2 cent

leadership of AI tokens

· Growth and capital inflow
The AI sector shows steady growth, adding billions to total market capitalization and attracting liquidity.
· Institutional recognition
Regulated investment products (e.g., Grayscale GTAO trust) have been launched on TAO, attracting significant capital.
· Technology and scarcity
The scarcity of TAO is supported by the first halving (50% reduction in emission), and Render represents a scarce resource—GPU computing power.
· Real utility (Utility)
Both projects solve practical problems: Bittensor creates a marketplace for AI models, and Render creates a marketplace for computing power. This aligns with the trend toward projects with tangible value.

Two perspectives on one technological revolution

· Choose Bittensor (TAO) if you believe in the future of a decentralized AI model and knowledge market, where value is created by algorithms and data.
· Choose Render (RENDER) if you believe in the future of decentralized physical infrastructure (DePIN), where value is created by distributed GPU power.

High technologies in the crypto industry

Projects Bittensor (TAO) and Render (RENDER) are technologically distinct but complementary leaders in the segment of decentralized computing and artificial intelligence (AI) in the crypto industry.

Simply put: TAO trades 'intelligence' of AI, while RENDER provides 'muscles' (computational power) for its operation

Thus, the transition of "miners" to TAO and Render is not a pool change, but the emergence of a new industry of decentralized computing. This segment is rapidly growing amid the demand for AI and has significant potential, but requires specialized knowledge and equipment. $TAO $RENDER

If you look at the situation in more detail, for one DDR5 module of the specified capacity in China, they ask between $5,000 and $7,000. In essence, when purchasing for data center needs, 100 such modules are placed in one box, and in this case, the cost already reaches around $700,000. For a comparable amount in Shanghai — not the cheapest Chinese megacity — residential homes are offered. Meanwhile, in many cases, when purchasing real estate in this city, you can even save compared to the mentioned amount.$TAO $RENDER

Balancing between the enormous demand from AI and the needs of the consumer market will be extremely difficult for manufacturers for several more years.

Users should prepare for the fact that a wide range of configurations and low prices for memory are unlikely to return in the near future. The industry has entered a prolonged period of structural deficit $RENDER $TAO

Micron says the memory market crisis won't end before 2028
Users should prepare for a prolonged acute shortage of certain components.
Amid the frenzy demand for AI systems memory, Micron, one of the key global memory producers, has found itself caught in a crossfire. Consumers blame the vendor for shortages and rising prices, believing it has abandoned the PC market for super profits from artificial intelligence.
In an interview with Wccftech, Micron's vice president Christopher Muro explained why the DRAM module shortage has become a structural issue and will last at least until 2028.

$TAO $RENDER

Instead of sporadic validator activity, current interest in TAO is likely driven by fundamental factors: scarcity due to the halving, growing institutional demand (especially in light of the ETF news), and overall confidence in the potential of decentralized AI.

Halving event: In December 2025, the TAO halving occurred, cutting the daily issuance of new tokens in half (from 7200 to 3600). Such events, creating scarcity of new supply, have historically supported price.
· Institutional interest: In early 2026, the TAO price rose 27% in a week. One of the main catalysts was Grayscale's application for the launch of the first U.S. ETF on Bittensor. Such steps

Data indicates that the majority of wallets that purchased TAO in the past 7 months are still not profitable or are close to breakeven. This reduces selling pressure, as investors typically avoid locking in losses.
· Limited issuance: Like Bitcoin, TAO has a capped supply of 21 million tokens. The scarcity of the asset attracts long-term investors.

to bold analyst predictions about the potential testing of levels at $3000-$5000.

According to forecasts, the retail price of Nvidia's flagship accelerator — the GeForce RTX 5090 — will rise to $5,000. Considering the recommended starting price was $2,000, this represents a 2.5-fold increase. It is noted that production costs for graphics cards have increased by 80% solely due to rising component prices for memory.

Rumors that Nvidia is reducing production of mid-range models — the GeForce RTX 5060 Ti and RTX 5070 — turned out to be only partially true. The production of accelerators is proceeding as usual, but the cards are not reaching stores. The majority of shipments are being intercepted by "custom AI factories" in Asia. These factories purchase gaming cards by the thousands, equip them with turbine-type cooling systems, and resell them to companies willing to use such accelerators for AI training. This creates an artificial shortage.

The source Newsis, citing insiders, reports that AMD will begin raising prices already in January 2026, with Nvidia following suit in February. The price increases will affect current product lines — Nvidia RTX 50 (Blackwell) and AMD Radeon RX 9000 (RDNA 4). Moreover, this will not be a one-time promotion: prices will be indexed monthly$TAO $RENDER

For example, some DDR5 memory kits with a capacity of 64 GB are now more expensive than a new graphics card – traditionally the most expensive component in a computer.

Previous-generation DDR4 memory, which is still being sold, is also becoming more expensive, although not as drastically; markups have already reached up to 50% per kit compared to prices in summer 2025.$TAO $RENDER

Memory prices have risen, however the latest news turned out worse than analysts had expected
The memory shortage will last until 2027-2028, it's already impossible to buy RAM modules at a reasonable price, and all this smells like a mining boom, but on a much larger scale.$TAO $RENDER

Japan has made a powerful leap, announcing the creation of a national AI initiative with a budget of $19 billion, led by SoftBank. The main goals are to catch up with the United States and China, ensure technological sovereignty, and give a boost to the economy. Construction of data centers and procurement of NVIDIA chips have already begun, with initial results expected in the public sector by 2026. This national project confirms that the 'AI boom' in the country is a reality at the governmental level.

Yomiuri: Japan to allocate $19 billion for development of national AI model

The Japanese government, together with the private sector, plans to spend around 3 trillion yen (approximately $19 billion at the current exchange rate) to create a national artificial intelligence (AI) model. The newspaper Yomiuri reported this.
It is noted that over 10 local companies, including giants such as SoftBank Corporation, will participate in the joint development. According to the publication, the government intends to cover about one-third of the costs, while the remaining funds will need to be invested by private firms.
It is expected that the development of Japan's own AI will begin as early as 2026. This will include, in particular, the construction of state-of-the-art data centers. In the long term, this AI model is also expected to be used in locally produced robotics.
$TAO $RENDER

The AI boom has made Tokyo Electron the third most valuable company in Japan — Sony has dropped to fourth place

Over the next five years, Tokyo Electron plans to increase capital expenditures and R&D spending by 80–90%, and hire up to 10,000 new employees. According to Jefferies analysts, Tokyo Electron's assets are not overvalued even at current levels$TAO $RENDER

🇯🇵 Japan: Retailers Are Really Buying Used PCs

In Japan, major stores have faced a severe shortage of components and complete systems, prompting them to turn to customers. The most striking example is the large Sofmap Gaming store in Tokyo's Akihabara district, which publicly asked customers to sell their old computers in January 2026.

· What they buy: They are willing to purchase almost any PC — gaming desktops, laptops, and even regular non-gaming models.
· Reason: Store shelves are empty, and new shipments are limited due to supply chain issues and high memory prices.$TAO $RENDER

Sharp increase in the cost of DDR5 standard RAM has given a second life to the AM4 platform. Journalists from the specialized portal Wccftech report this, citing sales data for CPUs from various overseas retailers.

For example, over the past week, the share of AM4 chips in sales at the large German retailer Mindfactory increased from 23 to 34%. Meanwhile, two processors for the older socket made it into the top five bestsellers. These chips are the Ryzen 5700X and Ryzen 5800X.

processors for the AM4 socket are indeed experiencing a surge in popularity, and as a result, sales and prices are rising (especially in foreign stores) due to the shortage and increased cost of DDR5 memory; AMD is even considering returning older models to the market to support users remaining on the DDR4 platform, which has become more attractive. This gives a second life to older, yet still powerful chips such as the Ryzen 5000 series, which are excellent for upgrading existing systems.
Why this is happening:
Rising DDR5 prices: More expensive DDR5 memory is pushing users toward the cheaper DDR4, thus toward the AM4 platform.
Cost-effective upgrade: For many users, switching to Ryzen 5000 with an AM4 motherboard and DDR4 memory is the best option for a significant performance boost without fully changing the platform.
Ongoing AMD support: AMD recognizes the value of AM4 and is considering releasing new or reviving production of older CPUs (such as the Ryzen 5000 series) to meet demand.
Key processors in demand:
Ryzen 5 5600, Ryzen 7 5700X, Ryzen 7 5800X/XT, Ryzen 9 5900X: These models have become bestsellers, offering excellent price-to-performance ratios.

processors for LGA 1700 socket (12th, 13th, and 14th generation Intel) may become more expensive, but the reason is not their popularity, rather global issues in the DRAM memory market, which causes DDR4 to become more expensive, and motherboards for LGA 1700 (especially DDR4) become scarcer and more expensive, indirectly affecting the cost of building systems with these processors. $TAO $RENDER

ASUS is trying to combat the DRAM shortage in the market by increasing production of motherboards supporting the older DDR4 memory standard. This is reported by the Wccftech portal, citing insider information from the Board Channels forum.

According to their data, the manufacturer plans to significantly increase shipments of AM4 motherboards with B550 and A520 chipsets in the first quarter of 2026. The company also intends to notably boost production of LGA 1700 motherboards based on B760M and H610M-G chipsets, supporting DDR4 memory.$TAO $RENDER

🌍 Global Context and Other Regions

The issue is not limited to Japan. It is part of a global shortage driven by speculative demand for AI components.

· Price surge: Since mid-2025, prices for DDR5 memory on the global market have increased 3.5 to 4 times, and SSD prices have roughly doubled.
· Non-market measures: In Taiwan, a key region for electronics manufacturing, some distributors have introduced unprecedented rules to combat the shortage: buyers can obtain DDR5 memory only by purchasing a motherboard in a 1:1 ratio.