IBM can't give away its chip business: report

IBM's attempt to offload its chipmaking business to GlobalFoundries saw it dangle a billion dollars in front of the putative purchaser, but that offer was rejected as inadequate. So says Bloomberg, quoting a “person familiar” with the failed deal. IBM's chippery chaps make fine product, but not so fine or in sufficient …

[Anonymous Coward]

I may be oversimplifying this

So IBM are currently leaking $1b-$2b each year thanks to their chip making business.

They tried to shift it to global foundries with $1b as a sweetener. Global foundries want $2 so they aren't losing money in the first year while they try and turn things around.

IBM said no, and so are going to hold on to the business, probably for another year or so, losing that extra $1b they didn't want to pay anyway, for nothing. Is that about the jist of i?

[Arthur 1]

Re: I may be oversimplifying this

I'm glad I wasn't the only one thinking along these lines as I read it. Sounds like a bargain for IBM.

I suppose it's possible the "believed" numbers are in the totally wrong ballpark. The only way this makes any kind of sense.

[Schultz]

"Big Blue is believed to leak a billion or two dollars each year"

<pendant>

Big Blue is believed to leak one or two billion dollars each year.

FTFY (or did IBM suddenly loose their famed accounting abilities / stopped counting the peanuts)

</pendant>

[wowfood]

Re: "Big Blue is believed to leak a billion or two dollars each year"

<pedant>

or did IBM suddenly lose their famed accounting abilities

</pedant>

[Steve Knox]

Re: "Big Blue is believed to leak a billion or two dollars each year"

<pendant>

....

</pendant>

Well? Don't leave us hanging!

[Denarius]

times have changed

once a foundry business with good chips would have been worth money. Now it cant be given away ? With incentives ? Says a lot about the success of makers of alternative chippery.

[asdf]

Re: times have changed

>Says a lot about the success of makers of alternative chippery.

No it says even with lots of automation as found in IBM's Fishkill fab (been there, pretty impressive) making chips in the first world (which IBM does more than most) is not really competitive with 3rd world child and slave labor still. Can thank Congress (at least in US) partially for that.

[Denarius]

Re: times have changed

Given the cost of the machinery, made in Germany ITIRC, the cost of labour would be very small change.

Taxes could be, if the first world companies paid much, which we all doubt. So that leaves quantity and is the cheaper alternative good enough ? Back to my original point. The alternative stuff is good enough or better and it shows how much times have changed. A big company and good product is not enough. Chippery has to be cheap and already a defacto standard. Not like 30 years ago when there were more types of CPU being brewed and sold, some in house and some by fabless vendors.

[Trevor_Pott]

Re: times have changed

"No it says even with lots of automation as found in IBM's Fishkill fab (been there, pretty impressive) making chips in the first world (which IBM does more than most) is not really competitive with 3rd world child and slave labor still. Can thank Congress (at least in US) partially for that."

Yeah, those human rights are such a drag...

You do realize you aren't the 1%, right? And that without all that nasty "interference" to enshrine human rights in law and then enforce it, you'd be tasting the whip too...

[JMiles]

Apple?

Tim Cook is often in the market for chip related things. Not sure he's interested in having an aging fab though but a US-based fab that Apple owned and ran for chip development and proofing? Could be a possibility, maybe even as a joint-venture with IBM.

[Robert Grant]

Re: Apple?

About time Apple makes all its developers write for a new CPU again.

[psychonaut]

Re: Apple?

i'll call him and let him know

[Anonymous Coward]

selling in bulk.

If I as an individual wanted to design a new product tomorrow without signing an NDA or paying licensing fees. IBM's chips are not what I would choose.

Amtel - yes

Microchip - Yes

ARM - yes

Intel - Yes

PowerPC - No

After three generations of hardware, I might start to look at using IBM chips. Or stick with what I know instead of a new architecture learning curve and adding new sw/hw bugs.

This is fine for IBM, they do not want to deal in small numbers (1k-100k units), but it is why they are making a loss.

[Patched Out]

Re: selling in bulk.

For radiation-hardened, high reliability space-based processors (i.e., Mars rovers):

Amtel - No

Microchip - No

ARM - No

Intel - No

PowerPC - Yes

If I was an astronaut travelling the 9-10 months to Mars, I would want the space qualified/proven version of the PowerPC running things on my spacecraft, no matter how out-of-date it is purported to be.

Newer isn't always better.

[Anonymous Coward]

Re: selling in bulk.

For radiation-hardened, high reliability space-based processors (i.e., Mars rovers):

PowerPC - Yes

Sorry to burst your bubble, wow brilliant IBM has the RAD6000 product range - great seller ?, large margin ? It is not a big seller and will not make a lot of money, if IBM charge too much for it people will go with LEON (OpenSparc), and pay peanuts.

It is exactly products like this that are enabling them to make such losses.

Look at the title - this is not bulk, nor high margin.

[Arthur 1]

Re: selling in bulk.

"For radiation-hardened, high reliability space-based processors (i.e., Mars rovers)"

Ah. That'll be optimizing for the common use case then.

[Pascal Monett]

Normal

Making things is so last millennium.

This millennium it's all about thinking up something slightly different than before (on a mobile!), patenting it and then suing the pants off anybody making anything like it.

[Erik4872]

Re: Normal

"Making things is so last millennium."

Don't you get it? Virtual software-defined hardware is the new new thing. In the cloud! This time it's different.

[amanfromMars 1]

Re: Normal which is ExtraOrdinary

10/10, and go straight to the top of the class, pass Go and collect 200 Bitcoin for that valid observation which really screws up the system, right good and proper, Erik4872.

[Anonymous Coward]

Give it to me

I will take the 1 Billion and free chip factory.

Then liquidate the chip factor and profit :)

[Arthur 1]

Re: Give it to me

It's very likely the contract it's sold under would require you to keep it running for some time.