Intel Makes Public Two More Data Leakage Disclosures

[SilverBird775]

You mean 10 years of telling their QA engineers to get out of the way because zomg performance!111!!!

The same haste development can attribute to software department in general, there is more features and more bugs that cannot be sufficiently explained as features. The internal bugzillas are drowning with critical errors and scratch feature implementations yet the products get sold to fund the completion. Of the marketing sin the "rolling release" was invented. Later on this creative model was extended with subscription. How beautiful it is, to pay forever for a never ending product!

The later news are giving an impression about hardware as a new subscription model to ever unfinished product.

[ferry]

I wonder how many of these recent flaws intel knew about way back in the design stage, and made the conscious decision to choose performance over security? I'd wager most or maybe all. Honest mistakes do happen from time to time, the original P5 FDIV bug, for example. But when fixing the "problem" creates a significant performance hit, Intel's faulty design seems intentional.

Nah, not intentional. These are side channel attacks that at the time nobody knew about. That's why you find them in other processors as well. But once you consider side channel attacks it's relatively easier to find more of them. Intel just has been pushing harder to optimize their hardware using technologies that are leaking data through side channels and now they get hit harder. Ask yourself why AMD choose to waste power and performance at the time? Did they know about the side channels? Or were they just lacking tech?

Fixing the problem through microcode updates replaces certain hardware by software, this slows the execution by itself, on top of other effects of partially undoing the optimization. Fortunately Intel will provide you with the opportunity to buy a new processor with fixed hardware :-)

[NotMine999]

Quoting Intel itself, "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)".

I am waiting for the day that Michael tests a new "latest & greatest, best thing since sliced bread" Intel CPUs with all of it's associated "security fixes" and finds that it's performance is rivaled or even beaten by one of those old "in order execution" Atom CPUs from Intel.

[mulenmar]

I think we will consider the P4 as a competitive CPU again soon.

P3 outperformed the P4 clock for clock and in watt/heat, so I'm glad I still have a Compaq business-class machine with a 1GHz P3 mostly maxed out (for its tech era, anyway) in storage.

Just needs some SSDs and a PCI slot SATA controller, maybe a better AGP GPU, and it'll be ready to roar!

Oh, right, web designers destroyed the Internet's "graceful fallback" guarantees.

Fortunately Intel will provide you with the opportunity to buy a new processor with fixed hardware :-)

And fortunately, I will not indulge in the consumerism they so desperately crave. I suggest everyone else refuse to buy Intel, too.

Get a TALOS machine if you can afford it, or some RISC-V or completely-FLOSS-support ARM machine if you can't. Heck, even one of those PowerPC machines the AmigaOS fans droll over despite the CPUs being intended for medical scanners and the embedded like.

Or get something older, if you don't care about the electricity bill and the associated ecodamage.

[nuetzel]

Two (???) steps forward, one step back...

[CochainComplex]

I doubt it, any QA in both software and hardware development (and in actual manufacturing) is going to find so. much. crap. regardless of how good your designers or production process are.

In cases like this, if your QA does not find a lot of stuff it's probably bad at its job. We aren't talking of basic easy dumb things with large tolerances like say sledgehammers, this is extremely complex electronics, making small mistakes is going to affect the end product in significant ways.

usually i would agree in a honest world it should be like this but you might have a look at the dieselgate case. the decision if a product is going to be released is done on the upper levels. it wouldnt be the first time in recent history that a product hit the market despite of severe protest by dev engineers and qa. (even if human lives are at stake)

[starshipeleven]

usually i would agree in a honest world it should be like this but you might have a look at the dieselgate case. the decision if a product is going to be released is done on the upper levels. it wouldnt be the first time in recent history that a product hit the market despite of severe protest by dev engineers and qa. (even if human lives are at stake)

I'm not sure how you can disagree, as I just said that QA is going to find issues if they are doing their job.

Intel has just decided to ignore QA, this is not the first and will not be the last time that QA gets overridden by management.

[Guest]

Quoting Intel itself, "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)".

Potentially, and then you they (and others) have to analyze and figure out exactly what is affected.

[Guest]

Yes, but many of those vulnerabilities do not affect AMD's x86 processors. That's why the "Swiss cheese" analogy is not accurately applicable to x86.

Yes, they just lack basic temperature and sensor reporting, their RNG sends back -1 (a bug that already happened 5 years earlier and was never fixed) and their software is buggy and broken.

Even with all of the security vulnerabilites, Intel is still the better choice, because of their professional and high quality software support.

AMD could easily change that, but they'd rather continue with their current and shitty software strategy.

[Guest]

completely-FLOSS-support ARM machine

Doesn't exist. Especially for running a GUI, no ARM based device is suitable. That Talos machine sounds good though, and pretty promising.