In the dog days of mid-August it was good of Intel to give us something to talk about by paying $7 billion for a software security firm.
There's not much to be learnt from the marketing-bollox of Intel's boss: "We have concluded that security has now become the third pillar of computing, joining energy-efficient performance and Internet connectivity in importance," said Otellini.
An unkind person might say that Intel has done a poor job of energy-efficient performance.
And though it's fair to say Intel proselytised WiFi connectivity - technologically, companies like Atheros got there first.
So after an indifferent performance in supporting the first two pillars of computing, maybe Intel feels it has a chance to succeed in the third - and only has to shell out $7 billion to achieve it.
And why $7 billion? I get my PC security for free - from AVG.
The wonderful thing about the Internet is that everyone queues up to demonstrate whether they're an idiot or a wise guy.
Opinion on Intel-MacAfee ranged from the idiotic to the wise.
So tell me which I am.
I reckon:
If there is an algorithm at the core of MacAfee which can be put onto a microprocessor with the effect of scanning the virus/bug database quicker and more efficiently than is done at the moment in software - then the Intel-MacAfee tie-up could be worth the investment.
If it is a marketing exercise by which Intel can say their chips deliver a more secure PC - then it's BS.
If it's a natural progression of integrating off-processor functions onto the processor, then it makes sense for Intel to keep adding value to its silicon.
If it's to keep bugs out of the mobile Internet then, in the absence of cell-phone bugs, it's deranged.
But if Intel screws up managing MacAfee - as Intel screws up most of its acquisitions - then none of the above will apply except the last.
Wasn't there a mention of security in the cloud..and how Intel wanted to be a leader there?
Irrespective of Intel's acquisition, security needs to get integrated into our PCs. As long as devices have the "option" to run security software, security remains un-integrated and so prone to subversion.
McAfee being bought over by Google or Microsoft made more sense - because they could claim to integrate it fundamentally into their products and services.
Intel buying McAfee seems a bit weird unless Intel wishes to deliver large swathes of software (for the cloud). Currently Intel delivers as much software as McD delivers healthy food. But you cannot criticize McD if they decide to acquire a health foods co, can you?
Well yes, FTM, I seem to remember McD buying some health food restaurants a year or so ago. But, so far as I'm aware, that strategy died the death. The curious thing about Interl/McAfee is that no one at Intel has yet come up with a convincing reason why they've done it.
Oops! I meant security needs to be integrated into PCs and internet experiences in general. And not into TVs (at least not as much, and not as long as Intel takes over the idiot boxes).
Do I need a beer - or is it a prophetic slip?
Yes, FTM, I've been wondering about the buggy TV. Odd, I thought. Then I remembered Intel has a deal with Sony to develop some kind of PC-TV integration and I thought, Holy Cow - x86s in TVs - now TVs will be crashing halfway through an episode of your favourite soap. But, riding to the rescue, Intel has now bought McAfee in a visionary move to protect TVs from bugs. I think you've hit the nail on the head here, FTM.
McAfee server protection products are used extensively to protect clients. Perhaps they didn't want those skills turned to protecting those same clients from SPARC and ARM based servers ?
$7bn is a lot though - they could have developed a proper optimised 64 bit server chip for far less than that !!
This is a classic case of modern management school dysfunction - that being 'no growth, can't innovate, don't know how to innovate, can't sit on the cash..crap lets just buy something.'
It's early to say but my impression is that this has deeper significance than folks are assuming. To me it suggests that Intel intends to build security hardware assist into their processors/chipsets, which has the potential for huge ructions in the industry.
This would be something like DEP (Data Execute Protection), where Intel tripped over another AMD innovation, but at a deeper and more complex level... say something which allows heuristic detection of suspect software behaviors and possibly triggers new CPU exceptions.
The software security vendors and processor competitors are going to be pissed that they need access to a new proprietary interface to compete. This is Intel's way of course, which we've seen time and time again, e.g. with PCI, AGP, PCIe etc.: create a new interface, set up a SIG and declare it an "open standard", but one where competitors can be excluded at will from the decisions and have no input into the definition... and Intel is in control.
It's a clever way of working around monopoly law and this behavior was part of AMD's complaint against Intel, and if I recall was not something they promised to stop doing. Perhaps I'm speculating too far but to me it makes sense and it hits the competition in the gut nicely.
To all those, Scunnerous and others (and maybe Intel) who think building malware prevention into the silicon is a good idea, I have one thing to say.
False positives. It only takes one. Imagine the recent McAffee disaster happening in the silicon. And don't try and tell me it can't/won't.
That's before we even consider DoS attacks making use of the capability.
Otellini: "We need antivirus, can someone buy me McAfee?"
Few hours later at intel middle management: "Done."
Otellini:"Great, which version?"
"Version ... ?"
Re: Mr Gibbins's interpretation
So much for the deeper meaning, eh Scunnerous (et al)...
ROTFL @ Edward.
If you have a load more like that one you could make a good living at industry events.