1Mon
I find it too uncertain. It can work, but it can also backfire. Data centers are all well and good, but it doesn't really make sense to rent them. Unless you don't expect to need the computing power in the medium term.
Basically, it's like a REIT on crack, so you have extreme short-term returns but also extreme cost of revenue because every 5-7 years everything is completely outdated and you have to modernize.
So there will probably be a spiral of more revenue -> more investment -> more revenue -> more investment and at some point you will have overinvested and it will knock you out.
It's therefore very risky to build up a lot of computing capacity if you don't have any super profitable applications yourself to guarantee utilization. And the logic that you can build an infinite amount of infrastructure and then Bitcoin will sort it out is of course also a cope.
Basically, it's like a REIT on crack, so you have extreme short-term returns but also extreme cost of revenue because every 5-7 years everything is completely outdated and you have to modernize.
So there will probably be a spiral of more revenue -> more investment -> more revenue -> more investment and at some point you will have overinvested and it will knock you out.
It's therefore very risky to build up a lot of computing capacity if you don't have any super profitable applications yourself to guarantee utilization. And the logic that you can build an infinite amount of infrastructure and then Bitcoin will sort it out is of course also a cope.
•
11
•
@Soprano Bitcoin mining is operated with the free capacities. That would be what you mean by "super profitable" applications...
••
1Mon
@Klein-Anleger1 Did you read the last sentence?
••
@Soprano oops, must have missed it somehow 🙈 Sorry
•
11
•
@Soprano I don't see it that way.
to rent them, because you can't start a project like this overnight without expertise, the necessary electricity contracts (which are becoming increasingly difficult) and locations that make sense, plus there is a kind of arms race here and the interference area of $META and co has to be served. This is much larger than the training area. That's also the reason why others like $WULF have a deal and $CIFR will probably be next. And yes, at some point it will probably come down to some kind of ride, but the demand in the cloud /AI area will not stop in the next few years and if there is no big tenant, then there will be many small ones. And $BTC and mining won't disappear overnight either.
to rent them, because you can't start a project like this overnight without expertise, the necessary electricity contracts (which are becoming increasingly difficult) and locations that make sense, plus there is a kind of arms race here and the interference area of $META and co has to be served. This is much larger than the training area. That's also the reason why others like $WULF have a deal and $CIFR will probably be next. And yes, at some point it will probably come down to some kind of ride, but the demand in the cloud /AI area will not stop in the next few years and if there is no big tenant, then there will be many small ones. And $BTC and mining won't disappear overnight either.
•
22
•1Mon
@BamBamInvest Well, the big players are quite capable of taking care of everything themselves. So a meta will probably not move in with Iris Energy, or what do you mean?
This Digital Realty-style server farm rental is particularly interesting for smaller players, but everything is risky there. So with Terawulf, it's really the optimal constellation that Google is involved and backing the deal, otherwise it would just be hot air. 3.7 billion over 10 years - Fluidstack has neither the 3.7 billion nor is it clear whether it will still exist in 10 years. But Google is doing it.
That's why I say yes. It's not completely unlikely that things will go the same way with Iris and that the Big Techs will put one of their protégés there and the share price will continue to rise. But yes, none of this is certain.
Ultimately, it's no different to real estate. You can just have a block of flats and in the best case scenario, the job center pays for everything and they don't care about the rent. But it can always happen that some rental nomads move in and tear your place apart.
This Digital Realty-style server farm rental is particularly interesting for smaller players, but everything is risky there. So with Terawulf, it's really the optimal constellation that Google is involved and backing the deal, otherwise it would just be hot air. 3.7 billion over 10 years - Fluidstack has neither the 3.7 billion nor is it clear whether it will still exist in 10 years. But Google is doing it.
That's why I say yes. It's not completely unlikely that things will go the same way with Iris and that the Big Techs will put one of their protégés there and the share price will continue to rise. But yes, none of this is certain.
Ultimately, it's no different to real estate. You can just have a block of flats and in the best case scenario, the job center pays for everything and they don't care about the rent. But it can always happen that some rental nomads move in and tear your place apart.
•
11
•@Soprano Why not? Anything is possible here, especially as there is no comparable capacity and there is also a severe lack of capacity in the US. $META recently entered into a 20-year deal with Constellation Energy to supply AI Data Center. Also another deal with Sage Geosystems.
These deals show that Meta is focusing on diversified, sustainable energy sources in the long term in order to avoid bottlenecks in the power supply. Colocation (renting space in third-party data centers) would allow Meta to scale capacity quickly without having to build everything from scratch - saving a lot of time and capital.
So why shouldn't this be an option? And since capacities are limited and $META is not the only player on the market, many are already in talks and $IREN is in a position of power due to the capacities that are already and will soon be available and does not need to sell itself short.
In addition, $IRENuses 100% renewable energy (e.g. hydropower in Canada and renewable sources in Texas), which fits in perfectly with
fits perfectly with Meta's sustainability goals.
Childress with 750 MW, expandable, and West Texas with 1.4 GW , also fits geographically with Meta's focus on eastern US locations.
It would also mean low upfront costs for $META, as $IREN provides the infrastructure (e.g. cooling, power redundancy) and has flexibility in the placement of servers or GPUs.
Advantages for Meta:
Meta could utilize capacity in months instead of years, as $IREN sites are already under construction or operational.
IREN's 100% renewable energy reduces Meta's carbon footprint and fits with regulatory requirements
Colocation converts capex into opex, which eases Meta's balance sheet by converting capex into operating expenses, which also leads to positive tax and operating results. $IREN low cost structure (through Bitcoin mining experience) could enable competitive pricing.  
Meta avoids power and land shortages that are increasing in the industry (e.g. delivery times for generators over 52 weeks.
I'm not saying it will $META, but there are some potential customers in the market.
Do a little research on available capacity of useful locations and data centers, colocation vacancy rates and power shortages in the US and you'll see what I mean.
I'm not so worried about that, but time will tell.
Mark my Words 😊 I think the game is only just beginning here ✌️
These deals show that Meta is focusing on diversified, sustainable energy sources in the long term in order to avoid bottlenecks in the power supply. Colocation (renting space in third-party data centers) would allow Meta to scale capacity quickly without having to build everything from scratch - saving a lot of time and capital.
So why shouldn't this be an option? And since capacities are limited and $META is not the only player on the market, many are already in talks and $IREN is in a position of power due to the capacities that are already and will soon be available and does not need to sell itself short.
In addition, $IRENuses 100% renewable energy (e.g. hydropower in Canada and renewable sources in Texas), which fits in perfectly with
fits perfectly with Meta's sustainability goals.
Childress with 750 MW, expandable, and West Texas with 1.4 GW , also fits geographically with Meta's focus on eastern US locations.
It would also mean low upfront costs for $META, as $IREN provides the infrastructure (e.g. cooling, power redundancy) and has flexibility in the placement of servers or GPUs.
Advantages for Meta:
Meta could utilize capacity in months instead of years, as $IREN sites are already under construction or operational.
IREN's 100% renewable energy reduces Meta's carbon footprint and fits with regulatory requirements
Colocation converts capex into opex, which eases Meta's balance sheet by converting capex into operating expenses, which also leads to positive tax and operating results. $IREN low cost structure (through Bitcoin mining experience) could enable competitive pricing.  
Meta avoids power and land shortages that are increasing in the industry (e.g. delivery times for generators over 52 weeks.
I'm not saying it will $META, but there are some potential customers in the market.
Do a little research on available capacity of useful locations and data centers, colocation vacancy rates and power shortages in the US and you'll see what I mean.
I'm not so worried about that, but time will tell.
Mark my Words 😊 I think the game is only just beginning here ✌️
••
1Mon
@BamBamInvest my professional experience with data centers in Germany is that large hyperscalers only rent built to suit and do not sit in ready-made DCs that do not correspond to their design. How do you see the case here at $IREN?
•
22
•1Mon
First of all, Iris is far too small to be relevant for Meta. Meta could cover a maximum of 5-10% of their computing needs if they were to use all of Iris' capacity at once.
Secondly, buying is always better than renting if you want to position yourself strategically in the long term. So you would either buy entire data centers from Iris or take over the company completely (which would certainly be better for both sides)
I'm not saying that Iris has no potential, but that I see it as speculative overall. There are already umpteen cryptominer / data center tenants and the question is how many are needed in the long term.
And all the "shovel manufacturers" for data centers were already played months ago, from electricity providers to cooling specialists - I think at some point some providers will simply be torn apart. In the meantime, I'm already going for "defensive" AI stocks because I know that the topic itself will remain a key technology, but I'm not sure whether the required computing power can continue to grow exponentially forever.
Secondly, buying is always better than renting if you want to position yourself strategically in the long term. So you would either buy entire data centers from Iris or take over the company completely (which would certainly be better for both sides)
I'm not saying that Iris has no potential, but that I see it as speculative overall. There are already umpteen cryptominer / data center tenants and the question is how many are needed in the long term.
And all the "shovel manufacturers" for data centers were already played months ago, from electricity providers to cooling specialists - I think at some point some providers will simply be torn apart. In the meantime, I'm already going for "defensive" AI stocks because I know that the topic itself will remain a key technology, but I'm not sure whether the required computing power can continue to grow exponentially forever.
•
11
•@Soprano all reasonable doubts, but as you say yourself, if it should/can only cover a maximum of 10% of the computing needs and the need will certainly become greater than smaller, they will be dependent on capacities, because you don't start such projects from scratch overnight. And buying/taking over won't work because the founders won't allow it and can prevent it due to their shareholdings.
I think that they will definitely be able to utilize their free and future capacities, because there will definitely be demand and they are up to date here, the data centers are designed and equipped for this. And it also has to be $META and not just 1 hyperscaler. Perhaps it will be used differently with AI Cloud.
But as you rightly say, anything can happen and it can lead to overcapacity in the short term. But in the long term I don't think so. And everything I read is that the data centers are fully utilized everywhere, even in the colocation area, there is little available capacity and locations that can be used sensibly for this, etc.
We will see, but in my opinion we are only at the beginning here and there are a number of companies, sectors and areas that can use free capacity and will have to at some point.
But to raise another point, which the last earnings and the earnings call also showed, in my opinion they will not be dependent on a colocation deal either and are currently transforming themselves into a neo cloud provider and could become a hyperscaler themselves 😊.
I just see a lot of imagination and possibilities here, through the partnership with $NVDA and how they equip their data centers, they are best in class with performance, equipment and locations and no one has their capacity available.
I'm excited to see where it goes. ✌️
I think that they will definitely be able to utilize their free and future capacities, because there will definitely be demand and they are up to date here, the data centers are designed and equipped for this. And it also has to be $META and not just 1 hyperscaler. Perhaps it will be used differently with AI Cloud.
But as you rightly say, anything can happen and it can lead to overcapacity in the short term. But in the long term I don't think so. And everything I read is that the data centers are fully utilized everywhere, even in the colocation area, there is little available capacity and locations that can be used sensibly for this, etc.
We will see, but in my opinion we are only at the beginning here and there are a number of companies, sectors and areas that can use free capacity and will have to at some point.
But to raise another point, which the last earnings and the earnings call also showed, in my opinion they will not be dependent on a colocation deal either and are currently transforming themselves into a neo cloud provider and could become a hyperscaler themselves 😊.
I just see a lot of imagination and possibilities here, through the partnership with $NVDA and how they equip their data centers, they are best in class with performance, equipment and locations and no one has their capacity available.
I'm excited to see where it goes. ✌️
•
11
•@financial_guru_1683 First of all, I have to say that I don't come from this area, I just read a lot about it and sometimes I don't have the technical understanding for it when it goes into great detail. But in my opinion we are entering a new era in which more and more power is needed and the interference area alone is huge. I don't know what the situation is like in Germany at the moment, but when it comes to AI, robotics, cloud, etc., huge capacities will be needed in the future which are not yet available and the choice of locations and maintenance costs are probably not affordable for small companies either, or easier to rent from outside.
Air cooling/liquid cooling is also an issue here for the provision of services, if I understand this correctly.
Building up such capacities costs a lot of time and money and is becoming ever larger due to the volume of data. Many industries and sectors will be dependent on ever-increasing capacities, which is why I still see a lot of growth here. The way the whole thing is accelerating now, I don't believe in overcapacity either, as the build-up also takes time.
It's not just about colocation deals and the company is now positioning itself as a neo cloud provider.
It remains exciting and things usually turn out quite differently than expected. The fact that $IREN
recognized this early on and positioned itself, I think they have an advantage/ head start here.
But I'm happy if you correct me or broaden my horizons if you come from the industry, I like learning here and your views certainly add value. 😊
Air cooling/liquid cooling is also an issue here for the provision of services, if I understand this correctly.
Building up such capacities costs a lot of time and money and is becoming ever larger due to the volume of data. Many industries and sectors will be dependent on ever-increasing capacities, which is why I still see a lot of growth here. The way the whole thing is accelerating now, I don't believe in overcapacity either, as the build-up also takes time.
It's not just about colocation deals and the company is now positioning itself as a neo cloud provider.
It remains exciting and things usually turn out quite differently than expected. The fact that $IREN
recognized this early on and positioned itself, I think they have an advantage/ head start here.
But I'm happy if you correct me or broaden my horizons if you come from the industry, I like learning here and your views certainly add value. 😊
•
22
•