USB 3.0 problems for Intel's Haswell

[ShintaiDK]

Nothing better than random rumour threads that run their own course.

 

[Abwx]

Have you actualy read the article before downplaying it as a rumour
because it doesnt suit your usual Intel magnificencing?..

They say that they did read in official docs...

"In the document, Intel purportedly informs manufacturers that

 

[OBLAMA2009]

BTW: Doesn't anyone find it strange that Intel considers it ok to release Haswell with such a hardware bug?

no i remember when i built my pentium 3 it had a bug that wouldnt let it turn off. they dont have very high standards for the first batch at all

 

[ShintaiDK]

Have you actualy read the article before downplaying it as a rumour
because it doesnt suit your usual Intel magnificencing?..

They say that they did read in official docs...

Sure, we never got fooled by such wording before.

/sarcasm.

And the information in the suppsoed "document" points to its bogus.

 

[KompuKare]

This could cost Intel 10's of millions of dollars. Maybe 100's of millions of dollars. 10's of 100's of millions of dollars. 100's of 10's of millions of dollars.

True but unlike certain other companies (hint: ~540nm), Intel does stand by their products when they have problems, even if it costs them serious amounts of money. (Although with SB-E and Intel VT-d they were slightly evasive but AFAIR they did offer replacements for those who asked.)

But something like this (if true) does show that even a company which has such a massive R&D can/will have issues. On these forums I sometimes think that the halo effect of top-end IB and SB-E performance makes people think that Intel is perfect (or has access to alien technology), but while their products perform really well smaller players seem to get better value for what R&D budget they do have. Which is not surprising, because a 90% good products is a lot easier to produce than one which 99% good.

 

[alyarb]

Oh no! This horrible bug might force slobs like Dennis Nedry to manage their peripherals. What a crime.

 

[podspi]

Oh no! This horrible bug might force slobs like Dennis Nedry to manage their peripherals. What a crime.

I know, right? Who leaves peripherals plugged in...

 

[frozentundra123456]

If true, seems like a pretty big bug to not get detected, unless they can get it fixed before release. They are already behind in mobile and need haswell as soon as possible and with no bugs that will encourage people to delay or buy arm. Do the new windows tablets have USB 3.0 ?

 

[inf64]

This is not a rumor,Mumak on XS confirmed this is really an issue(but a rare occurrence):

It's not the CPU, but the chipset (Lynx Point). And the problem doesn't occur on all USB3 devices, only certain ones depending on the level and magnitude of a signal during wake up from U3.

 

[Charles Kozierok]

Not at all. Processors ship with errata. Impressively large lists of errata. Intel will issue a microcode fix before Haswell even ships and this will be the only time anyone ever hears of it.

This, though I suspect most people will ignore your post and keep arguing about this relative non-issue.

 

[Fjodor2001]

This, though I suspect most people will ignore your post and keep arguing about this relative non-issue.

USB 3.0 devices not awakening from sleep is a non-issue?

Also, you did read this, right?

 

[Rvenger]

Disable sleep. Problem fixed

 

[Turbonium]

Not at all. Processors ship with errata. Impressively large lists of errata. Intel will issue a microcode fix before Haswell even ships and this will be the only time anyone ever hears of it.

What is a microcode fix?

 

[ShintaiDK]

What is a microcode fix?

Whener you start you system, the CPU loads the lastest microcode from BIOS. Its also placed in the OS for load there, but only the newest one is used.

The microcode resolves erratas with certain workarounds.

 

[Fjodor2001]

But where did you read that this can be fixed with a microcode fix? Not everything can be fixed that way you know...

 

[Magic Carpet]

Depends if this a bug with an pre-production stepping that they're sending out to partners for validation/review which will be fixed in a production stepping.

Isn't the production stepping already out?

 

[ShintaiDK]

Isn't the production stepping already out?

Not yet I think. The news is also dated.

It took 3 weeks to respin the old 6 series SATA bug.

 

[Fjodor2001]

Production stepping of what? Note that the latest info is that the HW bug is in the chipset (Lynx Point) and not the CPU (Haswell).

 

[Abwx]

Isn't the production stepping already out?

It is , since late Q3 2012.

It takes two months to manufacture a chip and by now
they have a few millions at most.

Production stepping of what? Note that the latest info is that the HW bug is in the chipset (Lynx Point) and not the CPU (Haswell).

Then it s the chipset that would need a revision , not the CPU...

 

[ViRGE]

But where did you read that this can be fixed with a microcode fix? Not everything can be fixed that way you know...

Virtually everything short of the 6 series degradation bug can be fixed with a microcode workaround. There have been USB issues in the past and that's exactly how they were dealt with.

 

[mrmt]

Isn't the production stepping already out?

They are. Intel took the pre-production charge in Q412, meaning that they should achieve qualification in Q113.

 

[IntelUser2000]

USB 3.0 devices not awakening from sleep is a non-issue?

Also, you did read this, right?

That's not what it said though. It says the issue could make applications relying on USB 3.0 to freeze when resuming from S3. That's a whole different issue from the device itself not awakening altogether.

Now that I think about it, perhaps it is a non-issue. How many people expect video to start playing as soon as you awake your computer? And out of that only when the video is contained in a USB 3.0 device?

 

[mikk]

If the chipset has the failure then it's a good news because a CPU respin needs much longer. Maybe it can be solved with a Bios or driver workaround.

 

[Fjodor2001]

That's not what it said though. It says the issue could make applications relying on USB 3.0 to freeze when resuming from S3. That's a whole different issue from the device itself not awakening altogether.

Now that I think about it, perhaps it is a non-issue. How many people expect video to start playing as soon as you awake your computer? And out of that only when the video is contained in a USB 3.0 device?

It also said the related applications would have to be restarted (e.g. to be able to play the video again).

Let's say you have an external HD connected by USB. Would you like to restart every application that has some content open on the drive every time the computer awakens from sleep?

Obviously this will affect different persons to a different degree depending on what use cases are affected by this. But I really don't see how people can say it's a non-issue!

 

[Phynaz]

It takes two months to manufacture a chip and by now
they have a few millions at most.

It may take GloFo two months, but I can guarantee you that it doesn't take Intel two months to fab a CPU.