"The world has $1 trillion of data center installed, and it ...

"The world has $1 trillion of data center installed, and it used to be 100% CPUs.
In the next 5-10 years, most of that $1 trillion...will be largely gen AI."
-Jensen Huang
___
CPU -> GPU
This means more Capex but also more revenue stream for cloud host companies.
It also means a restructuring or/and new infrastructure to be built. An incredible new revenue source for an already growing environment.
Let us dissect the winners for this shift from central processing unit to graphics processing unit.
___
You are business XYZ looking to build a new GPU based cloud system, what do you need and who do you need to work with?
1. GPU Manufacturers
Start by partnering with leading GPU manufacturers. Ex:
• $NVIDIA(NVDA.US)$ NVIDIA
• $Advanced Micro Devices(AMD.US)$ AMD
These companies produce high-performance GPUs and specifically designed for AI workloads. $Microsoft(MSFT.US)$ and a dozen other firms are rushing to be able to compete within this space.
2. Server Hardware Providers
Collaborate with server hardware providers to obtain the necessary infrastructure for your cloud system.Ex:
$Dell Technologies(DELL.US)$ Dell
$HP Inc(HPQ.US)$ Hewlett Packard Enterprise (HPE), $IBM Corp(IBM.US)$ International Business Machines
$Super Micro Computer(SMCI.US)$ Supermicro
All the above offer server solutions optimized for GPU computing.
3.Networking Equipment
Partner with networking equipment providers to ensure high-speed and reliable connectivity within your cloud system. Ex:
$Cisco(CSCO.US)$ Cisco,
$Juniper Networks(JNPR.US)$ Juniper Networks
$Arista Networks(ANET.US)$ Arista Networks
All offer networking solutions suitable for data centers.
4. Storage Solutions
Implementing efficient and scalable storage solutions is crucial for handling large AI workloads. Consider partnering with companies such as:
$Micron Technology(MU.US)$ Micron
$NetApp(NTAP.US)$ NetApp
$Pure Storage(PSTG.US)$ Pure Storage
$Dell Technologies(DELL.US)$ Dell EMC
The above will help with acquiring storage systems designed for high-performance data processing.
5. Software and AI Frameworks
Collaborate with software vendors and AI framework providers to offer a wide range of software tools and frameworks to your customers.Ex:
TensorFlow
PyTorch
or
$Microsoft(MSFT.US)$ Microsoft (Azure ML)
$IBM Corp(IBM.US)$ IBM (Watson)
The above libraries and companies offer popular AI frameworks and platforms.
6. Cloud Management and Orchestration
Invest in cloud management and orchestration software that allows users to easily provision and manage their GPU-based AI instances.Ex:
OpenStack
$威睿(VMW.US)$ VMware
$Red Hat (OpenShift)
These provide cloud management platforms suitable for deploying and scaling GPU resources.
7. Data Center Infrastructure
Ensure that your data center infrastructure is optimized for AI workloads. This includes power and cooling systems, rack designs, and other infrastructure components. Partnering with data center solution providers. Ex:
$Smart Sand(SND.US)$ Schneider Electric
$Vertiv Holdings(VRT.US)$ Vertiv
These companies can help you optimize the data center environment.
8. Cybersecurity
To ensure a safe and secure cloud solution, partner with specialized cybersecurity providers who offer services such as network security, threat detection, encryption, and access control.Ex:
$Symbotic(SYM.US)$ Symantec

McAfee
$Palo Alto Networks(PANW.US)$ Palo Alto Networks
$CrowdStrike(CRWD.US)$ Crowdstrike
___
As you can see, the new cloud infrastructure involves a lot of helps from a wide range of companies. All of these do not require AI companies to become successful.
They collect a tax of a sort on any company that wants to build the infrastructure of the AI age.
These are the companies that will likely be the winners of the “new age”.
___
AI age ETF
Most are looking at AI companies directly, for the reasons above, I believe it to be the wrong method to easily catch the next performers.
I created a new ETF with as many tickers as I could from the above, making sure not to include $NVIDIA(NVDA.US)$ for obvious reasons.
The result?
+62% ETF vs 35% $Invesco QQQ Trust(QQQ.US)$
The ETF pulled back slightly less at -34% vs -37% for Nasdaq but skyrocketed up almost double what the Nasdaq was able to achieve.
Custom ETF: