America's No. 1 financial conglomerate

Chapter 399 Taking it all

Gerstner didn't hesitate much about the selling price, but he hesitated a bit about Ernst's second condition: that IBM design a server specifically for him.

Ernst's demand for a custom server design essentially conveyed a key signal: the server product he needed differed significantly from IBM's current mainstream standard server models and could not be met by adjusting the parameters or optimizing the functions of existing products.

For IBM, this means breaking away from the existing production and R&D system, investing additional resources in specialized R&D work, and increasing R&D costs has become an unavoidable reality.

If time is tighter, customized R&D will not only directly increase R&D costs, but also face the dual risks of time cycle and changing requirements, which will undoubtedly further increase labor and R&D costs.

If Ernst adds new requirements during the R&D process, resulting in a change in requirements, IBM will need to redesign the solution, verify the technology, and test the prototype. These are all factors that need to be considered.

Therefore, Gerstner did not want to agree and instead proposed an alternative: "IBM can lower its price by 5%, which is another 5% lower than the lowest ex-factory price."

He wanted to compensate Ernst for his compromises on product standardization by offering price discounts, and tried to dispel Ernst's idea of ​​customized servers by using cost advantages.

However, Ernst shook his head and rejected the proposal almost without hesitation.

Price reduced by 5%?

Ernst wasn't looking for the short-term gains from a 5% price drop, but rather the potential increase of 10%, or even 2%.

0% additional fee.

The current server market lacks a clear product segmentation system, such as web servers, application servers, etc.

File servers and data servers are not available. The mainstream server products on the market are all general-purpose models, lacking in-depth adaptation to specific industry business scenarios.

Ernst's data center has three core requirements for servers: high stability, high scalability, and strong manageability.

High stability is the fundamental guarantee for the continuous operation of data centers. Servers are required to maintain an extremely low failure rate in a 24/7 uninterrupted operation mode. Even in the face of sudden hardware failures or software anomalies, redundancy design can quickly achieve failover to ensure uninterrupted business.

High scalability is necessary to cope with the continuous growth of data center business scale. As the amount of data continues to increase and business modules are gradually expanded, servers need to have flexible hardware expansion capabilities and software upgrade space, so as to achieve seamless expansion of computing resources, storage resources and network resources without interrupting existing business.

Strong manageability is related to the operational efficiency of data centers. It requires servers to have comprehensive remote monitoring, fault diagnosis and automated management functions, which can reduce manual operation and maintenance costs and improve overall operational efficiency.

High configuration, multi-core processor, dual power supply, and high durability are required, meaning that the failure of a single component should not lead to service interruption.

In addition, there are issues of heat dissipation and power consumption. Due to the high-density deployment of servers in data centers, heat dissipation and power consumption issues not only affect the stability of server operation, but more importantly, reduce operating costs.

For Ernst, pursuing ultimate performance was not so important.

It's not that high performance is unnecessary, but rather that stability must be the primary goal when balancing performance and stability.

A single business delay caused by insufficient performance may not seem like much, but a business interruption caused by stability issues can result in incalculable losses.

If we don't develop dedicated servers and save costs in some areas, the additional hardware alone will increase the price of the server by about 20%.

Compared to the increased costs and the potential risks of instability, what is the 5% price reduction proposed by Gerstner?

"8%, that's the limit I can offer." Gerstner made another concession.

Ernst's stance remained unwavering: "Even if you lower it by another 8%, I still won't agree."

Faced with Ernst's insistence on a special edition, Gerstner knew he had no leverage.

The hard drive business is definitely going to be sold, so IBM must secure the server business.

Seeing that Gerstner was still hesitant, Ernst issued a direct and pressure-based ultimatum to him: "How about I go talk to Hewlett-Packard and Compaq?"

Gerstner's face instantly darkened. HP and Compaq, as IBM's main competitors in the server market, both possessed the technological strength to rival IBM.

UNIX workstations running the HP-UX operating system and HP 9000 series servers are highly praised in the market for their stability and scalability, and are widely used in many industries.

Compaq's Compaq ProLiant series servers are also making a strong showing, featuring Intel processors and Windows...

The NT operating system officially brought the WinteI Alliance into the server field.

If Ernst were to switch to partnering with these two companies, IBM would be surpassed even in terms of market share.

"What exactly do you want to do by building so many data centers?" Gerstner couldn't help but ask the question that had been on his mind.

This question was not only a source of curiosity for Gerstner, but also a source of confusion for the entire IBM management team.

When asked by Gerstner, Ernst remained remarkably calm, even responding with a slightly sarcastic tone, "That many?"

He felt that IBM management's reaction was a bit of an overreaction. "I can't reveal the specifics of what they're doing, but I can tell you that this is not the limit, and there will be several more investments of this scale in the future."

IBM executives were stunned when they heard this; he really was just bragging without thinking.

They would have been skeptical of the thirty data centers had Ernst not been willing to include those order numbers in the contract.

The investment was simply too large, and the daily operating costs were too high.

When you say that this level of investment will increase several more times in the future, many people have one thought popping into their minds.

Is this American internet genius a fraud? He's talking nonsense.

Only Gerstner disagreed, not because he believed Ernst's words, but because he felt that none of this had anything to do with IBM.

Whether you brag or not is none of my business; I have to take what's in my mouth.

Since Ernst was unwilling to disclose his specific business intentions, Gerstner decisively gave up asking and instead shifted the focus of the negotiations back to the terms of cooperation.

"I also have two conditions."

Ernst showed some patience with the conditions proposed by Gerstner, nodding to indicate, "Please go ahead."

"The purchase quantities you mentioned earlier must be written in the contract. If your company fails to meet the purchase quantities within five years, I don't want GenTek; instead, Google must compensate IBM with no less than five billion US dollars in cash."

This is because they're afraid Ernst will transfer GenTek's assets, so they're using Google as collateral.

For Google, a prominent internet industry giant, five billion US dollars is not a big problem. With Google as collateral, the cooperation risk with IBM can be effectively reduced.

Ernst agreed almost without hesitation: "If I fail to meet the purchase quantity standards stipulated in the contract within five years, Google will compensate IBM with no less than five billion US dollars in cash. That's fine."

Ernst's quick agreement caused Gerstner to frown slightly. He was somewhat puzzled, not understanding why Ernst would repeat the conditions he had proposed.

Ernst chuckled inwardly. He had no choice but to repeat it; his secretary was taking notes, and he needed to reinforce the message to avoid making mistakes in the contract.

There is a big difference between procurement within five years and purchase within five years. In a business context, procurement usually refers to placing a purchase order, that is, forming a purchase intention and signing an order agreement, while purchase means completing the actual delivery of goods and payment.

Ernst never intended for those thirty data centers to be built within five years.

Based on the business scale and development plan of his company, three to five data centers are sufficient to meet the business needs, and there is no need to build more data centers on a large scale.

Even if cloud computing technology matures in the future, Ernst has no intention of immediately commercializing and opening it up to the public.

He can't possibly reduce the operating costs of internet companies, otherwise when the bubble bursts, there won't be a scene of carnage. If there isn't a scene of carnage, how can he buy at rock-bottom prices?

Therefore, in his strategic vision, the release of cloud computing technology will definitely take place after the bursting of the internet bubble.

At that time, a large number of Internet companies will go bankrupt due to broken capital chains, and a large number of high-quality assets will be sold at low prices in the market. Ernst can take this opportunity to make large-scale bottom-fishing and integrate into the cloud computing system.

At the same time, after the bubble bursts, surviving companies and newly established companies will inevitably join the cloud computing camp in order to reduce operating costs and enhance core competitiveness.

At that time, Ernst will be able to use cloud computing technology to form a strong technological barrier and market moat, firmly controlling his internet empire.

Even if other companies want to enter the cloud computing market, Ernst has already secured the largest and best market share by virtue of its first-mover advantage, forming an unshakeable competitive advantage.

"As for the second condition, IBM wants to sign a strategic contract with you, so that the servers your data center needs in the future must be from IBM."

When faced with Gerstner's second condition, Ernst did not refuse outright, but instead proposed an additional condition: "But there must be one more: the purchase of IBM servers must be in my best interests."

The meeting room fell silent once again.

IBM's management team all had one word on their minds: "Holy crap!"

This dog market not only wants IBM's hard drive business, but also has its eye on IBM's server business.

You want to have it all, don't you?