Recently, Samsung Electro-Mechanics announced that by 2026, the sales share of its high-end Flip Chip Ball Grid Array (FCBGA) substrates for server and artificial intelligence will exceed 50%.

FCBGA is an integrated circuit (IC) packaging technology，which involves flipping the chip and connecting it to the packaging substrate, then using spherical solder bumps to attach the package to the substrate.

It is mainly used in the packaging of high-density, high-speed, multi-functional large-scale IC chips, offering advantages such as high integration, small size, high performance, and low power consumption.

After a prolonged period of inventory cutting, the balance between semiconductor supply and demand sides has improved, with market demand gradually recovering.

The strong demand in fields such as high-speed network, server, smart driving, and optical module has continuously energized the development of high-multilayer high-speed boards and advanced HDI boards, which in turn is gradually boosting the prosperity of the packaging substrate industry.

As one of the main packaging methods for core electronic components like PC central processing unit, memory, and graphics processor, FCBGA boasts significant market potential in the development of 5G communications, artificial intelligence, virtual reality, and other fields.

Globally, IDM companies such as Micron, Infineon, and NXP have conducted extensive research and development in the FCBGA packaging field, while specialized packaging and testing companies like ASE Group, JCET, and Amkor have also developed various FCBGA technologies.

It is reported that numerous major international semiconductor companies, including Intel, Qualcomm, NVIDIA, AMD, and Samsung, are utilizing FCBGA technology.

Intel is one of the pioneers of FCBGA technology, first applying it to processors in 1997, while Apple is a loyal adopter of FCBGA technology, having used it in its processors from an early stage.

Data indicates that the global FCBGA packaging technology market will continue to grow rapidly in the coming years, with the market size expected to exceed USD 20 billion by 2026.

In face of such a highly potential opportunities, an increasingly more companies are channeling more efforts in developing FCBGA packaging technology, continuously facilitating its innovation and upgrade, and Chinese companies are also a part of this competition.

Currently, main companies engaging in FCBGA packaging substrates business in China include Fastprint, SCC, and FHEC (Forehope-elec), etc, which have disclosed their current progresses referring to FCBGA research and development.

Besides, Strongteam, a real estate company attempting to enter the semiconductor field, has set its sight on the FCBGA sector.

Fastprint disclosed that its low-layer FCBGA packaging substrates are currently in the small-batch delivery stage, with primary applications in the automotive and AI sectors.

SCC stated that it already has the capability of mass producing FCBGA packaging substrates with 16 layers and less, and the capability of sample manufacturing products with more than 16 layers.

The production line validation, sample delivery, and certification processes for various product levels have proceeded smoothly on track. Strongteam is actively transitioning into the semiconductor field and plans to invest in high-end FCBGA IC substrate enterprises.

Read more

• [News] Samsung Gains Early Market Entry Advantage in the Panel-Level Packaging Sector Ahead of TSMC
• [News] Intel Foundry Unveils its Expanded Process Roadmap for AI Era at Direct Connect 2024 Event

(Photo credit: Samsung)

Reuters previously reported that Intel is considering selling its stake in Altera, a FPGA (Field-Programmable Gate Array) manufacturer, as part of its business restructuring and cost-cutting efforts, as AMD and Marvell are said to be potential buyers.

As per a report from Economic Daily News citing sources, it’s believed that if the sale goes through, a significant portion of Altera’s orders could be redirected to TSMC, which would be highly beneficial for the Taiwanese foundry giant.

The same report indicated that Altera used to be a major customer of TSMC. However, after Intel acquired Altera in 2015, the orders were redirected to Intel. TSMC’s rapid growth, bolstered by orders from clients like Apple, AMD, and NVIDIA, helped mitigate the impact of losing Altera’s business though.

If Altera is no longer part of Intel, as it might be is acquired by companies like AMD or Marvell, which are currently key clients of TSMC, it is likely that Altera’s orders may return to TSMC in significant volumes.

Intel acquired Altera for USD 16.7 billion in 2015, and has previously indicated plans to sell a portion of its stake through an initial public offering (IPO), though no specific date has been set.

Citing sources familiar with the matter, Reuter’s report suggested that Intel’s plan does not currently include splitting up the company or selling its foundry business to buyers like TSMC, Reuters notes.

Intel had already begun segregating its wafer foundry business into an independent division and financials, starting from the first quarter of this year.

Per Reuters, the company has established a wall between its foundry and IC design business to ensure that the design division’s potential customers cannot access the confidential technologies of Intel’s foundry clients.

Read more

• [News] AMD Could Emerge as a Prospective Buyer for Intel’s Potential Sale of Altera
• [News] Intel Reportedly Mulls to Sell FPGA Unit Altera and Freeze USD 32 Billion German Project amid Crisis

(Photo credit: Intel)

AI chip giants NVIDIA and AMD have been under heated competition for a couple of years. NVIDIA, though controls the lion’s share of the market for AI computing solutions, had been challenged by AMD while the latter launched Instinct MI300X GPU in late 2023, claiming the product to be the fastest AI chip in the world, which beats NVIDIA’s H200 GPUs.

However, months after the launch of MI300X, an analysis by Richard’s Research Blog indicates that AMD’s MI300X’s cost is significantly higher than NVIDIA’s H200’s, while H200 outperforms MI300X by over 40% regarding inference production applications, which makes NVIDIA’s high margin justifiable.

AMD’s MI300X: More Transistors, More Memory Capacity, More Advanced Packaging…with a Higher Cost

The analysis further compares the chip specifications between the two best-selling products and explores their margins. NVIDIA’s H200 is implemented using TSMC’s N4 node with 80 billion transistors. On the other hand, AMD’s MI300X is built with 153 billion transistors, featuring TSMC’s 5nm process.

Furthermore, NVIDIA’s H200 features 141GB of HBM3e, while AMD’s MI300X is equipped with 192GB of HBM3. Regarding packaging techniques, while NVIDIA is using TSMC’s CoWoS 2.5D in the H200, AMD’s MI300X has been moved to CoWoS/SoIC 3D with a total of 20 dies/stacks, which significantly increases its complexity.

According to the analysis, under the same process, the number of transistors in the logic compute die and the total die size/total cost are roughly proportional. AMD’s MI300X, equipped with nearly twice the number of transistors compared to NVIDIA’s H200, therefore, is said to cost twice as much of the latter in this respect.

With 36% more memory capacity and much higher packaging complexity, AMD’s MI300X is said to suffer a significantly higher manufacturing cost than NVIDIA’s H200. It is also worth noting that as NVIDIA is currently the dominant HBM user in the market, the company must enjoy the advantage of lower procurement costs, the analysis suggests.

This is the price AMD has to pay for the high specifications of the MI300X, the analysis observes.

NVIDIA’s 80% Margin: High at First Glance, but Actually Justifiable

On the other hand, citing the results of MLPerf tests, the analysis notes that in practical deployment for inference production applications, the H200 outperforms the MI300X by over 40%. This means that if AMD wants to maintain a similar cost/performance ratio (which CSP customers will demand), the MI300X price must be about 30% lower than the H200. The scenario does not take other factors into consideration, including NVIDIA’s familiarity with secondary vendors, the Compute Unified Device Architecture (CUDA), as well as related software.

Therefore, the analysis further suggests that NVIDIA’s 80% gross margin, though might seem to be high at first glance, actually allows room for its competitors to survive. If NVIDIA were to price its products below a 70% margin, its rivals might struggle with negative operating profits.

In addition to achieving better product performance at a lower cost through superior hardware and software technology, NVIDIA excels at non-technical economic factors, including R&D and the scaling of expensive photomasks, which impact operational expenditures (OPEX) and cost distribution as well, while its long-term commitments to its clients, confidence, and time-to-market also play a role, the analysis notes.

Regarding the key takeaways from their latest earnings reports, NVIDIA claims the demand for Hopper remains strong, while Blackwell chips will potentially generate billions of dollars in revenue in the fourth quarter. AMD’s Instinct MI300 series, on the other hand, has emerged as a primary growth driver, as it is expected to generate more than USD 4.5 billion in sales this year.

Read more

• [News] NVIDIA’s Q3 Forecast Fails to Impress, while It Claims Blackwell to Bring Several Billion Dollars Revenue in Q4
• [News] Foxconn Confirms Shipment of NVIDIA’s GB200 to Start in Q4 as Planned

(Photo credit: NVIDIA)

The Wall Street Journal reported that OpenAI is in talks for a new round of funding, with tech giants Apple and NVIDIA both interested in investing in the AI research company OpenAI.

It’s reported that this investment will be part of OpenAI’s new round of financing, which will bring its estimated value to exceed USD 100 billion.

Sources indicated that OpenAI plans to raise billions of dollars, and venture capital firm Thrive Capital will lead this round of funding with a USD 1 billion investment. Microsoft, OpenAI’s largest shareholder, will also be a part of this round.

Reportedly, sources have revealed that Apple is currently in talks with OpenAI for the potential investment, while NVIDIA has already discussed joining the latest round of funding, who reportedly considered investing USD 100 million.

Although it is not yet clear how much Apple and Microsoft plan to invest, the point is that the three most valuable tech giants in the world would all become shareholders of OpenAI if these negotiations end in success.

In a memo on Wednesday, OpenAI’s CFO Sarah Friar stated that the company is seeking new financing but did not disclose specific details. Friar mentioned that OpenAI would leverage this funding to strengthen computing power and cover other operational expenses.

With the rise of the AI industry, Microsoft, Apple, and NVIDIA have also accelerated their pace in developing AI technologies.

Microsoft has invested USD 13 billion in OpenAI since 2019, holding a stake of 49% in this company. Apple, at its Worldwide Developers Conference (WWDC) in June this year, launched the Apple Intelligence system and announced a partnership with OpenAI.

As for NVIDIA, it has long been closely collaborating with OpenAI and has been highly active in making investment in this field. Its investment arm, NVentures, has invested in several AI companies since 2023.

Read more

• [News] NVIDIA’s Q3 Forecast Fails to Impress, while It Claims Blackwell to Bring Several Billion Dollars Revenue in Q4
• [News] NVIDIA’s Dominance Under Threat? Challengers Are Launching New Products to Compete

(Photo credit: OpenAI)

On August 27, AI server giant Supermicro was accused of accounting violations, inadequate disclosure of related party transactions, and evading sanctions by selling products to Russia by short-seller Hindenburg Research.

In addition, Supermicro announced on August 28 that it would delay the release of its annual report, potentially facing order withdrawals. Industry sources also believe this news presents a chance for Supermicro’s competitor Dell to gain market share.

Besides Dell, a report from Commercial Times also points out that Hewlett Packard Enterprise (HPE) could benefit from the shift in orders, potentially boosting shipments for its Taiwanese supply chain partners such as Wistron, Inventec, Quanta, and Foxconn.

The report from Commercial Times also cite sources, suggesting that this shift could provide Gigabyte, which is actively promoting its liquid-cooled products for NVIDIA’s H200 series, with  opportunities in the second half of the year.

Wistron, as a key supplier of motherboard and GPU accelerator cards for NVIDIA’s Hopper and Blackwell GPU, is not only a major supplier for Supermicro’s server motherboards but also for Dell. Its clients include HPE and Lenovo as well, which makes the company one of the primary beneficiaries.

Similarly, Inventec, one of the server motherboard suppliers, is also expected to benefit if the shift in orders boosts Dell, HPE, and Lenovo.

Moreover, one of Supermicro’s largest clients, CoreWeave, is transitioning to become a cloud computing service provider specializing in GPU-accelerated computing.

This shift has increased demand for GPU-accelerated computing and liquid cooling solutions. Reportedly, it’s believed that Gigabyte, which holds orders from CoreWeave, could be one of the biggest beneficiaries of the upcoming order shift.

Read more

• [News] Supermicro Faces Order Withdrawal Crisis, Potentially Benefiting Dell
• [News] Quick Couplings’ Buying Surge Reportedly Leads to AI Server Component Shortage Brought up by Supermicro

(Photo credit: Supermicro)