HIVE Digital Technologies said it has completed an inaugural research project using GPUs hosted in Asunción, Paraguay, in collaboration with Columbia University. The effort focused on iterative AI training runs performed remotely from New York and is now being submitted to NeurIPS, one of the major annual machine learning conferences.

For the AI infrastructure market, the development is less about a single model run and more about benchmarking and operational readiness. Independent, research-oriented testing can help quantify how well an installed GPU environment supports specific training workflows, including throughput and latency, and whether those results hold up under distributed usage patterns.

Columbia researchers use GPUs in Paraguay for NeurIPS-bound work

According to HIVE, the project centered on neural network pretraining research conducted by Columbia’s Department of Industrial Engineering and Operations Research. The team used HIVE GPU resources in Asunción while running iterative training experiments from New York.

The company framed the work as a “proof of concept” for intercontinental AI training, where researchers can run training jobs on geographically separated hardware. HIVE also said the research utilized code optimizations developed by Columbia to evaluate performance characteristics relevant to training workflows.

HIVE reported that, after normalization for each platform’s raw hardware performance, its A40 GPUs delivered training results comparable to newer-generation H100 GPUs in the targeted use case. The company described performance evaluation in terms of measured throughput and latency, along with tests for serving throughput and latency for a model configuration the team referenced as up to 1.4 billion parameters.

What the submission covers

While the filing does not provide full technical details in the release, it indicates the research addresses optimization methods for neural network pretraining under noisy conditions and explores accelerated algorithms designed to match performance characteristics of leading approaches.

HIVE’s announcement also notes that the work evaluated variants of the approach and tested performance for both training and serving settings, including standard throughput and latency tests related to LLaMA-style models. The reference to Muon in the release suggests the research is positioned within a broader trend in the field around scale-invariant optimization techniques intended to improve training efficiency and stability.

Performance benchmarks are used to plan larger HPC build-out

Beyond the academic milestone, HIVE said the measured performance data is intended to serve as a baseline for future expansion of its HPC and AI infrastructure in Paraguay, including what it describes as a “Gigafactory” for AI compute.

The company’s longer-term plan includes additional power and data center capacity at Yguazú, Paraguay. In the release, HIVE states that civil works for a 100 megawatt substation are complete, with commissioning expected during the summer and energization expected in September 2026. HIVE also said construction of a new Tier-III data center would begin in fall 2026, with an expected ready-for-service date in the second half of 2027.

For AI infrastructure operators, that timeline matters because GPU clusters require more than hardware procurement. They depend on predictable power delivery, cooling and redundancy design, network capacity, and operational workflows that can support both research experimentation and production workloads.

Why academic validation matters for AI infrastructure

GPU performance claims in the AI market are often difficult to compare across vendors, workloads, and environments. Even when systems use the same GPU model, results can differ due to software stacks, scheduling, network conditions, storage performance, and how optimizations are implemented.

In that context, HIVE’s emphasis on code optimizations and on measuring token-per-second, latency, and bandwidth reflects a practical approach to benchmarking. If the submitted NeurIPS work is presented publicly, other researchers and practitioners may be able to compare methodology, normalization choices, and evaluation settings.

However, it is still an evolving picture. A single collaborative study or an inaugural project generally does not establish a complete performance profile across all training regimes, model sizes, or operational constraints. What it can do is provide an early reference point for infrastructure readiness and for the feasibility of distributed training on installed hardware.

Market implications for Paraguay’s data center ambitions

HIVE’s expansion plan is also part of a broader push to localize AI compute capacity outside the most saturated markets. Paraguay has increasingly attracted attention due to its electricity profile and location, but the AI compute race is ultimately constrained by infrastructure, not just power availability.

By linking the research work to a planned substation and Tier-III data center, HIVE is positioning Paraguay as a location where international teams can access GPU compute for AI training. If the expanded infrastructure proceeds on the stated schedule, it may help reduce friction for researchers and enterprises seeking capacity without relying entirely on large hyperscale regions.

Still, timelines for construction and commissioning carry execution risk, and performance outcomes depend on ongoing software optimization and workload fit. The most relevant takeaway is that HIVE is using an academic collaboration to stress test both compute capabilities and the operational model for remote access to a distributed GPU cluster.

What to watch next

NeurIPS will be a key checkpoint for the work, as peer scrutiny may clarify technical assumptions and provide more granular evaluation results than the announcement itself. Investors and infrastructure buyers will likely also watch for updates on Yguazú commissioning milestones, as well as evidence that the performance baselines from Asunción translate to the scale of planned HPC deployments.

In the meantime, the release underscores an increasingly common pattern in AI infrastructure building, where operators seek credibility through benchmark-driven collaborations rather than purely marketing-led claims.

The Bank of England (BoE) published a policy statement and draft rules for systemic stablecoins on Monday, outlining how regulated pound-backed stablecoins would operate in the United Kingdom.

The BoE defines systemic stablecoins as those that are widely used in payments and may pose risks to the UK’s financial stability. HM Treasury is responsible for determining whether a stablecoin falls within the systemic regime.

Under the policy statement, systemic stablecoin issuers will be allowed to hold up to 70% of reserves in interest-bearing government debt, up from 60% under the previous proposal. Proposed holding limits have also been replaced with a temporary 40-billion-pound ($52.8 billion) issuance cap.

“This guardrail will be reviewed regularly and removed once risks to credit provision have been addressed,” the central bank said in a press release published on Monday.

The publication moves the UK closer to launching a dedicated regulatory framework for stablecoins, with the BoE aiming to finalize its rulebook by the end of 2026 ahead of a planned 2027 rollout.

Related: Critics tell UK Lords stablecoins are not future money

Bank shifts approach after industry feedback

The issuance guardrail replaces the holding limits proposed in the BoE’s November 2025 consultation, which would have limited individuals to 20,000 pounds per stablecoin and businesses to 10 million pounds per stablecoin.

Systemic stablecoins entail payments and retail-focused tokens. Source: Bank of England

At the time, the Bank argued the limits were needed to prevent large-scale shifts of deposits out of the banking system, which could reduce the availability of credit to households and businesses. Respondents to the consultation warned that the restrictions could limit the usability of stablecoins and create operational challenges for issuers.

The Bank said the new approach is intended to achieve the same policy objective while allowing unrestricted use by households and businesses.

The regime will apply only to stablecoins deemed systemic, while non-systemic stablecoins used mainly for crypto trading will remain under the Financial Conduct Authority’s supervision.

In May, Deputy Governor Sarah Breeden said the BoE was reconsidering its proposed holding limits and reserve requirements following feedback from digital asset companies, which argued that the restrictions could hinder adoption and make UK-issued stablecoins less competitive with dollar-backed rivals.

Cointelegraph is committed to independent, transparent journalism. This news article is produced in accordance with Cointelegraph’s Editorial Policy and aims to provide accurate and timely information. Readers are encouraged to verify information independently.

Peter Schiff has pushed back against Grant Cardone’s plan to combine real estate income with Bitcoin accumulation, arguing that the structure does not solve a real problem for property investors. 

The gold advocate made the comments after Cardone promoted a fund model that pairs income-producing properties with BTC holdings.

“Combining real estate with Bitcoin solves nothing,” Schiff wrote on X. 

He said Cardone’s argument rests on the idea that REITs need Bitcoin on their balance sheets so they can sell it later to pay for repairs and maintenance. Schiff rejected that view and said rental income already covers those ongoing costs.

Combining real estate with Bitcoin solves nothing. @GrantCardone claims REITs need Bitcoin on their balance sheets to sell so they can pay for repairs and maintenance. But real estate throws off rental income, which can be used to cover those ongoing costs. Happy to debate this.

— Peter Schiff (@PeterSchiff) June 21, 2026

Cardone fund pairs property income with BTC

Cardone Capital has been building a strategy that uses rental cash flow from multifamily properties to buy Bitcoin over time. The firm recently launched the $87.5 million 10X Space Coast Bitcoin Fund, which holds real estate and Bitcoin through a dedicated investment structure.

Cardone has argued that the model gives traditional investors exposure to Bitcoin without asking them to buy the asset directly. He has also said many investors in his Bitcoin-linked real estate funds did not previously hold crypto, making the structure a bridge between property investing and digital assets.

Meanwhile, the disagreement centers on whether Bitcoin adds value to a real estate model that already creates steady rental income. Cardone has criticized traditional real estate investment trusts because they must distribute at least 90% of taxable income to shareholders. In his view, that structure limits their ability to hold Bitcoin as a reserve asset.

Schiff disagrees with the reserve argument. He said property companies can use rental income for repairs, upkeep and maintenance instead of adding a volatile asset to the balance sheet. He also offered to debate Cardone on the topic, showing that the dispute has moved beyond a simple social media reply.

Broader Bitcoin treasury push continues

Cardone Capital has continued buying Bitcoin during market weakness. As previously reported by crypto.news, the firm bought another 282 BTC worth about $18 million as Bitcoin traded near $62,000. The purchase added to a position built through rental income from selected multifamily properties.

Moreover, as earlier reported, Cardone Capital held about 1,000 BTC after a $10 million purchase in January. The firm has targeted 3,000 BTC by the end of 2026 and 10,000 BTC over the longer term across multiple investment vehicles.

Real estate and Bitcoin remain a split topic

The debate reflects a broader split over Bitcoin treasury strategies. Supporters say Bitcoin can serve as a long-term reserve asset and may improve returns if property income funds steady purchases through market cycles.

Critics say the model adds price risk to an asset class that already has its own cash flow, debt, insurance and maintenance needs. For them, Bitcoin does not make real estate more efficient. It simply adds a new source of volatility.

A hawkish Fed. Rising bond yields. Concerns about Strategy (MSTR). Bitcoin BTC$64,045.81 already has plenty working against it. Now an ominous chart pattern is adding to the uncertainty.

The pattern is called a bear flag, and a breakdown could send the price of the largest cryptocurrency to as low as $54,000 initially, according to pseudonymous trader Doctor Profit, who called BTC’s bull-market peak at $126,000 and the subsequent selloff.

“Bitcoin is now forming a massive bearish flag on the daily timeframe,” the trader wrote on X. “My target is a dump to 54-56k region first before we move sideways once again and afterwards another leg down and the bottom is close in the region between 40-50k in my opinion.”

Drawn on a chart, the pattern looks like a flag on a pole that’s been flipped upside down. Here’s how it works: An asset drops sharply and then sees a relief bounce. The slide represents the pole and the bounce becomes the flag. When the price drops below the lower end of the flag, it deepens the selloff, with the downward move roughly the same size as the initial decline.

Key Takeaways

• Beijing designated 10 American enterprises for export restrictions, notably MP Materials and USA Rare Earth
• Restrictions prevent dual-use item exports from China to the designated companies
• China’s Finance Ministry simultaneously prohibited Chinese entities from purchasing from 46 U.S. corporations
• Measures represent Beijing’s retaliation after Pentagon designated Alibaba, Baidu, BYD, and other Chinese firms on its 1260H military roster
• Market experts characterize the restrictions as predominantly symbolic given minimal Chinese business ties for most affected firms

---

Beijing imposed export restrictions on 10 U.S. enterprises Monday, focusing on organizations connected to military operations, unmanned aerial systems, and critical mineral processing.

MP Materials and USA Rare Earth featured prominently among those designated. These organizations play crucial roles in the rare earth element extraction and magnet production pipeline, with MP Materials managing America’s sole operational rare earth mining facility.

The restrictions prohibit all dual-use merchandise exports from China to the designated enterprises. These items encompass products suitable for both commercial and defense purposes.

Applied Materials, Inc., AMAT

Additional companies facing restrictions include unmanned aircraft manufacturers Teal Drones and Jaia Robotics, electronic systems producer Aveox, Ball Aerospace and Technologies, plus Oshkosh Defense.

China’s Answer to U.S. Military Entity Designation

China’s Commerce Ministry justified the restrictions as necessary for protecting sovereign security interests and meeting global commitments. Officials characterized the decision as a response to Washington’s “antagonistic conduct.”

The Pentagon’s recent 1260H list update identified Chinese corporations allegedly supporting Beijing’s military apparatus. Notable recent inclusions featured Alibaba, Baidu, BYD, and NIO.

Beijing’s action represents a calculated response to that designation.

Simultaneously, China’s Finance Ministry announced procurement prohibitions preventing Chinese purchasers from acquiring goods from 46 American corporations, predominantly defense industry participants. Foreign-invested enterprises operating domestically in China with connections to those organizations remain unaffected.

Financial Markets Show Minimal Response

Equity markets demonstrated little reaction to the announcement. MP Materials and USA Rare Earth stock prices remained essentially stable after the disclosure.

Industry observers suggest the concrete ramifications of these restrictions remain constrained. The majority of designated American enterprises maintain negligible or nonexistent commercial operations in China.

George Chen, a partner with the Asia Group, characterized Beijing’s action as “measured” and “predominantly ceremonial.” He observed that most designated organizations focus on defense applications and maintained minimal Chinese trade relationships previously.

Han Shen Lin, another Asia Group partner, supported this assessment, noting the affected enterprises possess “minimal or zero substantial Chinese business footprint.”

The restrictions don’t generate immediate financial losses for most designated organizations.

Nevertheless, policy trajectory remains significant for market participants. Beijing demonstrates capacity to counter American blacklists with reciprocal limitations, particularly regarding defense technology, unmanned systems, and strategic minerals.

Organizations involved with rare earth elements and military procurement networks might gain advantages from sustained American initiatives to diminish dependence on China for essential materials.

Yet the commercial landscape grows increasingly intricate as both nations continue expanding their national security mechanisms.

This development continues an established sequence of reciprocal trade measures between Washington and Beijing that has intensified during 2026.

That key is meant to stay sealed inside secure hardware so the proofs can be trusted. With it exposed, the attacker could enroll their own provers as legitimate and sign fraudulent proofs that Taiko’s verifier accepted, then fake a bridge withdrawal that released real assets on Ethereum.

Taiko urged all users to withdraw from every bridge on the network, asked centralized exchanges to suspend deposits of its TAIKO token, and had its block producers stop making new blocks during the investigation.

By about 2 a.m. ET it said the exploit was contained and withdrawals through the main bridge and token vault were fully stopped. The exploiter had already moved about 2 million TAIKO, worth roughly $170,000, to an account on the MEXC exchange.

The dollar loss is small, but the flaw came from the same DeFi mechanism that have caused hundreds of millions worth of losses this year.

Forged cross-chain messages drained $292 million from Kelp DAO’s bridge in April and $11.4 million from the Verus-Ethereum bridge in May, the same failure where one chain is tricked into trusting a fake instruction from another. Bridges have produced more than $340 million in losses across at least 14 exploits in 2026, making it the costliest target in crypto. Taiko’s damage stayed contained mainly because the team caught and froze it within hours.

XRP price traded near $1.13 on June 22 after briefly slipping to about $1.12 during Sunday’s session. 

Buyers stepped in near that level and pushed the token back toward $1.15 within hours, keeping attention on the lower end of the range.

The move kept XRP inside the broad $1.10-$1.30 band that has guided price action for most of June. The token was down over 4% for the week and more than 13% over the past month, showing that the short-term rebound has not erased the wider weakness.

crypto.news data showed 24-hour volume near $1.28 billion, with XRP ranked sixth by market value. Its market capitalization stood near $70.28 billion, while fully diluted value remained above $113 billion. Circulating supply was about 62.05 billion XRP from a maximum supply of 100 billion tokens.

The support test matters because XRP has already struggled to hold higher levels this month. A prior move below $1.15 turned that area into the first resistance zone. Bulls now need to regain $1.15, then $1.20, before a stronger recovery setup can form.

XRP indicators show early recovery signs

The MACD shows a mild bullish turn. The histogram is slightly positive near 0.0045, while the MACD line sits around -0.0379 and above the signal line near -0.0424. That setup points to weaker bearish momentum and a short-term recovery attempt.

The signal remains early because both MACD lines are still below the zero line. That means momentum has not moved fully back into bullish territory. XRP needs stronger follow-through before traders can treat the setup as a confirmed reversal.

The RSI stands near 40.51, slightly above its moving average of 39.81. This shows some improvement from weaker levels, but the reading remains below the neutral 50 mark. Buying strength is present, but still limited.

A move above 50 on the RSI would give bulls a cleaner technical signal. Until then, the chart still favors caution. The token is no longer showing heavy downside pressure, but it has not yet shown enough strength to confirm a new uptrend.

Flows and derivatives activity improve

Fund flows offer one of the more supportive signals for XRP. As previously reported, XRP-linked products recorded about $10.66 million in weekly net inflows for the week ending June 18. That was close to the prior week’s $10.68 million.

Cumulative net inflows rose to about $1.45 billion, while total net assets moved closer to $1 billion. These figures show that institutional-style demand has not disappeared, even as the spot price trades well below last year’s highs.

Derivatives activity also picked up. Coinglass data showed XRP volume rising 50.17% to $2.08 billion, while open interest increased 1.23% to $2.66 billion. Options volume rose 19.06% to about $609,170, and options open interest increased 0.75% to $65.47 million.

Higher volume and open interest can support sharper price moves, but they do not show direction by themselves. If long positions build while spot demand stays weak, volatility can rise on both sides. Traders will watch whether open interest grows with price recovery or with another failed bounce.

Analysts watch $1.36 and $1.08

Analysts remain split on whether XRP is building a base or forming another pause inside a downtrend. Javon Marks said XRP’s breakout remains valid and kept a long-term measured move target near $17. He wrote that traders are watching for “another >12X” move if the setup continues.

$XRP‘s breakout stands which means the measured move target near $17 does as well.

In response to this break, we’re watching for an another >12X (1,100% Increase) to reach it.

Advertisement

Keep in mind, we precisely called the move from ~$0.50 to $3.30+.

The move just doesn’t look done… pic.twitter.com/vdVi4pdAMs

— JAVON⚡️MARKS (@JavonTM1) June 21, 2026

That target remains a projection, not a confirmed path. XRP would first need to clear several nearer resistance levels, including $1.15, $1.20 and $1.30. The larger bullish case becomes harder to defend if the token loses the lower range.

Another analyst using the name Batman pointed to a short-term compression phase. He said XRP still has an ascending demand trendline while descending resistance squeezes price action. He set the “breakout threshold” at $1.36 and the “invalidation” level at $1.08.

Those levels give traders a clear map. A move above $1.36 would suggest that buyers have taken control of the range. A loss of $1.08 would weaken the structure and could open the door to a deeper support test.

For now, XRP’s position remains mixed. The token has defended $1.12, flows have improved, and MACD is turning slightly higher. At the same time, RSI remains weak, price stays below neutral momentum levels, and the wider trend has not recovered.

The next move depends on whether buyers can turn the rebound into a sustained close above $1.15 and $1.20. If they fail, XRP may keep moving sideways near support. If selling returns below $1.10, the $1.08 invalidation level could become the next test.

Financial regulators in South Korea are pushing for broader reporting requirements on crypto transfers to further align with global Anti-Money Laundering standards for digital assets.

South Korea’s Financial Intelligence Unit (FIU) raised proposals to expand the Financial Action Task Force’s (FATF) Travel Rule requirements to smaller crypto transfers during a plenary meeting in Paris last week, according to an announcement on Monday.

The crypto Travel Rule is a global AML standard that requires crypto exchanges to share sender and recipient information for transfers above certain thresholds. It is designed to improve the traceability of funds moving between platforms.

South Korea already applies Travel Rule requirements to crypto transfers above 1 million won ($650), and the latest proposal calls for extending those obligations to smaller transactions.

Ongoing gaps in global oversight and DeFi risks

The FIU said Travel Rule obligations should apply to both originating and receiving crypto asset service providers (CASPs) to close gaps in cross-border transfers.

The FIU also called for stronger action against offshore and unregistered crypto platforms, citing increased misuse in illicit finance cases and risks of regulatory arbitrage.

FIU Commissioner Lee Hyung Ju at the FATF plenary session in Paris. Source: FIU

Beyond the Travel Rule discussion, FATF also approved a new report examining risks associated with decentralized finance (DeFi), according to the FIU.

Related: South Korea police raid Bithumb over lawmaker hiring favoritism probe: report

FIU Commissioner Lee Hyung Ju welcomed the adoption of a DeFi-related report during FATF discussions. However, he said regulatory arbitrage across jurisdictions mainly stems from differences in licensing, supervision and offshore oversight.

Seven years after FATF extended Travel Rule scope to crypto

The proposal was part of broader discussions on the implementation of FATF Recommendation 15, the international standard updated in 2019 to apply AML measures to crypto assets and CASPs.

Seven years after FATF extended its AML framework to cover crypto assets, global implementation of Recommendation 15 remains uneven, according to a targeted update by FATF in 2025.

Source: FATF

The FATF assessment found that 49% of jurisdictions were only partially compliant with requirements for CASPs, while 21% remained non-compliant as of April 2025, leaving only about 29% of jurisdictions rated largely compliant or compliant.

Magazine: China’s 107 Bitcoin memory thief, Bithumb CEO booked: Asia Express