A TPod for Mr Schwartz and other musings
Someone at SUN truly needs to start taking people like Paul http://blogs.zdnet.com/Murphy/?p=511
to heart and start changing their priorities
IMHO the data center culture is the primary sustainer for propietary closed software/IT systems in general and the low ROI for IT projects overall. SUN needs to focus on where the customers for low cost computing are going not their existing big box folks like GE.
If Jonathan is really serious about the "long tail" http://blogs.sun.com/roller/page/jonathan?entry=the_longest_tail then lets see some real out of the box INTEGRATED thinking, action and most importantly flawless execution from the hardware,software and business guys at SUN to return to its roots!
The T1 and Open Solaris are a step in the right direction. But technology is only one part of the equation and SUN does a very good job at that. However you cannot change the price point of one part of the experience, without reexamining the costs of the ecosystem that it resides in, as Paul points out in his blog.
Let's take Paul's spec since I consider him a reasonable person and translate what it would mean to some one delivering IT services small business or say IT in China or India.
For e.g. the question in their minds would be how can they plug in a bunch of these http://www.storagereview.com/articles/200601/WD1500ADFD_2.html and use ZFS get RAID and back up for cheap.
The point for small business users, the geeks at home and the "long tail" users in India/China etc. is that the T1 IS the data center .
So if SUN haven't gotten rid of the legacy IT bling bling like rackmounts and associated accessories, IS the new T1 equation really game changing??!!
The destruction of the data center (and the culture that goes with it, primarily in accessibility and close minded approaches) ) is what long tail computing is all about.
SUN seriously needs to rethink price/design points and form factors taking the entire ecosystem into account. Whup ass devices at 3k price points not just CPU but everything CPU, disk, wireless, mobility, networking. And it needs to happen yesterday, not with Russian style planning in 5 year projections on power points. The marketing question that needs to be asked is how can I make my product relevant to some one who has a disposable income of $1 a day.
For stuff like byte code based mobile code computing(ala Jini) the hardware world is just taking off but the key is going to be very high volume, T1 type chips at ecosystems that are extremely low power, in ipod nano form factors and free cell phone price points.
CELL BTW IMHO was IBMs sad attempt at it probably because Sony drove the design specs.
There are a gaziliion small businesses, device manufacturers and geeks who would want a T1 in a Mac mini form factor/feature set, a kind of stripped down 4 core T1000). More brownie points if it has built in mesh networking and Tarantella enabled remote desking. Or just bundle it with this http://www.tadpolecomputer.com/html/products/mobile/comet12/ maybe with a Blu-Ray burner and SUNs growth in "long tail" buyers would be far more significant than the IT edifices.
Low cost computing clusters are not going to come together in permanently organized form factors like racks. They are going to be more like automata and fractals linking together on a as needed basis to produce results and then move on to new worlds to conquer. More bazaar like to use an oft quoted cliche.
Another example as Ben points out on the Ultra 820 http://www.cuddletech.com/blog/pivot/entry.php?id=412 and highlighted by the responses to him, too often the product is far away from a home enthusiast requirement and designed by guys who measure performance by how fast MS Word boots up.
How accessible is this hardware to the conversation drivers in the GLOBAL long tail marketplace? What is the point in all that cool software if the cool hardware is not accessible and by accessible I mean both in terms of price points and form factors. the Ultra 20 was good but it missed some key trends, small form factors, mobility and wireless !! how about a basic laptop running the T1 or a cube with builtin wi-fi and blue tooth.??
Why not a laptop range based on the T2 or the Turion dual cores based on the subscription model. The margins would still be as good as the low end boxes that SUN ships.
BTW same thing applies to the Galaxy boxes. They're good. But not even a single blade product, this is Andy for crying out loud people are expecting a lot more! Maybe age is catching up with him or he is just getting lazy ;-)!!
I mean HP ate up the Opteron market with vaniila designs even while pitching Itanic with a straight face!! That is brilliant marketing.
Sun has always geared its products to the hardware/unix geeks that drive technology just as apple is targetting the music geeks unfortunately i fear a significant amount of the dot com poisoning still afflicts their DNA.
Sun has all the technology pieces right now and even the people with the vision to see where it can go e.g. http://blogs.sun.com/chandan?anchor=calculating_the_aio_card. But the question is whether they have a Steve Jobs kind marketing chutzpah and savvy to turn on a dime and create a TPod market place for its geeks! The monetization will take care of itself
Throw in some SPOTs(http://www.sunspotworld.com/), Jini driven Mindstorms and the possibilities are endless. Sometimes I think the SUN guys are so busy playing with and improving their cool toys that no one has time to put it all together for the rest of us!
Instead of opensourcing the chip design go a step further and open source a TPod reference hardware spec or specs with a CDDL/GPL 3 like license.
Let a thousand flowers bloom just like the PC world which BTW actually seems to be regressing in terms of the hardware design openess.
SUN's got world of hardware excited with the T1. Now for achange they need to follow through on the execution and finish to reap the benefits!
BTW as reflected by various posts on the net(including Jonathan's) the finance dept at SUN probably needs a serious re-org since they are still working with the dot bomb mindset when treating their geek customers as opposed to the GE's of the world. They need to change to the low cost high volume long tail mindset.
Also their new CFO needs to be some one who can force accountability on their financial execution and make their quarters look good. When HP accountants with their disastorous Compaq/Itanic expeditions or IBM with Las Vegas style pension fund gambling can make their books look good, I cannot understand how the GAAP numbers for a cash flow positive company can look this pathetic.
Now another topic regarding the SAN cult in corporate IT and some quotes by Scott McNealy http://www.eweek.com/article2/0,1895,1907211,00.asp on exit costs.
The reality is that the SAN has truly replaced the mainframe and servers as the pinnacle of ivory tower data centers, not surprising considering the mainframe origins of the industry leaders :-)
Sun also has to understand their commitment to bleeding edge technology and openness (which creates a low exit cost from them unlike the mainframe guys) comes at the price of eternal vigilance with respect to what is happening in the tech market place. If their marketing/techies folks spend too much time at the ivory towers and not enough at the edge then they miss out on the conversation and the market place drivers.
Another development is the the PC guys have figured out how to expand their turf into the ivory tower, the magical solution to all performance, reliability and political issues is the one server/ per project/ per app/per server/per VP/per Director/per project with redundant everything
Of course these days open source is the fig leaf to cover the bloat, with the IT lambs clueless that this is just the "Global Services" rod being rammed up their collective arses or some wannabe $500/hr FUDmeisters http://www.linuxcio.com/VenusMars.html in the employ of some well funded printing ink manufacturers (with a willingness launch lawsuits to prevent consumer choice. e.g. http://www.hp.com/hpinfo/newsroom/press/2002/020128b.htm) driving their own hidden corporate agenda.
Mr Finks recent ads on slashdot btw come pretty close to Roland Piquepaille posts in annoyability ratings!
With all these corporations being run by technologically brain dead accountants/lawyers who wouldn't realise that their 1 million mainframe was replaced by a clustered PC server environment costing even more sometimes just in terms of windows licensing or open source with proprietary hooks(Your first hit is free!!) life goes on with business as usual.
Once the implementation bills start coming in and the red ink starts flowing the numbers don't look very pretty forcing the above mentioned accountants to run to their cookie jars, fuzzy math recipes and the miracle Wall Street quarterly 1 penny profit model - layoffs!!
But these guys are still small potatoes compared to our telco IT overlords who even in the impending world of google and skype, order Itanic 1 Superdomes by the dozen for lunch without even checking the sell by dates.
Like obesity, IT bloat at corporations is a technology crisis in the US. BTW if you work at place that reverberates with words like EAI/MQ Series your company my friend has been infected. Get out of there as soon as possible if you do not want your brain to vegetate.
In my opinion SUN trying to sell anything to these guys based on cost/performance is a lost case simply because like it or not SUNs only marketing department is Scott McNealy's mouth with the user friendliness of a bull horn and corporate IT works more like Congress. You want to sell Niagara to IT overlords in the US do it like the big guys. Get the honchos to the golf courses, NASCAR/NBA tickets or wine and Curly depending on their tastes :-)
Can you tell I'm a corporate IT pessimist :-/ Now satire aside is it any wonder that IT is getting outsourced. With the current business model in place outsourcing is almost a no brainer as an option since IT has become a financial blackhole that will require a million consulting serfs to tend to it rather than the automation of human drudgery that its supposed to deliver.
Sid
PS: A suggestion to Jonathan. Drop the "participation" "Web 2.0" memes they do not capture the vision. "Long tail computing" as a meme on the other hand does create a visual metaphor and escapsulate the ideas of the current trends much better.
to heart and start changing their priorities
IMHO the data center culture is the primary sustainer for propietary closed software/IT systems in general and the low ROI for IT projects overall. SUN needs to focus on where the customers for low cost computing are going not their existing big box folks like GE.
If Jonathan is really serious about the "long tail" http://blogs.sun.com/roller/page/jonathan?entry=the_longest_tail then lets see some real out of the box INTEGRATED thinking, action and most importantly flawless execution from the hardware,software and business guys at SUN to return to its roots!
The T1 and Open Solaris are a step in the right direction. But technology is only one part of the equation and SUN does a very good job at that. However you cannot change the price point of one part of the experience, without reexamining the costs of the ecosystem that it resides in, as Paul points out in his blog.
Let's take Paul's spec since I consider him a reasonable person and translate what it would mean to some one delivering IT services small business or say IT in China or India.
For e.g. the question in their minds would be how can they plug in a bunch of these http://www.storagereview.com/articles/200601/WD1500ADFD_2.html and use ZFS get RAID and back up for cheap.
The point for small business users, the geeks at home and the "long tail" users in India/China etc. is that the T1 IS the data center .
So if SUN haven't gotten rid of the legacy IT bling bling like rackmounts and associated accessories, IS the new T1 equation really game changing??!!
The destruction of the data center (and the culture that goes with it, primarily in accessibility and close minded approaches) ) is what long tail computing is all about.
SUN seriously needs to rethink price/design points and form factors taking the entire ecosystem into account. Whup ass devices at 3k price points not just CPU but everything CPU, disk, wireless, mobility, networking. And it needs to happen yesterday, not with Russian style planning in 5 year projections on power points. The marketing question that needs to be asked is how can I make my product relevant to some one who has a disposable income of $1 a day.
For stuff like byte code based mobile code computing(ala Jini) the hardware world is just taking off but the key is going to be very high volume, T1 type chips at ecosystems that are extremely low power, in ipod nano form factors and free cell phone price points.
CELL BTW IMHO was IBMs sad attempt at it probably because Sony drove the design specs.
There are a gaziliion small businesses, device manufacturers and geeks who would want a T1 in a Mac mini form factor/feature set, a kind of stripped down 4 core T1000). More brownie points if it has built in mesh networking and Tarantella enabled remote desking. Or just bundle it with this http://www.tadpolecomputer.com/html/products/mobile/comet12/ maybe with a Blu-Ray burner and SUNs growth in "long tail" buyers would be far more significant than the IT edifices.
Low cost computing clusters are not going to come together in permanently organized form factors like racks. They are going to be more like automata and fractals linking together on a as needed basis to produce results and then move on to new worlds to conquer. More bazaar like to use an oft quoted cliche.
Another example as Ben points out on the Ultra 820 http://www.cuddletech.com/blog/pivot/entry.php?id=412 and highlighted by the responses to him, too often the product is far away from a home enthusiast requirement and designed by guys who measure performance by how fast MS Word boots up.
How accessible is this hardware to the conversation drivers in the GLOBAL long tail marketplace? What is the point in all that cool software if the cool hardware is not accessible and by accessible I mean both in terms of price points and form factors. the Ultra 20 was good but it missed some key trends, small form factors, mobility and wireless !! how about a basic laptop running the T1 or a cube with builtin wi-fi and blue tooth.??
Why not a laptop range based on the T2 or the Turion dual cores based on the subscription model. The margins would still be as good as the low end boxes that SUN ships.
BTW same thing applies to the Galaxy boxes. They're good. But not even a single blade product, this is Andy for crying out loud people are expecting a lot more! Maybe age is catching up with him or he is just getting lazy ;-)!!
I mean HP ate up the Opteron market with vaniila designs even while pitching Itanic with a straight face!! That is brilliant marketing.
Sun has always geared its products to the hardware/unix geeks that drive technology just as apple is targetting the music geeks unfortunately i fear a significant amount of the dot com poisoning still afflicts their DNA.
Sun has all the technology pieces right now and even the people with the vision to see where it can go e.g. http://blogs.sun.com/chandan?anchor=calculating_the_aio_card. But the question is whether they have a Steve Jobs kind marketing chutzpah and savvy to turn on a dime and create a TPod market place for its geeks! The monetization will take care of itself
Throw in some SPOTs(http://www.sunspotworld.com/), Jini driven Mindstorms and the possibilities are endless. Sometimes I think the SUN guys are so busy playing with and improving their cool toys that no one has time to put it all together for the rest of us!
Instead of opensourcing the chip design go a step further and open source a TPod reference hardware spec or specs with a CDDL/GPL 3 like license.
Let a thousand flowers bloom just like the PC world which BTW actually seems to be regressing in terms of the hardware design openess.
SUN's got world of hardware excited with the T1. Now for achange they need to follow through on the execution and finish to reap the benefits!
BTW as reflected by various posts on the net(including Jonathan's) the finance dept at SUN probably needs a serious re-org since they are still working with the dot bomb mindset when treating their geek customers as opposed to the GE's of the world. They need to change to the low cost high volume long tail mindset.
Also their new CFO needs to be some one who can force accountability on their financial execution and make their quarters look good. When HP accountants with their disastorous Compaq/Itanic expeditions or IBM with Las Vegas style pension fund gambling can make their books look good, I cannot understand how the GAAP numbers for a cash flow positive company can look this pathetic.
Now another topic regarding the SAN cult in corporate IT and some quotes by Scott McNealy http://www.eweek.com/article2/0,1895,1907211,00.asp on exit costs.
The reality is that the SAN has truly replaced the mainframe and servers as the pinnacle of ivory tower data centers, not surprising considering the mainframe origins of the industry leaders :-)
Sun also has to understand their commitment to bleeding edge technology and openness (which creates a low exit cost from them unlike the mainframe guys) comes at the price of eternal vigilance with respect to what is happening in the tech market place. If their marketing/techies folks spend too much time at the ivory towers and not enough at the edge then they miss out on the conversation and the market place drivers.
Another development is the the PC guys have figured out how to expand their turf into the ivory tower, the magical solution to all performance, reliability and political issues is the one server/ per project/ per app/per server/per VP/per Director/per project with redundant everything
Of course these days open source is the fig leaf to cover the bloat, with the IT lambs clueless that this is just the "Global Services" rod being rammed up their collective arses or some wannabe $500/hr FUDmeisters http://www.linuxcio.com/VenusMars.html in the employ of some well funded printing ink manufacturers (with a willingness launch lawsuits to prevent consumer choice. e.g. http://www.hp.com/hpinfo/newsroom/press/2002/020128b.htm) driving their own hidden corporate agenda.
Mr Finks recent ads on slashdot btw come pretty close to Roland Piquepaille posts in annoyability ratings!
With all these corporations being run by technologically brain dead accountants/lawyers who wouldn't realise that their 1 million mainframe was replaced by a clustered PC server environment costing even more sometimes just in terms of windows licensing or open source with proprietary hooks(Your first hit is free!!) life goes on with business as usual.
Once the implementation bills start coming in and the red ink starts flowing the numbers don't look very pretty forcing the above mentioned accountants to run to their cookie jars, fuzzy math recipes and the miracle Wall Street quarterly 1 penny profit model - layoffs!!
But these guys are still small potatoes compared to our telco IT overlords who even in the impending world of google and skype, order Itanic 1 Superdomes by the dozen for lunch without even checking the sell by dates.
Like obesity, IT bloat at corporations is a technology crisis in the US. BTW if you work at place that reverberates with words like EAI/MQ Series your company my friend has been infected. Get out of there as soon as possible if you do not want your brain to vegetate.
In my opinion SUN trying to sell anything to these guys based on cost/performance is a lost case simply because like it or not SUNs only marketing department is Scott McNealy's mouth with the user friendliness of a bull horn and corporate IT works more like Congress. You want to sell Niagara to IT overlords in the US do it like the big guys. Get the honchos to the golf courses, NASCAR/NBA tickets or wine and Curly depending on their tastes :-)
Can you tell I'm a corporate IT pessimist :-/ Now satire aside is it any wonder that IT is getting outsourced. With the current business model in place outsourcing is almost a no brainer as an option since IT has become a financial blackhole that will require a million consulting serfs to tend to it rather than the automation of human drudgery that its supposed to deliver.
Sid
PS: A suggestion to Jonathan. Drop the "participation" "Web 2.0" memes they do not capture the vision. "Long tail computing" as a meme on the other hand does create a visual metaphor and escapsulate the ideas of the current trends much better.
posted by Sid Art at 12:26 AM
14 comments
