New hardware required to fix Haswell USB 3.0 connection bug

[Shawn Knight]

A few weeks ago we learned that Intel’s upcoming Haswell platform was plagued with an annoying USB 3.0 bug. The chip maker didn’t immediately fess up to the issue but we’ve since heard that it is indeed a real cause for concern – so much so that it will require a new chipset revision to fix, according to Fudzilla.

If you recall, the bug rears its head when a Haswell system is woken up from an S3 (suspend to RAM) sleep state while a USB 3.0 device is attached. The attached device will disconnect itself, forcing the user to reconnect in order to access the data once again. It’s not a major issue as data isn’t compromised but if you wanted to pick up working where you left off on data from a USB drive, it’ll quickly become an annoyance if nothing else.

Motherboard partners are reportedly continuing ahead as planned with production on word from Intel that Haswell desktop components will be ready in June, probably launching at Computex 2013. Instead of delaying the platform’s launch, Intel will simply address it in a future chipset revision which means the first batch of motherboards will likely carry the bug.

Back in 2011, early models of Intel’s Sandy Bridge platform were found to have a faulty SATA 3Gb/s controller that would cause gradual degradation and eventual failure over a long period of time. Intel was forced to issue a general recall as part of a move that is believed to have cost the company upwards of $1 billion.

Permalink to story.

https://www.techspot.com/news/51895-new-hardware-required-to-fix-haswell-usb-30-connection-bug.html

 

[Jad Chaar]

Good to hear they are on the job. But dang, I feel bad for whoever gets a faulty board.

 

[treeski]

It's annoying, but it wouldn't prevent me from picking up a new computer/motherboard or anything.

 

[Burty117]

Good to hear they are on the job. But dang, I feel bad for whoever gets a fault board. I wonder if a chipset driver could fix this for people who buy the first batch

Well no, the entire point of the article was to point out it needs a hardware fix and not a software one?

Anyway,
Yeah, I still plan on getting this at launch as well, I rarely use S3 sleep anyway and if I do, everything I have is USB3 and I don't use external drives making this even less annoying for me.

 

[Jad Chaar]

Well no, the entire point of the article was to point out it needs a hardware fix and not a software one?

Anyway,
Yeah, I still plan on getting this at launch as well, I rarely use S3 sleep anyway and if I do, everything I have is USB3 and I don't use external drives making this even less annoying for me.

Yeah, but I still dont see why this cant be fixed without a firmware update.

 

[Burty117]

Yeah, but I still dont see why this cant be fixed without a firmware update.

Meh, we'll leave the Intel Engineers to decide on that one xD
From what I gather though, it sounds more like an architecture flaw and since they've already started production on the chipsets, they will simply fix it in a future hardware revision, a firmware upgrade cannot fix a physical issue. :/

 

[Jad Chaar]

Meh, we'll leave the Intel Engineers to decide on that one xD
From what I gather though, it sounds more like an architecture flaw and since they've already started production on the chipsets, they will simply fix it in a future hardware revision, a firmware upgrade cannot fix a physical issue. :/

Yeah, true.

 

[hood6558]

I never use sleep states, and anyway my Ivy Bridge Z77 board does the same thing (drops USB 3.0 drive after idle periods) as do many other configurations, so it's nothing new or serious - just another FUBARed release from Intel. I'm sure they have a few extra billion stashed away for these all too frequent contingencies. Except this time they're not recalling the chips, so it will only cost them the price of launching a new chipset, not buying back any bad hardware. It will inevitably slow Haswell sales at launch, since few will have forgotten the last fiasco. Maybe they'll all wait for Broadwell and just pass on Haswell. This is bad news...

 

[Skidmarksdeluxe]

It's annoying, but it wouldn't prevent me from picking up a new computer/motherboard or anything.

+1 I don't use S3 sleep anyway so it's a non issue for me.

 

[Vrmithrax]

As others have noted, S3 sleep in desktops seems rather unused, so Intel shouldn't have too much of a fiasco if that is the only function involved in this bug. That said, I'm assuming this issue will not plague notebooks or mobile platforms? I could see where it would be a potential disaster to have a sleep state bug in that arena.

 

[Guest]

This is the problem. They know there is a problem but are going ahead with the release anyway. This is saying if you don't like it don't buy or if you bought it sue us to meke us recall it. AMD is not a competision anymore so we can do as we please. Sue us or buy amd if you don't like. Hope they loose more then a billion. At least a good amout that will hurt. I am getting tired of crap releases for they are rushed to production and out the door for top dollars. I rather wait and get a good product that will last me. My uncle by a computer for $2 on garage sale. This thing is from like 2000. It still works. Not a problem with it. It has the original hard drive. This is a 13 year old system. My mac book pro is on the secon logic board and the video is failing on the second logic board. The hard drive crashed on light use after 1.5 years. I am on the second hard drive. The drive was a toshiba. I gat a hitachy in there currently. As far as the hard drive we are talking about catastrofic faliure where the hard drive wont even boot. I recently used it as a time machine drive the drive backed up and them lost large chunks of the data. About %10 was usable and had to wait for long periods of recovery. The drive out of the $2 machine has 0 errors. Strange ha. And mind you that I have a 250 gig hard drive in that machine. The mac book has a 750 drive and 2.5 inch. So we may look at 5 to 7 times the arial density. But I am willing to bet that the Hitachy drive which was manufactured by Japan and not Tai Land will last longer. I am in a way glad that Tai land got fluded. That was just the computer god puting an end to crappy products. I guess there is a god. We may pay more for hard drives but will only have to buy once or twice every 10 years as versus every year.

 

[treeski]

This is the problem. They know there is a problem but are going ahead with the release anyway. This is saying if you don't like it don't buy or if you bought it sue us to meke us recall it. AMD is not a competision anymore so we can do as we please.

It's annoying that they didn't catch the bug, but considering the bug is more annoying than anything else, I think people would be much more put off if they didn't release Haswell asap. It's a very anticipated CPU release.