NVDA
NVIDIA
-- 120.150 TSLA
Tesla
-- 281.950 SMCI
Super Micro Computer
-- 42.950 PLTR
Palantir
-- 84.770 AMD
Advanced Micro Devices
-- 99.510 
Is it a long-term contract on the scale of 1.2 trillion yen with Microsoft and Coreweave for AI infrastructure: The Information Report
The Information reported on the 2nd that Microsoft $MSFT plans to sign a long-term contract totaling 10 billion dollars (about 1.2 trillion yen) from 2023 to 2030 in order to use CoreWeave's GPU cloud infrastructure as an execution platform for the AI model. The agreement is expected to provide GPU computing resources on the scale of approximately 1.4 billion dollars per year.
The highlights of this agreement are summarized in the following 3 points:
1) Secure stable resources through a 7-year long-term fixed contract
2) Large-scale orders to GPU cloud specialists
3) Diversification of supply sources by utilizing infrastructure other than Azure
Microsoft $MSFT currently requires large-scale GPU computing resources, centering on OpenAI's AI model. The company's CEO Satya Nadella stated at the financial results briefing the other day that “demand for AI workloads is expanding at a pace that exceeds expectations,” and this agreement is positioned as a response to rapidly increasing demand.
[Impact analysis on related companies]
1 ️ ︎ Microsoft $MSFT
The recent growth rate of Azure sales is about 30%, of which AI-related demand accounts for about 3%. The following effects are expected from this agreement:
- AI-related revenue: multi-billion dollar annual growth potential
- Azure sales: the added effect of 2-3% annual growth
- Operating margin: 1-2% room for improvement
2 ️ ECT NVIDIA $NVDA
CoreWeave is one of NVIDIA $NVDA's main customers, and it has achieved three times more cost efficiency than conventional cloud providers with infrastructure optimized for AI. According to this agreement:
- Data center segment sales: growth accelerated by more than ten percent per year
- Optimizing investment plans by securing long-term demand
- Possibility of expanding supply to new AI-specialized providers
3 ️ ECT Core Scientific $CORZ
The company announced a 500MW HPC hosting agreement (cumulative revenue 870 billion yen) with Coreweave last week. This deal may result in additional agreements with CoreWeave:
- Annual growth rate of HPC/AI business: potential to accelerate to 30-40%
- EBITDA margin: 5-7% improvement expected
📍 Changes in market structure and future prospects
Conventionally, in the cloud computing world, tripartite competition between Amazon Web Services $AMZN, Microsoft Azure $MSFT, and Google Cloud $GOOGL has continued. However, with the advent of generative AI, this composition is changing dramatically.
The essence of this change lies in the shift from “general purpose” to “specialization.” What is required in the AI era is not just computing power. It is an “AI-specific infrastructure” that maximizes NVIDIA $NVDA GPUs and enhances power efficiency to the extreme.
In this context, the rise of CoreWeave is indicative. The company has achieved three times more cost efficiency than conventional cloud providers with infrastructure optimized for AI. This overwhelming efficiency is thought to have driven Microsoft.
Also, the strategic change of Core Scientific $CORZ can be said to be a movement that symbolizes this change in market structure. The company is proceeding with a shift from an unstable bitcoin mining business to an AI/HPC hosting business based on long-term fixed contracts. The 500MW contract with CoreWeave is an indication of the effectiveness of this transformation. What is particularly important is the 1.2 GW power generation contract capacity that Core Scientific $CORZ has. What determines the winners in the AI era is not simply technical ability. The question is whether stable AI infrastructure and power supply can be secured.
So how should investors make decisions?
The most important thing is choosing a “timeline.” In the short term, this structural change will be accompanied by various disruptions. However, if you look at it on a 3-5 year timeline, the winners should be clear.
Particularly noteworthy indicators are:
- GPU utilization efficiency
- Electricity procurement costs
- Long-term contract acquisition status
- Ability to respond to technological innovation
Changes in these indicators are extremely important in measuring the competitiveness of each company. Then, that change will materialize as a result of a scramble over a new “petroleum” called “computing resources.”
Decisions to accurately capture these changes and utilize investment opportunities at the appropriate timing are required.
The highlights of this agreement are summarized in the following 3 points:
1) Secure stable resources through a 7-year long-term fixed contract
2) Large-scale orders to GPU cloud specialists
3) Diversification of supply sources by utilizing infrastructure other than Azure
Microsoft $MSFT currently requires large-scale GPU computing resources, centering on OpenAI's AI model. The company's CEO Satya Nadella stated at the financial results briefing the other day that “demand for AI workloads is expanding at a pace that exceeds expectations,” and this agreement is positioned as a response to rapidly increasing demand.
[Impact analysis on related companies]
1 ️ ︎ Microsoft $MSFT
The recent growth rate of Azure sales is about 30%, of which AI-related demand accounts for about 3%. The following effects are expected from this agreement:
- AI-related revenue: multi-billion dollar annual growth potential
- Azure sales: the added effect of 2-3% annual growth
- Operating margin: 1-2% room for improvement
2 ️ ECT NVIDIA $NVDA
CoreWeave is one of NVIDIA $NVDA's main customers, and it has achieved three times more cost efficiency than conventional cloud providers with infrastructure optimized for AI. According to this agreement:
- Data center segment sales: growth accelerated by more than ten percent per year
- Optimizing investment plans by securing long-term demand
- Possibility of expanding supply to new AI-specialized providers
3 ️ ECT Core Scientific $CORZ
The company announced a 500MW HPC hosting agreement (cumulative revenue 870 billion yen) with Coreweave last week. This deal may result in additional agreements with CoreWeave:
- Annual growth rate of HPC/AI business: potential to accelerate to 30-40%
- EBITDA margin: 5-7% improvement expected
📍 Changes in market structure and future prospects
Conventionally, in the cloud computing world, tripartite competition between Amazon Web Services $AMZN, Microsoft Azure $MSFT, and Google Cloud $GOOGL has continued. However, with the advent of generative AI, this composition is changing dramatically.
The essence of this change lies in the shift from “general purpose” to “specialization.” What is required in the AI era is not just computing power. It is an “AI-specific infrastructure” that maximizes NVIDIA $NVDA GPUs and enhances power efficiency to the extreme.
In this context, the rise of CoreWeave is indicative. The company has achieved three times more cost efficiency than conventional cloud providers with infrastructure optimized for AI. This overwhelming efficiency is thought to have driven Microsoft.
Also, the strategic change of Core Scientific $CORZ can be said to be a movement that symbolizes this change in market structure. The company is proceeding with a shift from an unstable bitcoin mining business to an AI/HPC hosting business based on long-term fixed contracts. The 500MW contract with CoreWeave is an indication of the effectiveness of this transformation. What is particularly important is the 1.2 GW power generation contract capacity that Core Scientific $CORZ has. What determines the winners in the AI era is not simply technical ability. The question is whether stable AI infrastructure and power supply can be secured.
So how should investors make decisions?
The most important thing is choosing a “timeline.” In the short term, this structural change will be accompanied by various disruptions. However, if you look at it on a 3-5 year timeline, the winners should be clear.
Particularly noteworthy indicators are:
- GPU utilization efficiency
- Electricity procurement costs
- Long-term contract acquisition status
- Ability to respond to technological innovation
Changes in these indicators are extremely important in measuring the competitiveness of each company. Then, that change will materialize as a result of a scramble over a new “petroleum” called “computing resources.”
Decisions to accurately capture these changes and utilize investment opportunities at the appropriate timing are required.
References
Disclaimer: Community is offered by Moomoo Technologies Inc. and is for educational purposes only.
Read more
Comment
Sign in to post a comment
Kimihiko OP : Microsoft $MSFT has been facing a situation where the demand for AI capabilities outpaces the supply. It has been reported that between 2023 and 2030, there is a possibility of spending up to 10 billion dollars to rent AI servers from CoreWeave supported by Nvidia $NVDA.
Kimihiko OP : AI search startup Perplexity supported by NVIDIA $NVDA, which has recently been discussing funding with a valuation of 9 billion dollars, has announced that it is processing more than 0.1 billion search queries every week.