Let’s face it: most of the best graphics cardsare already extremely costly. But think about what would happen if conditions deteriorated further? What a terrible idea.
Regrettably, based on recent reports, various high-end chips could become more expensive, beginning with the upcoming product generation. This is likely due to TSMC reportedly increasing the costs of its next-generation process node.
TSMC’s new generation of wafers could see a price increase of as much as 50%
TSMC is the major player behind the products developed by companies such as Nvidia, Apple, AMD, and even Intel, serving as the world’s leading contract chip producer. Its “process nodes,” including N3 or N2, represent advanced manufacturing technologies that reduce transistor size, resulting in faster chips, improved energy efficiency, and typically higher production costs.
As the leading supplier of chips globally, TSMC secures significant business and enjoys considerable autonomy in determining its pricing. This further supports the concerning report fromChina Times, shared by Dan Nystedt on X.
The newest Nvidia graphics chips are manufactured using a 5nm process, however TSMC has already started producing a 3nm node, with 2nm expected in the near future. These more advanced nodes are now anticipated to significantly increase in cost.
The N3P 3nm process is reportedly approximately 20% pricier than the prior (5nm) version. In contrast, the 2nm technology is now expected to raise costs by a significant 50%—and this is compared to the N3P process, not the 5nm node that is currently found in consumer devices. The overall price hike could be substantial.
Nystedt states that TSMC’s clients, Qualcomm and MediaTek, have already paid much higher prices for the N3P process; Qualcomm is reported to have paid an additional 16%, while MediaTek: An impressive 24%. This pricing variation can result from various reasons, and receiving volume-based reductions is common within this industry. Nevertheless, even major companies like Nvidia may find it difficult to secure a significant discount from TSMC this time.
The China Times reports that N2 output has been strong, indicating there will be sufficient quantities available. However, due to high demand, TSMC doesn’t have to compromise with any particular client — it can set prices as high as possible.
This is unfortunate for gamers.
Who will pay for these rising prices?
Chips begin their existence as silicon wafers, which are thin sections of extremely pure silicon that are manufactured with billions of tiny transistors. These wafers are subsequently divided into separate chips, which serve as the central processing units in CPUs, GPUs, and a wide range of other devices.
In a graphics card or processor, multiple chips can be combined, occasionally using a “chiplet” design where several smaller dies are connected rather than depending on a single large piece of silicon. This method helps lower expenses, enhance production efficiency, and optimize the use of each wafer, which becomes increasingly significant as wafer costs rise.
For corporations such as AMD, Intel, and Nvidia, these increased costs present challenging decisions. They have the option to either take on the extra cost themselves, transfer it to customers by increasing product prices, or seek out efficiencies in other areas, like shrinking the die size or utilizing chiplets more extensively. A smaller die size does not automatically result in reduced performance, since advancements in manufacturing allow engineers to fit more transistors into the same space. This usually leads to improved speeds and better efficiency, even if the chip occupies less physical space.
So, if the report is accurate, can we anticipate higher prices in the consumer market, beginning with the next generation of CPUs and GPUs? It’s uncertain. Without an official statement from TSMC or one of its clients, we can only guess. That said, I wouldn’t be surprised if prices rise once GPUs move to the 2nm process.
Gaming is becoming less accessible for PC users due to rising costs.
The issue here is that there’s very little flexibility remaining regarding components designed for everyday users.
If TSMC actually increases the price of its wafers by 50% in the next generation, that would be a significant increase — but GPUs and CPUs are already costly. Nvidia’sRTX 5090begins at $2,000, but the actual cost is typically around $2,500 or more. Then there’s theRTX 5080that’s valued from $999 to $1,300, and theRTX 5070 Ti, which you’ll frequently see at its $749 MSRP these days.
That’s a significant amount of money for a single part. If these price hikes come to pass, many gamers may need to avoid purchasing top-tier GPUs and instead go for more standard options, such as theRTX 5060.
The positive aspect is that frame generation technology enhances the capabilities of these more affordable GPUs — however, it isn’t available in all games, and it’s possible that you’d prefer a GPU that can handle games without depending on AI. Nevertheless, this is the current reality, and due to increasing prices, we’ll all need to adopt technologies like DLSS and FSR even more frequently.
I can envision that the price hike would mainly affect the top-tier models, with significant cost changes for the RTX xx90 and RTX xx80. Nevertheless, the lower models might also experience increases, and if the RTX xx60 rises from $300 to $400, that would still be a considerable issue.
At this moment, there’s no reason to worry, and there’s definitely no need to start purchasing GPUs now out of concern for rising prices. It will take some time before the 2nm process is applied to consumer products. Nevertheless, when that time arrives, we can only hope that manufacturers are fair and do not shift the full burden of the increased costs to consumers.
Latest Technology News, Sent to Your Email
