IBM in talks to buy SUN

Well not two days after I say how their cloud innovations may be too late in coming for SUN... now IBM is in talks to purchase them. How very interesting.

This potential deal is interesting on multiple fronts.

• RISC (Not Risk as in risky but RISC as in Reduced Instruction Set Computer) This potential merger creates a big home for the remaining non-Intel chip guys. While HP does have their own this combination of fanatical Sun worshipers and the IBM "tried and true"ers is a potentially strong combo. Very different audiences and users though with one driven by tech innovation in Sun users and the other in the "never been fired for picking IBM" mentality.
• Java - Not a lot of the press on the topic brought this up. I guess because it is a shared standard and all that rigmarole. The reality though is that Sun still drives Java in a lot of ways. IBM on the other hand has been a big Java pusher on multiple fronts for development tools, open source etc. This could be a big deal. Of course there is always the gotcha side of the new languages such as SCALA coming up in popularity for the many-core problem.
• Open Source - Sun was big big in the .com era and went down when the .com boom went up. Since then they pushed into Open Source fairly successfully and actually have several software products that they are responsible for that have decent success. Grid, MySQL, even Solaris (a religion among some by itself) are big software assets. Of course the hard part is that IBM has their own efforts in these areas.
• People Logistics - If Sun has an office somewhere, so does IBM. The overhead portion of this could be a big savings opportunity. Not geeky though so I will stop there.
• Competition - Some of the press has stated this is a response to Cisco's entry into the data center space for servers. I doubt that but it does narrow down the big data center players. HP, IBM... ... well...

All told I think this could be good for Sun and IBM. If done right. (isn't that always the trick though... if it works everyone is a genius if it doesn't they were doomed from the get-go.) It gives me some hope for these products and companies for the future and in any case it will be fun to watch.

Friday Link Day - Microsoft to buy Yahoo!?

News has started to come out that Microsoft is looking to resume talks to buy Yahoo! I think someone has Google envy.

Along the lines of the silicon related post from earlier this week... IBM announced that they have devised a way to create vacuum spaces in chips that will allow them to use less power and run faster. The chips even "assemble themselves" to a certain degree.

Getting tired of the laws of physics holding them back with silicon Intel is working on Laser chips now. NEC is saying that laser chips could power petaflop-class chips. Optical chips may still be a few years away but the idea sure is fun.

TPF as a SOA platform

Really... I am not kidding. For those of you who read my posts at all regularly you know that the majority of my time is spent in Open Systems. But in reality there is still a lot of data and processing that occurs on the mainframe. I caught the following press post on Information Week and simply had to read it.

IBM Opens Up System Z Mainframe To SOAs

Now, reading into the press release it's much more fluffy than the title sounds. But really, would we expect IBM marketing to not be able to do their jobs? (I think they were the original inventors of F.U.D., right?)

Still, it does make you think about what kind of data source a mainframe could be for any high volume system. The ability to handle large amounts of data and transactions in a stable manner is a consistent throttle that we encounter in the Open Systems world. Now the cost/benefit needs to also be compelling since there are many other ways to solve the problem then simply making the iron bigger.

I had an opportunity to talk to some IBMers lately and they are still big on big boxes, just now they want to slice those boxes into 100s of virtual servers and provision them dynamically. I found it ironic that when you consider the open source revolution pushing hard away from big iron, we are now pushing closer to it. The IBM Z series and the IBM P series even sit in the same "refrigerator boxes" now. (For those who are not Z or P savvy Z=Mainframe, P = Open Systems) So you could walk out on a raised floor and not even know if it's an open system or a mainframe system... now it's just a system.

But then, isn't that how a service consuming customer views it?