops basic pc knowledge and claims to work for intel
>/pol/ laps it up harder than 3 day old jizz off the floor of a truckstop bathroom
>>
>>64131725
Your games will be affected, just not as much as data centers will be.
>>
So if I get this right the biggest problem is not us casual users being affected, instead, big companies.
>>
>>64131656
Pol never listens... it's usually dismissed as "larping garbage"
>>
>>64131510
They are, and from the sounds of it it's similar to the intel one in that it's an architectural problem, but it has no patch and is much harder to execute
>>
>>64131108
top kek
>>
>>64131771
Most websites you use, including this shithole, have Intel hardware

At best they'll run slow, at worst they'll cave completely
>>
>>64131705
THE FUTURE OF USA AND ISRAEL
>>
On an i5 6600k with a GTX 660 on Win7. I probably won't update windows as I haven't updated it in ages anyway.
I was thinking of upgrading to a 1070, but with this news I might go with an RX480
>>
>>64131629
>near zero on all 3 variants

you were in denial, you had a glimpse of the light and now you're back in stage 1 denial.
>>
File: meltdown.png (21 KB, 642x472)
21 KB
21 KB PNG
intel is fucking done
>>
>>64131749
>The announcement has the attention of alrge security auditing companies
>Major data center giants like Amazon and Google are concerned enough to be seen publicly asking about potential impact
>Microsoft and to a lesser extent the Linux community is up in arms about rolling out a patch ASAP
>Exaggerated

How do you function with this level of delusion? Is day to day life difficult?
>>
>>64131555
>>64131579
>>64131580
>>64131590

So malware locally installed on a machine can peer into the cache regardless of firmware and software and be used for corporate espionage?

>tfw we just updated all our workstations

So there are 3 different types of side attacks that peer into the cache, and only amd is immune to 1 of them.

So in this "backdoor" that the 3rd type of attack that only effects intel CPUs was just a huge telemetry function which allowed them to make "better" performance numbers?
>>
>>64131679
Explain.
>>
>>64131794
Guess that damage control on CNBC worked well.
>>
>>64131783
what's stopping them from just changing to intel?
>>
>>64131778
No, they arent.
Google's Project Zero confirmed.
AMD is clean.
>>
IF Meltdown affects AMD there's no way anyone would trust them EVER again. Let's just wish they didn't fuck up.
>>
File: 1510021433132.jpg (37 KB, 960x540)
37 KB
37 KB JPG
>soon the only safe way to browse the internet will be on your old C64
>>
>>64131013
Pretty sure there isn't a Spectre fix yet, so nobody is sure of its potential performance impact.
>>
>>64131008
great! i hope they get rid of javascript first.
>>
>>64131733
Know that already. Intel just doing some damage control.
>>
File: intlel tweet.png (446 KB, 662x1408)
446 KB
446 KB PNG
>>64131327
>>
>>64130690
Lol, I don't care, my Athlon x64 x2 is still performing well with 8600GT nVidia.
>>
Fucking sigh.

First Heartbleed, then Krackattacks, now THIS?

Maybe I should just give up computing.
>>
>>64131807
changing from intel*
>>
>>64131802
The intel ME is a separate processor with it's own OS. It had bugs but they were "fixed" with a firmware update some months ago
This bug is about the main system processor that runs all your shit.
>>
>>64131765
B-but muh always right
pls pay attention to us
>>
This is fucking huge.

Desktop, Laptop, and Cloud computers may be affected by Meltdown. More technically, every Intel processor which implements out-of-order execution is potentially affected, which is effectively every processor since 1995 (except Intel Itanium and Intel Atom before 2013). We successfully tested Meltdown on Intel processor generations released as early as 2011. Currently, we have only verified Meltdown on Intel processors. At the moment, it is unclear whether ARM and AMD processors are also affected by Meltdown.

Which systems are affected by Spectre?
Almost every system is affected by Spectre: Desktops, Laptops, Cloud Servers, as well as Smartphones. More specifically, all modern processors capable of keeping many instructions in flight are potentially vulnerable. In particular, we have verified Spectre on Intel, AMD, and ARM processors.

Which cloud providers are affected by Meltdown?
Cloud providers which use Intel CPUs and Xen PV as virtualization without having patches applied. Furthermore, cloud providers without real hardware virtualization, relying on containers that share one kernel, such as Docker, LXC, or OpenVZ are affected.
>>
>>64131789
nigger this is CPU not GPU
>>
>>64131814
TempleOS doesn't have virtual memory either so I guess that's a "modern" "alternative".
>>
>>64131696
they're not lying, just using sneaky linguistic tricks: they grouped together the exploits, one of which affects intel, one of which affects everyone
>Recent reports that these exploits are caused by a “bug” or a “flaw” and are unique to Intel products are incorrect.
and then they mention their competitors in the next paragraph to associate those names with all the exploits, even though their competitors are only affected by one.
>>
>>64131767
Online games gonna sucks
>>
>>64131765
The info about ME posted there is from before all the ME vulnerabilities were found, or people even knew what exactly the ME was able to do.
>>
>>64131800
They all affect Intel, one of them may or may not affect aMD.

Only way to patch it is to shut down speculative execution, resulting in nominal changes to "average" user performance but likely to greatly impact companies, data centres and servers as it's "workload dependent".

Basically RIP Intel
>>
File: 1514489876717.png (48 KB, 800x729)
48 KB
48 KB PNG
>>64131105
>>
>>64131814
A 286 supports arachne browser
You can use that
>>
>>64131300
>@20:58
fucking kek
>>
>>64131813
Spectre only.
>>
>>64131807
Maybe you didn't know this but hardware costs money and changing could cause all kinds of issues
>>
>>64131832
That shit would be expensive and time consuming. hey'd have to rebuild from scratch
>>
>>64131851
I really do believe that TempleOS could be much more than it currently is.
>>
>>64131344
Yes, these X% benchmarks are just averages to mask what's taking time.
Need hard data.

In worst case, think of real-time data streaming LOB services doing tons of system calls, those will be in for a surprise when amazon reboots...
>>
>>64131642
shame.
Would have been a good excuse to bring back those i686 distros to life
>>
>>64131001
Are phone posters in trouble too??
>>
>>64131847
Yes but according to the other thread Nvidia GPUs use a lot of syscalls which will suffer from this update, ergo your nvidia GPU gets fucked over too
>>
>>64131747
They have been lying a lot longer than that.
>>
>>64131795
Why are those security auditing companies so terrible compared to NSA?
>>
kek, even mads stock is dropping now.
>>
File: feelsbadman.png (74 KB, 800x600)
74 KB
74 KB PNG
>>64131600
>sitting here on my thinkpad with a third generation intel like
>>
>mfw secure browsing will only be possible via tablet or sim free phone for the next 5+ years

>mfw I have no face
>>
>>64131834

>fixed
>>
>>64131888
>installing the update
>>
>>64131858
In laymen terms, chop your arm off to save the body?
>>
>CPUs have been stagnant for years
>PCs keep getting slower since they can't keep up with the software bloat
>now this
JUST
>>
>except Intel Itanium and Intel Atom before 2013
I fucking knew it. Buying a netbook with a N270 was the best computing decision I ever made. This shit is rock solid inside and out.
>>
Project Zero's paper on the subject

https://googleprojectzero.blogspot.com
https://googleprojectzero.blogspot.com
https://googleprojectzero.blogspot.com
https://googleprojectzero.blogspot.com
>>
Would spectre work on current gen consoles? Could it be used to make piracy possible?
>>
>>64131904
Because NSA has more than likely been working hand in hand with Intel utilizing this "newly found" security flaw to spy on people. It doesn't take a shit load of technical know how to break into a building when you have the fucking keys.
>>
>>64131918
I won't, but I might as well get an AMD GPU in case I'm forced to update in the future. I'm not gonna purposefully put myself in a risky position like that
>>
File: ded.jpg (13 KB, 270x278)
13 KB
13 KB JPG
>>64131858

None affect AMD64. Stop spreading lies.
>>
>>64131807
COMPABILITY
>>
File: 1505442077045.gif (2.29 MB, 695x392)
2.29 MB
2.29 MB GIF
who else /comfy/ watching this meltdown?
>>
>>64131913
Spectre affects ARM too
>>
>>64131932
Itanium will rice from the ashes like a phoenix!
>>
>>64131866
Holy fuck, he refuses to answer!
>>
>>64131920
In laymans terms, put a lock on your front door - stops people robbing you but makes it take longer and more effort to get in and out of your house, and those going in and out a lot will be worse affected than those who only use the door occasionally.

Up to now you were safe because nobody knew your door was unlocked the whole time.
>>
mark my words, intel is working as hard as they can to find a flaw in AMD chips which they're then gonna give to "security researchers". this news is gonna come out to take heat off intel.

this issue is massive and intel will go to any lengths to reduce the impact.

just fucking watch
>>
MyCPU is safe.
www.mycpu.eu
>>64131932
Shame that the performance sucks so much
>>
>>64131182
how they will push their brainwash without tech?
>>
>>64131949
COMPATIBILITY*

fuck I can't spell
>>
>>64131951
provide proof.
This thread is full of "only ARM64 no ARM32" "AMD IS ALSO AFFECTED"
But nobody cites the fucking claims
>>
There goes the dream of ever increasing CPU performance and exponential growth even sooner then expected. It was a nice time /g/
Now we will be stuck in regression for years
>>
File: 1514607214814.png (974 KB, 900x900)
974 KB
974 KB PNG
>>64131950
it's like a train wreak in slomo
>>
>>64131869
>>64131873
Well rip

Maybe something good will come from this, the harsh cold made the vikings, as they say, or something like that.
>>
WHERE IS THE WINDOWS ME UPDATE
>>
>>64131982
See
https://spectreattack.com

Desktop, Laptop, and Cloud computers may be affected by Meltdown. More technically, every Intel processor which implements out-of-order execution is potentially affected, which is effectively every processor since 1995 (except Intel Itanium and Intel Atom before 2013). We successfully tested Meltdown on Intel processor generations released as early as 2011. Currently, we have only verified Meltdown on Intel processors. At the moment, it is unclear whether ARM and AMD processors are also affected by Meltdown.

Which systems are affected by Spectre?
Almost every system is affected by Spectre: Desktops, Laptops, Cloud Servers, as well as Smartphones. More specifically, all modern processors capable of keeping many instructions in flight are potentially vulnerable. In particular, we have verified Spectre on Intel, AMD, and ARM processors.

Which cloud providers are affected by Meltdown?
Cloud providers which use Intel CPUs and Xen PV as virtualization without having patches applied. Furthermore, cloud providers without real hardware virtualization, relying on containers that share one kernel, such as Docker, LXC, or OpenVZ are affected.

What is the difference between Meltdown and Spectre?
Meltdown breaks the mechanism that keeps applications from accessing arbitrary system memory. Consequently, applications can access system memory. Spectre tricks other applications into accessing arbitrary locations in their memory. Both attacks use side channels to obtain the information from the accessed memory location. For a more technical discussion we refer to the papers ( Meltdown and Spectre)
>>
When are chineses going to build their own microprocessors and save us from the kikes?
>>
>>64131854
People have been suspicious of ME since it was introduced. Maybe even before. That's what anon meant by "basic knowledge"
>>
File: thisisfine.png (310 KB, 580x282)
310 KB
310 KB PNG
>>64131950
This is us.
>>
>Android is affected
>carriers update at glacial pace
time to take all sensitive information off my phone i guess
>>
Well, I refuse to lose that 30% I fucking paid for.

Guess I'll just circlejerk the same websites I do now with JavaScript turned off for the next few years.
>>
I'm not too concerned of your average desktop home user but what about at enterprise level?

Major lawsuit incoming?
>>
>>64131019
(E) All of the above.
>>
Tabletfag here, am I affected?
>>
Right now I have Intel Core i5-4670K Haswell in an ASRock z87 mobo.
I have a spare AM3+ motherboard (MSI 870A-G54)

It can support an FX-8350 with a bios upgrade (i have a spare phenom 2 to upgrade it first).

Should I pull the trigger and get the FX, or save another $200 for Ryzen 5 and a new mobo?
>>
File: Intel headquarters 2020.jpg (304 KB, 1800x1200)
304 KB
304 KB JPG
>Yes our products are fucked
>But have you heard of these side-channel attacks that also fuck our products up?
>>
>>64131524
nope

>>64131576
lying faggot

>>64131724
almost, all 3 of them for AMD x86_64, arm_amd has the spectre vulnerability though.
>>
>>64132021
>lawsuit

For what? The processors are working exactly as their specs say they should. This is a feature, not a bug.

They're only patching it because that feature has a massive security risk attached, but if anyone had bothered to ask they'd have known that from the start. Because as they said, working as advertised
>>
>>64132037
Ryzen
>>
>>64131835
they always wanna act like they're the first to find out the truth.
>>
>>64131019
Since 1995.
>>
>>64131019
Anything intel branded faster than 133mhz.
Mhz. That's not a typo.
>>
>>64131019
Pentium Pro and later
>>
>>64132021
I can see it. Most times companies like Intel are "too big to sue". But when you have behemoths like Google and Amazon bringing their legal team to play, we'll be watching the court case proceedings like a prize fight.
>>
>>64131979
It's easier to hide their lies without it.
>>
>>64132053
for not disclosing the vulnerability is something that can have a case, but Intel will say that they didn't know anyway.
>>
>>64132073
>he didn't have a 233MHz Pentium 1
>>64132034
Intel CPU? Yes
>>
>>64131986
underrated

technical recession incoming
>>
>>64131413
Take that Brian & Intel shills!
>>
>>64132053
Are you retarded? A huge fucking security hole that needs a band-aid fix that decreases performance by 30% isn't something "working as intended".
>>
>>64130919
What chips are we talking about here? I don't know much about server hardware
>>
>>64132085
What if that vulnerability was forced by a Governmental agency?
>>
>>64131967
But from what I'm also reading, is that the fix for the processors dramatically cutting performance. Is this in overall data center performance, or individual workstation performance?
>>
>>64131970
AMD should hit these fucks with a libel lawsuit, based on whats in these slides.
>>
File: 1506547454349.gif (1.93 MB, 460x259)
1.93 MB
1.93 MB GIF
>apple slow down their phones
>a few weeks later intel slow down their cpus
Hmmm.
>>
>>64132098
It's actually up to 80% depending on the workload.
>>
>>64131836
What about dumbphone?
>>
>>64131836
except not AMD and ARM, but amd_arm. There's a huge difference.
>>
is ios/iphone affected by this?
>>
>>64132003
>just Sold a Mini PC with an Atom 330 for 99€
>IT will be worth 5x that by next week
>>
>>64132114
Data centers are fucked the most.
>>
>>64131816
Holy shit, is this a blessing in disguise?
>>
>>64131936
lol don't worry they don't give a shit about your vidya
>>
>>64132112
Then we nuke the country responsible off this planet
>>
File: 1515010344363.jpg (142 KB, 708x396)
142 KB
142 KB JPG
MicroShit patch just came out

Intelfags go get your 30% fix

https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892
https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892
https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892
>>
>>64131816
Javascript is generally not large in size or performance heavy for servers. It's more of a client/browser issue, performance wise.
>>
Am I getting my hopes up too far for thinking this could cause some major damage worldwide?
>>
>>64131300
Is Ronak Intels house Pajeet or something?
>>
File: Loud_Howard_Drunk.png (144 KB, 270x350)
144 KB
144 KB PNG
>>64132118
>80%
>mfw
>>
>>64132112
That would be so fucking good, if Intel would just flat out admit that US government forced it.
>>
>>64132117
Gotta slow them down so you have some reason to upgrade. Websites can only stress your CPU so much
>>
>>64132150
where's the W7 fix
>>
>>64132150
I thought they were rolling out the patches on next Tuesday? What does these micro patches do?
>>
>>64132114
As an layman my understanding is that this speculative processing thing is what made Intel processors so good, because they select data ahead of it actually being needed or something

This is what gives people access and so it needs to be shut off. That's the "patch".

Without this the processors need to wait to get an instruction before processing data (or however it works) meaning they now are less efficient and performance suffers. It's apparently workload-based, so PCs will barely notice it, but data centres and servers will be fucked as it fucks their performance at every single operation.
>>
Quick rundown

There are three bugs
>Spectre version 1:
Affects everyone but easily patchable (no perf loss)
>Spectre version 2:
Affects everyone, not patchable but hard to execute in a real world environment and can be mitigated heavily with patches

>Meltdown
Intel only. Patches fuck perf hard in some usecases. Not relevent to AMD/ARM
>>
>>64131932
>except Intel Itanium and Intel Atom before 2013
Price gonna rise up for these
>>
Is Itanium affected? What about PowerPC? Sparc?
>>
>>64132145
I don't see anything stopping the people trying to crack consoles from using this exploit if it works and gives them access to the console
>>
>>64132152
Not at all. This could be used to trigger a war.
>>
>>64132151
Don't listen to them. They are shilling against opensource (JavaScript most always presents us the ability to examine the code, since it is largely a client-side language).
>>
>>64132202
nobody knows, because nobody cared enough to check
>>
>>64132213
Hint,hint.
>>
>>64132003
>https://googleprojectzero.blogspot.com
I read the google paper and they say that AMD is only vulnerable to one attack, and specifically: "If the kernel's BPF JIT is enabled (non-default configuration), it also works on the AMD PRO CPU."

So it's not even a ryzen cpu, and it looks like a simple fix in any case.
>>
File: hqdefault-4.jpg (26 KB, 480x360)
26 KB
26 KB JPG
Is it safe
>>
>>64131169
You must not be reading this thread (or thinking clearly for that matter). The big issue everyone has been talking about recently has been an issue specific to Intel CPUs. Intel, slinging mud, has exposed a flaw that every modern day processor has, including AMD and ARM processors.

Though the Intel specific problem sucks, this new flaw has the opportunity to btfo this entire planet.

Read >>64130838
and https://googleprojectzero.blogspot.com/2018/01/reading-privileged-memory-with-side.html
>>
So as long as I don't visit websites, I'm safe, right?
>>
>>64132244
Was it made after 1993?

If yes, then no
>>
will this hurt my gayman fps??
>>
>>64132191
So the CPU was predicting what data would be used and had it on standby for if/when it would be used, but this in and of itself is the vulnerability and now taking it away will cripple performance.

It was kinda like Intel Optane but instead of it being for your Disk Drive, its for your CPU
>>
>>64132208
oh, i think i misunderstood your comment. yeah, if you could fiddle with memory, you could, theoretically, crack consoles without modchips.
>>
>>64131342
Don't make statements without providing source or your opinion is immediately discarded.
>>
>>64132202
Itanium no because no speculative execution. Same as pre 2013 atom.
>>
>>64132213
Now you're just trying to make me happy
>>
>>64131982
>>https://archive.4plebs.org/pol/thread/117886401/
https://spectreattack.com/spectre.pdf

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
>>
>>64132240
Yes, the spectre bug is much less serious short term since it's hard to pull off, but it can only be mitigated not fixed without hardware redesigns.
>>
File: LUL.png (21 KB, 112x112)
21 KB
21 KB PNG
>>64132240
>shintel is falling like Hillary did in the election
>>
>>64132257
only by 30-65%, no biggie
>>
File: int.jpg (70 KB, 898x628)
70 KB
70 KB JPG
>>64132112
S-shut it down!
>>
>>64132258
>So the CPU was predicting what data would be used and had it on standby for if/when it would be used, but this in and of itself is the vulnerability and now taking it away will cripple performance.

Exactly

Technically all it's doing is putting Intel back on a level playing field with everyone else. Problem is that we've grown accustomed to the superior world that Intel provided and going backwards will suck donkey dong
>>
>>64131413
> However, for both ARM and AMD, the toy
example as described in Section 3 works reliably, indi-
cating that out-of-order execution generally occurs and
instructions past illegal memory accesses are also per-
formed.

The shill thinks we don't read sources. Ha, fuck off.
>>
Is this on the hardware level or what?
>>
>>64131001
googleprojectzero blogspot co uk/2018/01/reading-privileged-memory-with-side html
>>
>>64132129
see >>64132003
>>
>>64132255
It is not x86 (ia64)
>>
>>64132044
That is a picture from liberated Aleppo actually
>>
>>64132313
Their attempt to actually leak memory failed
>>
>>64132244
Maybe. Nobody's sure.
>>
>all I do is play games on my PC and deal with low security tasks
>updating OS will patch out performance
>no benefit to me
wtf
>>
>>64132323
Hardware architecture design level
>>
>>64132176
Must have been severe for them to release it asap.
>>
>>64132323
Ring 0. The same level that a malicious bit of code executed by your CPU (that you can't detect in any way) can overwrite your entire motherboard BIOS with all 0's if it wants and permanently brick your PC.

The problem is just Ring 0 access. It's the fact that a rogue Java script has the ability to drop a payload that can utilize the exploit. All from Ring level 3.
>>
>>64131836
>>64132003
>except Intel Itanium and Intel Atom before 2013
Gonna stock up on those.
Seems like the libreboot thinkpads are dead.

>all modern processors capable of keeping many instructions in flight
Does it refer to OoO only ? Does it have a technical name ?
>>
>>64132323
read anything in the thread?
>>
>>64132232
VIA decided to go for the security by obscurity approach I see.
>>
>>64131681
except consoles run AMD apu's and Nvidia tegra based chips.
>>
>>64132401
>Seems like the libreboot thinkpads are dead.
My T60 and X60s are likely going to be fucked up very badly by this.
>>
>>64132309
I've also heard that this is some CIA psi-op that intel made to give them backdoor into people's hardware to around the OS.

is AMD guilty of this as well?
>>
>>64132401
God i hope not some one needs to benchmark a libreebooted system with the fix
>>
THAT DIDN'T TAKE LONG

>INTC INVESTOR ALERT: Law Offices of Howard G. Smith Commences Investigation on Behalf of Intel Corporation Investors

https://www.businesswire.com/news/home/20180103006309/en/
>>
>>64132381
>>64132390

shit, no fixing it then

>>64132405
I only read the headline, chill out dickshitter
>>
How do I protect myself against this exploit without this shitty patch? Would disabling JavaScript do it? Would most websites work without JavaScript?
>>
I can't wait until the inevitable JS-based exploit that somehow fucks your whole kernel right up.
>>
>>64132438
AMD is only marginally affected by one of three bugs. Intel is affected by three, including the most serious.
>>
>>64132462
You can, patch.
or you can wish nothing happens.
>>
>>64132448
Did you listen to the conference call? They lied straight to them and then refused to answer any of their questions or concerns
>>
>>64132466
i'm sure such a thing will eventually turn up, but that is not what is happening here
>>
>>64132466
If privileges ever meant anything that shouldn't be possible. But due to this hardware level fuckery, who knows.
>>
>>64132390
>a rogue Java script
Java is not a scripting language.
>>
>>64132462
Completely isolating your computer. Have fun offline, and better not use external storage mediums either.
>>
>>64131469
Remember the 6 billion Hz, goy. OY VEY
>>
>>64132462
Block all ads, disable JS globally.

If the site needs JS, enable it for that time only (like to post a comment on 4chan), then disable it after.

And don't visit shady websites.
>>
>>64132293
>only that 60 fps will go to 40 or 21 fps, no biggie
>>
Good thing I just bought a itanium 9700 series based server in oktober.
>>
>>64132509
You know what I'm getting at. Java based websites. I.e. more than we care to admit.
>>
>>64132416
Have you tested and made sure the exploit doesn't work?
>>
>>64132492
>>64132518
>>64132526
Fuck it I'm running CommonSense™ 2017 and it hasn't let me down yet, I'll take my chances.
>>
File: 211c16bfhb001.jpg (52 KB, 550x481)
52 KB
52 KB JPG
what the flippin frigg is going on
this looks bad and i'm worried
>>
>>64132244
Probably yes. Definitely yes if you have earlier than Poulson.
IIRC Itanium didn't even have out-of-order execution before Poulson.
>>
>>64132540
The vast majority of websites do not require Java. In fact I don't even think Firefox supports Java out of the box anymore.
>>
File: 1488276778500.jpg (35 KB, 600x520)
35 KB
35 KB JPG
So all shIntel was doing this whole time was CHEATING their performance by intentionally using a security flaw as boost

really, REALLY makes me think
>>
>>64132191
This couldn't be further from the truth. Why are you even on /g/? Holy shit what is happening to this board?
>>
>>64132570
That's literally true.
>>
>>64132547
Your PC hasn't had an easily exploitable flaw like this until now either. You don't even need to browse the wrong sites to get fucked by this.
>>
>>64132585
Someone linked a scary thing on /tv/ and now I'm here
>>
Who the fuck is paying for fixing this mess?
>>
Is SPARC affected ? Older ones use in-order but due being a barrel processor it keeps instructions decoded.
>>
>>64132596
Literally as in the actual meaning of the word?
>>
>>64131001
Obviously untrue. Project Zero executed variant 1 on an FX
>>
>>64132626
i don't know but it better not be the government. i had a thought that trump has something to do with this because maybe obama used this backdoor to spy on him
>>
>>64132626
You.
>>
>>64132613
Which is why I said block ads and disable JavaScript globally.
>>
>>64132635
Yes.

>>64132632
No.
>>
>>64131794
>5%
>done
I mean, I hope they are. But call me when it's double digits.
>>
>>64132642
>FX
Lol
Also seems to be patchable by software with no perf hit.
>>
>>64132614
Then go back there. What you describe is called prefetching and something every high performance CPU does since like three decades. And Intel's CPUs still do it even with the patch. Performance would be fucking abysmal otherwise. What the patch does is separating the page table between user and kernel space, now you got two the CPU has to switch between. That causes the performance hit.
>>
File: 1502474562221.jpg (69 KB, 775x837)
69 KB
69 KB JPG
>Attacks using JavaScript.

>In addition to violating pro-cess isolation boundaries using native code, Spectre at-tacks can also be used to violate browser sandboxing, by mounting them via portable JavaScript code. We wrote a JavaScript program that successfully reads data from the address space of the browser process running it
>>
>>64132655
Yeah, your browser is the only attack vector, you got it. Scriptkiddies will have a lot of fun with people like you once the meltdown toolkits come out.
>>
>>64132613
>Your PC hasn't had an easily exploitable flaw like this until now either
But it literally has for the past 20 years
>>
>>64132645
dumbass
>>
>>64132703
Wake me up
>>
If this kills javascript what alternatives do we have
>>
>>64132719
not an argument
>>
>>64132711
Pray tell then, how else can people be infected?
>>
>>64132716
Which doesn't matter since nobody knew about it. Now they found out, the documentation is getting released and soon there will be toolkits around so everyone that wants to try it can exploit the flaw.
>>
>>64132703
STOP BROWSING THE INTERNET NOW
>>
>>64132655
How does one disable JS globally?
>>
>>64132626
Google, linux kernel devs from red hat, intel, amazon, students from multiple universities. Microsoft. Everyone.
>>
>>64132763
MY HAND'S TOO SWEATY TO PULL THE PLUG
>>
>>64132746
Everything that connects your PC to other systems. Mainly mailing clients and external memory
>>
>torrenting is now impossible without heavy CPU load
>encryption is now impossible without heavy CPU load

THANKS BASED INTEL!
>>
>>64132771
uMatrix or noScript.
>>
>>64132800
>Mainly mailing clients and external memory

Not sure you mean by the last one, but I don't use a mailing client.
>>
>>64132759
How do you know someone hasn't been exploiting it for years now without it becoming public knowledge?
>>
>>64131815
Spectre cannot be fixed. Entirely new processor designs are needed.
>>
>>64132771
>>64132703
Can I get some sort of chrome/firefox extension to toggle it?
>>
>>64132642
Variant 1 is fixable.
It's 2 that affects all chips and can't be fixed.
>>
You can be attacked by parsing CSS, not only running JS. You can't really escape.
>>
File: 1513008209986.jpg (368 KB, 840x700)
368 KB
368 KB JPG
I hope this won't count as boycotting an Israeli company.
>>
>>64132342
That is a photoshoot that a journalist put together so he could try for one of those "award winning photo" awards

It's not genuine, the entire thing is staged
>>
>>64132847
>>64132822
>>
>>64132771
If you have to ask that, then I suggest you don't. You won't know how to fix broken websites once you disable it.
>>
This is trumps fault
>>
File: 1513467728749.jpg (402 KB, 1536x2048)
402 KB
402 KB JPG
>>64132811
Works on my machine desu.
>>
Lads I'm on a pirated Win7 SP2 that I never updated. Is there a way to just get the patch that Microsoft are releasing without having to install a load of updates? I wanna avoid those telemetry updates they added a while back.
>>
>http://ieeexplore.ieee.org/document/466952/
>Date of Conference: 27-30 June 1995
>The CPU of the multi-chip module processor has a superscalar, speculative issue unit, and an out-of-order execution datapath.
Well fuck
>>
File: 64553024_p0.png (86 KB, 700x988)
86 KB
86 KB PNG
Retroware fags were right again!
>>
>>64132842
Unlikely but possible. We don't know, but if it happened it wasn't widespread
>>
>>64132847
see >>64132816
>>
>>64132872
At what CPU performance?
>>
>>64132861
Why not?
>>
>>64132390
>Next generation ILOVEYOU virus
>>
>>64132861
my fucking sides, it probably does
burgers need to get the fuck off Israel
>>
>>64132860
This is true.
>>
>>64132887
Time to break out those PA-RISC workstations out of the closet.
>>
>>64132860
what the fuck
>>
>>64132866
Didn't know there were such things for pictures that don't shit on the Syrian government
>>
>>64132914
Burgers can't stop and won't stop funding Israel
>>
File: 86f2c88b80.png (88 KB, 1150x705)
88 KB
88 KB PNG
intel dug so hard they found one amd cpu that amd themselves must've forgotten existed.

bulldozer mobile, bga socket
>>
>>64132861
Buy israeli or else, goy
>>
>https://en.wikipedia.org/wiki/Comparison_of_CPU_microarchitectures
We're all fucked bois
>>
File: ThinkingSpinner.gif (240 KB, 320x320)
240 KB
240 KB GIF
>>64132842
Maybe this is how China got ahold of the F-35 and other Lockheed Martin datasets without triggering alarms immediately.
>>
>>64132860
I guess I have no choice but to update, then. In one fell swoop, my Nvidia card as been rendered useless.

I can't even comprehend this pain.
>>
>>64131019
everything since out of order execution was added, because CPU does not check if the predicted memory area from branch prediction belongs to the process that invoked it. Hence, it allows to read that memory before flushing it.
>>
>using password manager means nothing now
rip
>>
File: 1.png (58 KB, 819x494)
58 KB
58 KB PNG
Looks like Itanium, Pentium Pro and Cyrix are also fucked.
>>
>>64132671
>>64132852
Thank you for the clarification. Are there any benchmarks that have been conducted on meltdown patches for regular consumers? I knew I shouldn't have gone fucking intel this year.
>>
Anons, time to make /g/pus. Let's set this shit up. Let's take back the free world.
>>
>>64132878
Did HAL get it right, though?
>>
I don't see how Intel is gonna survive this desu, they are really fucked
>>
>>64133070
Let's make the logo!
>>
>>64133070
Let's face it, if there will ever be a crowd sourced CPU it will be made by /a/ because its required to get that bootleg Chinese cartoon sharing symposium back up. /g/ is not only to incompetent but also has no reason to do so.
>>
>>64133044
sauce?
>>
File: 1513725245727.jpg (33 KB, 604x526)
33 KB
33 KB JPG
>all these retards using effect instead of affect
lrn2english
>>
>>64133139
That's a screenshot from wikipedia.
>>
>>64133139
>>64132984
>>
At this point just burn everyrhing and start rebuilding.
>>
File: Jewish-trick free.jpg (48 KB, 580x347)
48 KB
48 KB JPG
literally /ourguys/
>>
AND SO IT BEGINS
https://www.youtube.com/watch?v=bReA1dvGJ6Y
>>
File: A humble suggestion.jpg (42 KB, 599x708)
42 KB
42 KB JPG
It's out
https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892
>mfw it'll be tommorow before benchmarks are out
>>
>>64133221
>uses speculative execution
No
>>
File: jesus2.png (3 KB, 638x399)
3 KB
3 KB PNG
>>64130858
Temple OS runs everything is ring 0 so everything can read and manipulate all the memory anyways. God told Terry this would happen!
>>
>>64133263
I need 8.1 or 7 patch
>>
I don't care about this cyber"security" bullshit. I have Kensington locks, it doesn't matter to me.
>>
>>64132963
that's incorrect, what they fx pro is bristol ridge (replying to myself)
>>
>>64133300
Ok memes have gone too far
>>
Please tell me the Z80 is safe.
>>
>>64133469
It’s not
>>
>>64133205
Sometimes you get something even better this way
>>
but i don't want to take an fps hit in my gayman :(
>>
>>64133263
How long does it normally take it to actually show up in Windows Update?
>>
>>64133602
The security update is set to automatically roll out to all Windows 10 desktops today at 5 PM ET, though we’re not seeing it on any of our systems just yet. It’s also set to become available on other Windows 7 and Windows 8 systems, though it will not appear automatically on these systems until Tuesday, January 16.
>>
>>64133469
You'll have to patch your TRS-80 but the Sega Genesis and Master System should be fine unless you're using flashcarts.
>>
>>64133635
Sauce
https://www.onmsft.com/news/microsoft-to-release-emergency-windows-update-to-address-intel-amd-processors-security-concerns
>>
>>64133635
it's 7:46 PM NY time dingdong
>>
>>64131660
If you "just" got it, maybe you can still return it
>>
File: deviance.jpg (15 KB, 480x255)
15 KB
15 KB JPG
>>64131659
100%
>>
>>64130690
holy fuck holy FUCK
https://github.com/Eugnis/spectre-attack
>>
How the fuck do I update pirated Win7
>>
>>64131659
>>64131729
During the CIA leaks, it was found the CIA had found vulnerabilities in AES256 and hadn't bothered to tell anyone about it.
>>
>>64133854

just buy a key from pajeet for 2 dollars, still been using mine since forever.
>>
>>64133854
Pirated as in you use some edited iso downloaded from torrent or clean windows iso with daz loader?
>>
>>64133635
Well, it's been three hours and I don't have it. I assume the "roll out" is gradual then. Hopefully it's here in the morning so I can get to benchmarking.
>>
>>64133854
I installed The Crew when it was free on Uplay and that malware updated my win7 and made it non genuine again so try that.
>>
>>64133914
>I assume the "roll out" is gradual then
Yes.
>>
>>64133854
>https://www.ebay.com/sch/i.html?_nkw=windows+7+key&_sop=15

No one is too much of a poorfag for this
>>
>>64133917
That's not how that works.

>>64133914
Also, forgot to ask, anyone know a formula for excel that'll print out the inverse of what I put in? i.e. 10 to 9 outputs 10%
>>
>>64133973
I'm not even joking. The pc restarted during installation and some kb(bunch of nunbers) thing was downloading and installing.
>>
File: 1501937218780.jpg (112 KB, 1152x864)
112 KB
112 KB JPG
>>64131457
>>64131367
>>
>>64133897
Clean iso with some sort of activator I think
>>64133884
Do I have to reinstall to do this? I really don't wanna reinstall.
>>
>http://elbrus2k.wikidot.com/elbrus-2000
>Speculative execution explicitly program controlled
Does this mean we can turn it off ?




Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.