What to know:
- Nvidia AI server prices could rise by up to 15% as memory costs increase.
- Vera Rubin and Grace Blackwell systems are facing higher costs due to heavy DRAM demand.
- Samsung, SK Hynix, and Micron are seeing strong AI-driven memory demand outpace supply.
Nvidia Corporation has allegedly been issuing notices to some of its largest clients of increases in the cost of AI servers driven by its chips.
Bloomberg sources say that some Nvidia-based server prices due to be delivered in early next year may be rising as much as 15%. The extent of the price rise would depend on the type of chip and the quantity and type of memory contained within the servers.
Nvidia Faces Higher Costs as Memory Price Rises
These pricing hikes pertain to the systems that are designed using the latest Nvidia AI hardware, namely the Vera Rubin and Grace Blackwell systems.
These pricing details were supposedly communicated to customers by server makers who assemble the computer systems for large-scale data centers like Microsoft, Google, and Oracle.
One of the most important reasons for these costs to be on the high side is the memory requirement. The company’s accelerators consume a lot of DRAM.
Samsung, SK Hynix, and Micron dominate the memory market across the globe. Even though the production of memory is on the rise, there is an even higher increase in demand for memory from AI data centers than the supply.
This situation has seen memory prices rise and enabled big memory companies to have more control over the prices of AI hardware.
According to Micron’s CEO, Sanjay Mehrotra, memory plays a very significant role in the AI industry because of the need for larger, faster, and more efficient systems.
Also Read | Coinbase One Unlocks Lucrative 6.5% APY on $500K USDC
Nvidia Faces Rising Costs Across the AI Boom
The gross profit margin of Nvidia also stands among the highest in the semiconductor industry, at approximately 75%. In addition, Nvidia’s AI chips can cost several tens of thousands of dollars each, depending on the chip.
In light of the fact that the company decided to pass the cost of higher hardware prices to its customers, we can understand the growing pressure on the AI ecosystem.
This isn’t just a problem with Nvidia. Large tech firms have been hit by increased prices due to the rising demand for more sophisticated chips and other hardware.
Amazon, Microsoft, Google and Meta are building their own AI chips in an attempt to minimize dependency on Nvidia. Nonetheless, these companies buy hardware from the company and compete with each other for memory chips needed to make AI systems.
Nvidia Earnings Add to the Focus
The timing of the price increases becomes particularly relevant since the market is now keenly following the financial performance of Nvidia.
Nvidia is set to announce its earnings for the fiscal second quarter next week. The stock ended Friday’s trading session at $214.70 after dropping in value for six straight days, its longest run of losses since 2022.
Earnings reports might give investors a better understanding of how increased prices of memory and servers affect the company’s operations.
The critical point for investors is going to be whether increasing cost levels start to negatively affect Nvidia’s margins or whether the demand for AI technology allows Nvidia to pass on most of the cost increase to customers.
In case the demand for AI infrastructure stays strong, increasing cost levels can indicate that companies continue spending a lot on data centers.
This will add to the general conclusion that memory, advanced chips, and servers have become among the most crucial components of the AI infrastructure race.
Also Read | Pi Network Price Eyes $0.15 as PI Approaches Key Resistance
Be the first to comment