The Cloud, page-8

Ok, nice story Interloping. Good chat, most probably unrelated to Weebit.  And I'd go through every line of your well reasoned discourse and respond but I am too busy quoting a managed applications and technology service for a national customer right now. In other words I do this cloud and managed stuff for a living. I prefer to deal with cloud realists than cloud evangelists.

There is the story of the cloud, then there is the reality of the cloud. At some point, everyone has to pay for what they use, plus render a profit to the provider. And what they use is still quantified as compute units, network units and storage units. Also Application function usage or user counts for SaaS. And somewhere to grow compute units a box has to be bought and installed, for greater bandwidth more network bought, and for storage more memory or disk or whatever acquired. At some point you cant put a box in and someone  has to spend millions of dollars on a new datacentre or more space in one. They are the building blocks, and the provider wears the capex or lease. Just like in the traditional model.

I brought tape into it because it still holds by far the bulk of data. That is changing quickly and not for long, but, its not all "archival". It is in some cases still a potential 3rd tier or 4th or 5th. BTW, most data is "archival" the moment it is generated, since the likelihood of it actually ever being used for anything meaningful or looked at again once it becomes the input to some greater value information product is most likely nil. And a great bulk of data that is being kept will never be used. Its existence is partly because it is being stored because it "might" be used. It is expensive to keep data close to hand that may never be used, hence the discussion on storage hierarchies. The problem with "the cloud" right now is the lack of value being placed on this information in regards to it potential for adding future value. In the near future, a lot of companies will go broke holding data on expensive storage just in case it might be useful some day.

>>means for the end-user is YOU don't know where the data physically sits - and you don't care (beyond knowing it is secure)
Now the whole cloud storage you dont care where it is thing. When real decisions are made about data, that is rarely the case.Serious companies want to know precisely where their most important asset will live. Not a week goes by where data soverignty isnt raised as an issue in my working world.  Some situations allow for that, but its far less common than you think. Knowing your data is "secure". What does that really mean? Secure from access over the net? Secure from the employee who walks out with the disk? Secure from the earthquake & tsunami about the hit the West Coast of the USA?...and your Multi-site facilities are in San Fran and LA. But you dont care where the data is because its "secure". Just saying, the cloud concept doesnt mitigate the need for exactly the same due diligence on your information assets as if you were on-premise.

There are very few companies on the bleeding edge of information science, whether in the cloud or not. Most people are merely recreating their original apps environments on new kit up there. Perhaps adding some more analytics or replacing their existing ones, or replacing their old ERP with a worse cloudified one.In that sense the only differentiating feature is its multi-tenancy capability.
They are not truly exploiting anything, merely seeking a more flexible IT model in terms of scalability and a means of turning capex into opex.  So its always been in business and again these people are still the trend followers. They didnt push into the cloud for competitive advantage (ie an investment), they pushed into the cloud to just save a buck. And they dont always succeed at that even.

Also the question of governance comes into play. You can look at the mass of data but it isnt accessible in the way you'd think due to logical business and other boundaries. The hard reality of this hits home to many companies after their first cloud dabble. If it werent so, you couldnt explain the number of companies still running on-premise or in hybrid cloud models to protect their jewels, the information that is really used to run their businesses.

The real information exploitation companies are actually very small in number.

On your question of the cloud requiring more storage due to multi-site redundancy. Some of that is due to bad information management in the first place, some due to costly n+1 technical approaches in like in HPC scenarios, where cost doesnt matter. Hadoop File System for instance is not the way it is due to some clever n+1 redundancy strategy, the architecture is merely highly wasteful to keep things in memory. Them storing something in 4 places for safety when in all likelihood you dont need the information at all is just a costly waste. And besides, most good business maintain a similar plan for their valuable data anyway, its not exclusive to the cloud.

Now, I dont want to define weebit's market. I want the market to do that. If it ends up in every macbook or next to a bunch of AI accelerators or as longer term faster lower power storage in a data centre. Whom am I to question that? Perhaps a new level of storage with profoundly good long term economics will allow some more wasteful information management practices to flourish without killing the owners.