Talk of an imminent sale of Sun to Google has been swirling around trading floors and Silicon Valley for more than a week. Shares of Sun, which has a partnership with Google to develop and distribute each other’s technology, spiked up about 4 percent last week as a result of the rumors. The speculation got even more legs after Google purchased Writely, a maker of a web-based word processor that some people viewed as a product to be added to Sun’s StarOffice suite, which Google may help distribute. It’s also convenient that Eric Schmidt, Google’s chief executive, is the former chief technology officer of Sun.
Unless Sun-Google wants to be the next AOL-Time Warner or HPaq.
Lets refresh, Google uses mostly whitebox Linux servers and Python internally. These technologies are nearly alien to Sun, which is built almost entirely upon Solaris and Java. Google is predominantly a services company, while Sun is a hardware/software company.
Google doesn’t need to buy Sun to get access to OpenOffice.org, because OO.o is already open sourced. Also, it does not appear that Google is going the way of using OO.o directly. Instead, they are recreating it for the web by buying and creating the components required. This is a far more sensible strategy.
So, really, why would Google buy Sun? What would they get for their money?
This is just another groundless Google rumor.
Google posibly uses as much as, or more Java in their servers. And they do not care about what os they actually use so much. what difference does it make if they use Solaris? these are server systems, not desktop after all. Maybe more stability and scalability would come with Solaris if you ask.
But for that rumor, i would say “unlikely”. they can use Sun products without buying sun anyway.
Google posibly uses as much as, or more Java in their servers. And they do not care about what os they actually use so much. what difference does it make if they use Solaris? these are server systems, not desktop after all. Maybe more stability and scalability would come with Solaris if you ask.
But for that rumor, i would say “unlikely”. they can use Sun products without buying sun anyway.
At the moment, Google very much does care about what they run on their servers. Remember, Google isn’t running stock Linux, they employ a good number of kernel developers and they use some very inhouse extensions (i.e. the Google File System). Why would they spend the time/money to port all of that work to Solaris? Regarding stability/scalability, who runs a more scalable web cluster than Google right now? They have well over 150,000 systems running in one giant computing cluster. They don’t really care about stability of either software or hardware because all of their systems are designed to route around failures.
Go read their Google File System and Map Reduce papers at the Google Labs website. Sun and Solaris offers nothing to them that they don’t have with their home built systems and custom Linux.
it is very unlikely that Google will do a file system that only works in Linux.. show me a pointer here.
Here is the white paper they wrote about it: http://labs.google.com/papers/gfs.html
And what do you think, they write a distributed filesystem for handling petabytes of data and they didn’t optimize it 100% specifically for the operating system it’s going to be running on? Oh, and if this helps, here’s the job opening for Linux kernel developers: http://www.google.com/support/jobs/bin/answer.py?answer=23725.
did you actually read that article? they are not really talking about a linux dependency there, contrary they had some problems with the linux’s behavior so thay had to fix it to fit the file system design. Posibly they are looking for engineers to overcome this kind of issues..
So, really, why would Google buy Sun? What would they get for their money?
A fellow R&D based company, with lots of smart people, who build big-ass computers*. You don’t think that Google would benefit from some Niagara stuffed Sun Fire’s?
I don’t think Google is relgious about Linux on whiteboox. If they could get a sweet setup doing Solaris on massive SPARC they’d go for it.
Now I don’t think it’ll happen (though it’d be good for Sun so I wish it would), but I think anyone who thinks there is nothing in it for Google is full of it.
* I realise that both SGI and IBM fall under this category but SGI is really a skeleton, while Sun make not be peak it’s still got life to it. IBM I think is a little too large to take over, even for Google.
It would be a very interesting move.
I was very impressed with Gmail. Most of us thought we’d be stuck with crappy webmailers like Hotmail and the new yahoo (It’s still crap filling up with more crap each day). Google has a lot of potential. They’re the Apple of the Web.
Google could put Sun back on the map. Business is ripe for new approaches. And not just the consumer section.
Edited 2006-03-14 19:20
Perhaps what inspires these rumors is that no one can see how Google is going to tie it all together. They buy some of this here and that there, they launch different ideas as betas, etc., but where is the overall masterplan? People seem to be expecting that the plan will suddenly emerge as a whole operating system and platform. But maybe it will never happen like that because Google isn’t going to become “just another” OS company.
Besides, Google is a publicly traded company these days. Given the close links among senior executives of Google and Sun, Google could face a real storm of criticism that this was all about sweet deals among old pals, etc. So trying to buy Sun might backfire badly anyway.
well actually buying a company that provides and online word editor makes a lot of sence for google and so do all the rest of the purchases. Google is an online service provider and the de facto world’s information hub. It is a lot faster and easier for them to purchase a small company that has a technology that they need than to write everything from scratch ( especially when MS is right behind you ). Oh and I highly doubt that google has any plans on providing an online OS much like MS would like the world to believe.
Google would have to be an idiot to buy Sun. The money is in smart ideas on how to use the network effect on the Internet, not in building boxes. Google would be much smarter in pursuing a Web based MS Office replacement. If they can conqueror that, then move onto Google OS. A web calendar and web storage are obvious short term moves. Smartest move from where I sit would be to put major resources behind turning Firefox into a platform instead of just a browser.
That joint announcement was a charity act from Eric Schmidt to his old firm, nothing more.
Smartest move from where I sit would be to put major resources behind turning Firefox into a platform instead of just a browser
Firefox already is very much a platform. I’m currently working on a project using PHP+XUL+Javascript on Firefox and it is extermely useable today as a rich web application development platform. XUL (XML User-interface Language) is extremely capable for creating complex user interfaces. The biggest problem I’ve come across with this combination is an overall lack of good, thorough, and complete documentation. So far the best source of XUL documentation I’ve found is XUL Planet http://www.xulplanet.com/
Firefox, Thunderbird, NVu, and Seamonkey (the old monolithic Mozilla suite) are all written in XUL, just to attest to how powerful XUL can be.
The Firefox platform still has a ways to go before you can build a Web Office replacement on it. It is missing core components needed to efficiently build a word processor or spreadsheet. But things are moving in the right direction to ultimately support a web office. Moving Firefox onto the XULrunner platform will be a big boost.
Google could do a lot to help this process along. More than just give Mozilla money, Google is sitting on hordes of excellent engineers.
Firefox, Thunderbird, NVu, and Seamonkey (the old monolithic Mozilla suite) are all written in XUL, just to attest to how powerful XUL can be.
And the latest one, the XUL iTunes clone, ‘Songbird Media Player’.
I don’t see what sun brings to the table? They are really good at laying people off. That’s about it.
Seems like an insane match. Have the most successful, agile, profitable internet company in the world take over a struggling, (still) fat server manufacturer with no vision?
http://boondoggle.wordpress.com
“fat server manufacturer with no vision? ”
you described IBM actually. Sun is pretty innovative if you ask me.
If they could stop Scott McNealy from talking smack by getting Sun, I say go for it!
http://news.com.com/2100-1030_3-6049511.html?part=rss&tag=6049511&s…
@Last Software…3D-design toolmaker
I blogged about this earlier today, http://uadmin.blogspot.com/ , here are my toughts.
I think the biggest piece that Google gets out of a merger with Sun is access to the guys that made the ultraT1 chip; Google is rumored to have 150,000 servers that use a lot of power and resources. The UltraT1 is just the platform they need to reduce this by a factor of 5 and there power requirements by a factor of 20 this alone would go a long way to paying for the purchase. If they had a modified version of the UltraT1 that was more suited to there code it could do even better. Anyone care to do the math on what it costs to run 150,000 500watt white box servers not to mention the AC costs, and networking infrastructure, they could reduce it by a factor of 5, that means 80% more scalilibility or 80% smaller power bill.
Second large piece would be the StorageTek, Google is about information and data storage. It has a lot of servers, and I bet more than 75% percent of there servers are not backed up, the data is just replicated everywhere, not a great idea for long term business. Then of course Google could move into the data back up field much like its mini-google search appliances, Google could market data backup appliances to businesses. What happens when 100,000 gmail users loose there data?
Then we have Sun’s programmers and engineer. Google’s current method putting each new employee through 10 plus interviews just doesn’t scale when they want to grow. If they merged with Sun they would have 25,000 bright employees and no headaches of interviewing them all. Many of Sun’s programmers are the best in the tech world. Brian Cantrill the developer of Dtrace comes to mind. Solaris is the current state of the Art operating system. It has more features than Linux, is more stable and scalable. The only thing that Solaris is missing is hardware support. Of course if you are Google and you buy your servers 1,000 at a time, it’s not a problem to pick hardware that is supported. Then we have Sun’s support people, I bet that Google is struggling with supporting its customers of the search appliance as well as in house users. Sun’s support people are used to dealing with enterprise clients that expect 4 and 5 9’s of uptime.
Sun also has lots of people that are familiar with the cell phone business; perhaps Google wants to get more involved with that, since there are more cell phones than computers. Google is already a house hold name on the computer, perhaps its time for them to move into the cell phone platform.
Then we have Sun’s grid network and thin client technology. Google could market the grid to home users, perhaps giving away or renting the client and accessing Google’s suite via Sunray’s. It’s the perfect platform, plug in a router and a Sun(Google)ray 10 seconds later the customer is on google/ig that’s modified with a Google suite including Google word processor, gmail, Google calendar, Google phone and some other web 2.0 apps. No configuration, nothing to break, perfect for grandparents. Google could deploy this technology everywhere, homes, libraries, fast food restaurants, coffee houses, once Google started to mass produce them they could get down to $100 and display on digital TV’s with just 3 wires to connect, not to mention the lock in sure it will give you a way to save your data to other devices, but once you are hooked the Google Ray will as common place as the Remote control, and google search engine is today.
You may say that most of this could be had for less money if you buy the pieces. Google doesn’t have a money shortage. It has a talent shortage, hiring the personnel to manage and merge all the little bits is extremely hard and time consuming given its current hiring practices. Sun employees work together quite well for being a company of 25,000 or so employees. Try and merge 12 startups and have them work as well together and it wont go that smoothly.
Your entire comment would fly completely in the face of how Google works. They don’t want 25,000 bright workers, they put everyone through 10 interviews because they only want the smartest people available. If anyone in any of the 10 interviews isn’t sure of you, you don’t get hired. Google doesn’t want centralized storage, they went and designed and implemented their own file system to handle redundancy everywhere. Why then go and put a single point of failure into it? And the idea of backing up much less restoring petabytes of data (yes, I said petabytes) is comical at best. Google already employs a good percentage of the smartest computer scientists in the world. They’ve already built one of the biggest computing clusters in the world. What does Sun offer them?
Considering Sun’s track record, I’d say it would be like Google committing suicide trying to buy and integrate Sun.
Definitely one of the best comments I’ve read on here in a long time. +1
Indeed. JamesD’s response was better commentary than most of the articles being posted.
Google is standardizing on Java (though indeed PageRank is written in Python I believe), and Sun has expertise in building Internet infrastructure while Google is the master of maintaining it.
Also, Google’s only other source of revenue apart from ads is selling the Search Appliance, which is hardware. Sun’s future is in volume commodity servers, which is right up Google’s alley.
I’d also say Google and Sun have similar visions in making the ol’ adage of “the network is the computer” happen.
However, I doubt Google would do it. Sun would be a huge anchor around the neck of what is now a very agile and healthy company.
Google would probably buy Red Hat or Novell before they would buy Sun.
Now, I do believe that in a few years, Sun is likely to croak and its IP will be bought up piecemeal. Google might take a few bites then.
But who knows … if grid computing takes off (as a result of everything moving to SOA), then Sun is definitely set up for a huge comeback with their Niagara line. And OpenSolaris has a future as a server OS since it has a Linux compatibility layer.
Edited 2006-03-14 20:48
> Google would probably buy Red Hat or Novell before they would buy Sun.
Why buy Open Source?. What does buying Redhat give you that you can’t already get from ftp://mirrors.kernel.org?
If they want something bad from Open Source, get Linus.
They’d do it to get the employees, similar to how they have hired developers of open source projects they liked (such as GAIM).
If they want employees, just go raid Redhat. It’s not like Redhat could have made employees working on open source sign “non-compete” and “ndas”!!!!
Anybody would give their left kidney to work at Google if you are still allowed to work on open source.
Let’s see: CA vs NC – I’d take Cali
GOOG vs RHAT – I’d take GOOG
Ian Murdock has an interesting piece on Google on his blog.
http://ianmurdock.com/?p=313
Basically, anybody that uses Gmail to backup important data is an idiot. It’s not Google’s fault if your data is lost. I noticed that people have talked about Google being a good fit with Sun because Google is about accessing stored data, while Sun has the hardware to store large amounts of data. However, if Google has problems keeping your data (Google reserves the right to modify, suspend, or discontinue the Service without notice per the terms of service), what good would an investment in Sun do?
So with which money exactly is google going to buy sun?
Market capitalization is not the same as actual money in the bank…
Google currently has $9.5 billion in assets according to their Sept 2005 SEC filings. Sun’s market cap is $15 billion. So buying a controlling stake in the company is not necessarily out of their reach.
I am not sure that Google can actually afford to buy Sun Microsystems right now. Also unless Google has a desire to become a hardware provider buying Sun makes no sence. It would be a much better move to ally with Sun but that’s just MHO.
Google + Sun = Google-san
Next step – move HQ to Japan!