This is the most recent computer not vulnerable to Spectre. It is 16 years old, and has a 900MHz processor

Yep

>t. never read the exploit whitepaper
Sure thing buddy, skiddies are now hackermen all they have to do is write shit js and hack the planet
i'm starting to believe spectre and meltdown are a jew plot to force you to buy new cpu's

Attached: 1543353347670.jpg (267x323, 7K)

>Spectre
Has that ever been exploited in the wild for any practical attack? I doubt it.

Zen2 and Ice Lake (or was it Cannon Lake? Fuck Intel and their lakes) are supposed to have hardware mitigations for it, but nothing has been completely solved.

The Pi doesn't use speculative execution, so it cannot be affected by these attacks. There are ARM chips that can be affected, but not the ones in the Pi.
Basically, it literally lucked out because the CPUs in the Pi are cheap and slow.

Could you at least read what Spectre is and widen your view on computers before posting dumb shit like this?

what is power9?

risque v when?

I don't care about Spectre.
The chances that anybody will be able to break the software/firmware patches protecting against Spectre, and then both actually steal anything from me, a literal nobody, and that the information will be valuable, are so remote, I will gladly expose myself to the possibility for the convenience of modern processing power. Furthermore, anybody in similar circumstances (having patches and being an uninteresting target) who wouldn't is likely a paranoid.

they won't steal the info from you personally, they'll steal it in a batch from information holders like banks/online stores

And it smells like my armpits, fuck that shit

What about the A53 core? It's in-order, doesn't that mean it doesn't do speculative execution? That would make it immune.

Attached: gunbuster_explosion.webm (640x480, 506K)

That's a fair point, but it doesn't address the thread we're in right now, which is clearly geared for end-user security, or the numerous threads we've had before where people are looking for personal computing devices that are Spectre-immune.
Plus, I have no control over these businesses such as banks which have become borderline essential for modern life. Hence, worrying about it isn't productive anyway.

as far as i know this can only read data stored in variables which were deleted afterwards, and only if they were in the same scope as the actual program. it can't do shit for any other things

Raspberry Pi uses ARM so actually it would be the most recent computer not vulnerable to Spectre.

Atom N270 isn’t vulnerable to speculative execution attacks because it’s an in order processor.
Intel was hilariously cheap when they designed that dumpster fire.

And it will inevitably die of sudden GPU failure, as well. coincidentally, my 900 MHz iBook G3 just died today. I shimmed the GPU to get it working again, but it will inevitably fail once more.

BTW, the iBook pictured is not a 900 MHz model. It's actually a 2001 500 MHz model.

Why are Spectre threads always just a bunch of dumbasses listing literally every non-x86 processor/computer they've ever heard of once?

Affected by specter?

POWER9 BRO

It's equally amazing how little things have changed in the last 5 years

AMD is not impacted by Spectre or Meltdown at all unless you count having physical access

Holy shit you are autistic OP if this is actually your laptop. Go buy a normal PC.

What causes the failure?

Shitty BGA soldering job. All iBook G3's, aside from the toilet seat ones, were affected.

Fuck. This explains why the guy I sold mine to emailed me several months later demanding a refund. This was 10 years ago, and I never responded, but holy shit was he pissed off, and he accused me of ripping him off. I never understood why until now.

It's not your fault. Not sure if it was on Apple, or AMD, but the problem persisted up through the early iBook G4 series. I don't believe the powerbook line was ever affected.

It was one of those early iBooks. I genuinely had no idea but I dodged a bullet apparently. I even tried to sell it on eBay the first time, but the buyer backed out. Ended up selling it on the local classifieds site instead.

>he puts his actual credit/debit cards online
What a dumbfuck. You deserve whatever happens to you. Should've used pre-paid gift cards.

Raspberry Pi processors do SISD (single instruction, single data) so they aren't vulnerable.

>POWER
>ARM
>RISC-V
>not a mention of SPARC with CPUs only a couple years old that are not vulnerable to Spectre
Jow Forums really is something, specially turning US hours

looks COMFY af user

Meltdown doesnt matter

>Ahem

Attached: 51LmoKyKoLL.jpg (500x500, 63K)

Shilling for AMD doesn't make you any better than shilling for Intel, faggot.
amd.com/en/corporate/security-updates

Ever actually used a T-series UltraSPARC machine at home? They're shit. Fascinating and fun to work with, but shit, and useless as desktops.

However they are still worthless garbage toys, I'd honestly still rather have an iShit even if the benchmarks aren't as good.

Wasn't this era of iBooks hideously unreliable?

Why bother with something literally worse than a "toy"?

Raspberry Pis are immune to Spectre. 4 1.4 GHz cores in the 3B+.

Yes, I was browsing an archive of apple's website from the early 2000s, and I saw a link for an iBook logic repair program extension.

Nobody is going to use a RISC-V or 15 year old PPC laptop either.
SPARC is open though and here right now. Modern "open" POWER and ARM boards even aren't.

I use an IBM Thinkpad X23 running OpenBSD.

Attached: 031_017.jpg (720x488, 63K)

>what is site isolation

I have a desktop 1200 mhz AMD that not on the list. It was made around 2000-2001 I think, so 19 years ago. Why post the probably the slowest processor from three years later?

>Nobody is going to use a RISC-V or 15 year old PPC laptop either.
Nope, but those that do will still get better use out of them than a T-series chip that can't even outrun a fuckin' cheap old ARM9 SoC on single-thread integer.
>SPARC is open though and here right now.
Unfortunately SPARC is literally dead as Oracle finally beheaded the shambling corpse of Sun's hardware division a few years ago, and none of the newer out-of-order models are invulnerable anyway.

Yeah, this thread is full of pseudo-intellectuals crawling all over each other to list everything Intel doesn't make to look woke as always, and you could probably get some mileage out of a T2+ system with a proper setup, but at the same time it's not really surprising people don't usually bring them up. They're just not built for the kind of work we do on a desktop, and recent desktop SPARC systems are hideously overpriced and underpowered.

>muh megahurts

>implying people need to use the latest and greatest computers to be safe on the Internet
You can use ancient computers for most Internet tasks, I have a Transmeta running OpenBSD, come hack me.

Vault 7 proves that security through obscurity is actually a viable strategy.

>Unfortunately SPARC is literally dead as Oracle finally beheaded the shambling corpse of Sun's hardware division a few years ago
SPARC is open you dip, that's literally the point. It's already here, unlike RISC-V, just make your own hardware, what the RISC-V tards are trying to do and failing.

>SPARC is OPEN!
nobody's sliced one open and modeled the entire circuit, to see if there's anything secreted away inside

wait you don't believe a company which has likely received a national security letter do you?

RISC-V is at least coming along, it's hardly 'failing' either

you're a shill, it's easy to tell because of your style, you're not a good one either, who do you work for? NSA?

>JustIntel®Problems

I just said the SPARC T-series performance is really shit compared to other options, most of which are also old, and that's from experience.
I see what you mean now, but I still don't see how that makes up for its deficiencies. Unless you're trying to secure some kind of integer-heavy embarrassingly parallel system to a ludicrous degree, SPARC systems are still pretty sub-optimal.

>you're not able to do that on the 16 year old computer either
I'm not sure if you're just dumb or severely inept.

If you're not doing Modern Web Javascript anything made after 1997 or so is fine.

techarp.com/guides/complete-meltdown-spectre-cpu-list/2/
I looks like its on the list, user.

You're face.

Nobody wants to steal you're pedophile weeaboo shit, dumbfuck

Umm...user...most modern CPUs are patched against Spectre

that's just not true, the automated hammering on every open port by bots for the last 20 years is the proof

everything is to be hacked dummy

Use a firewall then like literally everyone else instead of wasting time and money on hipster ISAs.

i do but the point is that everybody is always hacking everything, telling a person nobody's trying to hack them is bullshit

No, SPARC is open, as in literally GNU licence. You can get all the schematics for the chips, how to make them and related firmware for free right now and produce or modify for your own production. POWER has a similar open intivive, but not *as* open IIRC.

>computer
that's ishit on your picture

Oh zoomers.

>SPARC is open, as in literally GNU licence
yeah but that doesn't magically guarantee that there isn't hidden circuitry on the chip

You can go through all the material if you think there is. If there was though, I'm sure someone would already be talking, as the material is available for anyone.
I highly doubt you'd even be capable of hiring a fab to even make you SPARC silicon if you're this dumb though or even make a softcore.

I'm just going to assume there is and that "the material" isn't the exact "material" that gets etched into silicon

Well, it's not far fetched that someone would create SPARC CPUs and embed their own shady shit in in. But then same could be said for RISC-V production silicon.
That's the same as in compiling a program from source you can go through yourself and see before using the program, against using a pre compile binary from someone else.

Plus:
>and that "the material" isn't the exact "material" that gets etched into silicon
Unless you think the fab adds something themselves, which would probably be actually easily spottable on the long run, the material you give to a fab to create the chip, is what you get.

He's right tho, the ibook g3 is a piece of shit. No TLS 1.2 support and 128mb of ram makes it indeed incapable of browsing the web. And don't even try to tell me that you can install linux on this shit because it's simply not true, debian, mint and arch all have install issues and sever graphics glitches and don't even get me started on the pkg repos. Trying to use your g3 as a daily driver is completely pointless and only done for the sole purpose of saying "look I use this old g3 as my daily driver, am I cool yet?". No you're not, you can't even wifi with this shit.

That's why I don't trust processors made too long after 9/11 when they passed the secret laws requiring spyware on the chips. Also it wasn't feasible back then to put backdoors in anyway.

Nobody cares

>easily spottable
only if you plane off the cpu one molecule at a time and reconstruct the actual circuits you photograph

>Nobody cares
>thread has 81 replies

I have non-x86 workstations from the early (before mid) 2000's that are capable of even playing YouTube.

There's more to computers from 16 (or more) years ago than G3.

Of people saying it doesn't matter

You're going to run debug routines on your brand new chip you had a fab, fab you. We are talking about very low level things here, CPUs without even firmware onboard while you develop them.
Plus yes, you're probably going to go over the silicon with a microscope too.

>implying the fabs aren't in on it
>implying they even know what they have
think like a glownigger nigger

techspot.com/news/78917-google-researchers-software-patches-never-fully-protect-against.html

I thought that was the whole point of your argument. You realize you do the debugging and development, not the fab, when you hire them to fab your designs?

that's why smt is turned off in openbsd now and they flush every cache every time

so you get a huge verilog file and then what

manually read every line?

Basically, more likely with a team though. If we talk about SPARC, the ones that modified it or if from the ground up, you should already know everything.
Your verilog file doesn't really matter, if anons argument was that fabs can secretly embed things into chips you hire them to make based on your own designs you already know are clean.

That I still highly doubt though, imagine the backslash, plus it's hard for a fab to later deny if found out after you ordered tens or hundreds of thousands of chips from them, think major financial loss for them after something like that comes out.
Companies who create CPUs have it much easier, since they can masquerade backdoors as other things, like IME.

No they don't know everything. You're talking hundreds of millions of transistors and probably a mile of interconnects in a modern processor.

The only true trust is making the CPU yourself or relying on processors from before the government mandated backdoors. Everything else is a meme.

Come home.

Attached: S_Zilog-Z84C0006PEC.jpg (283x176, 18K)

it was, but i used it when it was still relevant

>The only true trust is making the CPU yourself
That's what I'm talking about. But you have to fab them somewhere, anons argument, to my understanding, was that fabs could put something in it without your knowing. You realize even when you'd use something like open SPARC, you're still going to design the chip yourself for fab, you're just basing it on a ISA that's already made and that material is not that big. Laying down things like those millions of transistors is done while designing the layout, the material just says how many of what kind and where their I/O goes.
Plus there's no way of telling what CPUs from the past could have backdoors. Even before government mandation.

device makers already encrypt and obfuscate SoCs and FPGAs to stop the chinks from sneaking shit in, unless you own the fab you're not at full trust

to trust the cpu you have to design and build it yourself like the russians and chinese and japanese and even germans do

Bullshit. I had a 1Ghz+ iBook G4.

That's what I've been talking about.

No, he's actually wrong, though. The Debian installer works just fine now, and even if it didn't, you could use a Debian Jessie image, and upgraded it to Sid after install. As for the graphics problems on G3 iBooks, I won't dispute that. However, you can avoid that by just running it in radeonfb mode. Not perfect, but it does work. Just did it the other day no problems.