HIVE Digital Technologies became the latest bitcoin miner to lean fully into the AI compute boom on September 10, 2026, when its wholly owned subsidiary BUZZ High Performance Computing (BUZZ HPC) named AI and HPC industry veteran Mark Volk as Senior Vice President, Revenue. The announcement, made from the company’s San Antonio, Texas office and confirmed by outlets including StockTitan and The Globe and Mail, came bundled with a bigger number: HIVE says its combined bitcoin mining and GPU cloud operations have now surpassed $1 million in average daily revenue.
That figure matters beyond HIVE’s own balance sheet. It’s a data point in a much larger story playing out across the crypto-mining sector, where companies that spent years stacking ASIC rigs are now racing to convert spare megawatts and data-center shells into AI compute capacity. HIVE, listed on the TSX and Nasdaq under the ticker HIVE (BVC: HIVECO, FRA: YO0), is one of the more aggressive movers in that shift, and the Volk hire signals it’s now staffing up specifically to sell that capacity rather than just build it.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What HIVE Announced on September 10
According to the company’s official release, BUZZ HPC brought on Mark Volk as Senior Vice President, Revenue, tasking him with leading the subsidiary’s global go-to-market strategy. His mandate spans enterprise, government, sovereign AI, hyperscale, and research and academic customers, a deliberately broad list that reflects how HIVE wants to sell its GPU capacity: not just to one type of buyer, but to whoever needs compute and is willing to sign a long-duration contract for it.
Volk reports to Craig Tavares, BUZZ HPC’s President and Chief Operating Officer, and will be responsible for commercializing the unit’s GPU-powered infrastructure across four product lines: sovereign AI cloud, GPU-as-a-Service, colocation, and hybrid AI solutions. The company’s public materials describe him as an “AI and HPC veteran” without detailing his prior employers, so his exact career history before BUZZ HPC isn’t part of the confirmed record.
The same release disclosed that HIVE’s combined hashrate services (bitcoin mining) and GPU cloud services have surpassed $1 million in average daily revenue. Since August 21, 2026, HIVE says it has produced an average of approximately 12 bitcoin per day, representing roughly 2% of global bitcoin network production, while its GPU cloud operations have generated approximately $100,000 in average daily revenue on their own.
Who Is Mark Volk, and What He’ll Actually Do
Volk described his own job in blunt commercial terms. “My mission is to build the commercial engine around it: win customers, build partnerships and convert our infrastructure into large-scale, long-duration contracts across government, enterprise and hyperscale markets,” Volk said in HIVE’s September 10 announcement. He also framed the broader opportunity in infrastructure terms, saying, “Compute is becoming one of the world’s most strategically important infrastructure assets.”
That framing lines up with how HIVE’s leadership is talking about the hire. Frank Holmes, HIVE’s Executive Chairman, put it plainly in comments carried by Proactive Investors on September 12: “We have the infrastructure and Mark’s going to build that revenue engine for us.” The quote captures the split that now defines HIVE’s business: one team builds and operates data centers and GPU clusters, another sells access to them. Volk is being brought in specifically to run the second half.
This is a familiar pattern in the AI infrastructure buildout. Hardware and power capacity have become commodities that need dedicated sales organizations to monetize efficiently, the same shift that’s driving demand for GPU allocation deals covered in reporting on Megaspeed’s $2 billion Nvidia chip purchase and similar moves by Asian compute resellers.
Breaking Down the $1 Million-a-Day Number
The headline figure blends two very different businesses. Bitcoin mining still supplies the bulk of it: at roughly 12 BTC produced per day, and with bitcoin trading in a range that keeps daily mining revenue in the high hundreds of thousands of dollars, hashrate services remain HIVE’s larger single revenue stream. GPU cloud services contribute the remainder, at an average of approximately $100,000 per day as of the September disclosure.
That split is worth sitting with, because it tells you where HIVE actually is in its transition, not where its press releases suggest it’s heading. GPU cloud revenue is growing far faster than mining revenue in percentage terms, but it’s still the smaller slice of the pie in absolute dollars. The Volk hire is a bet that the ratio flips over the next few years as BUZZ HPC’s contracted capacity comes online.
From Bitcoin Rigs to GPU Racks: The Pivot Explained
HIVE describes itself as a global provider of sustainable digital infrastructure and AI compute, and its strategy is straightforward on paper: take power contracts and data-center sites originally built for bitcoin mining, and layer GPU clusters on top of the same electrical and cooling infrastructure. Mining rigs and AI accelerators have very different compute profiles, but they share the two things that are hardest to secure quickly: cheap power and physical space with adequate cooling.
BUZZ HPC is the vehicle for that second business line. It’s a wholly owned subsidiary of HIVE Digital Technologies, and its financials are broken out separately in HIVE’s quarterly and annual filings, which is how outside observers can track the pivot’s actual progress rather than taking the “AI compute” branding at face value.
FY2026 Financial Results: The Numbers Behind the Headline
HIVE’s fiscal year 2026 results, published in its FY2026 earnings release, show total company revenue of $297.8 million, up 158% year-over-year. BUZZ HPC’s slice of that was $19.5 million, up 94% from $10.0 million in FY2025. HIVE attributes that growth to the deployment of an Nvidia H200 GPU cluster and strong demand through a GPU marketplace it uses to sell spare capacity.
| Metric | FY2025 | FY2026 | Change |
|---|---|---|---|
| Total company revenue | Not separately broken out | $297.8 million | +158% YoY |
| BUZZ HPC (GPU/AI) revenue | $10.0 million | $19.5 million | +94% YoY |
| Q3 FY2026 total revenue | — | $93.1 million | Record quarter |
| Q3 FY2026 gross operating margin | — | $32.1 million | Over 6x YoY |
| Q3 FY2026 BUZZ HPC revenue | — | $4.9 million | Against $2.3M direct costs |
Those numbers explain why HIVE is investing in a dedicated revenue organization now rather than later. A 94% year-over-year jump in HPC revenue is a meaningful growth rate, but $19.5 million is still a small business relative to the company’s overall $297.8 million top line. Scaling that further requires exactly the kind of enterprise and government sales relationships Volk has been hired to build.
Q3 FY2026: Six Times the Margin
HIVE’s most recent quarterly results, detailed in its Q3 FY2026 release, show revenue of $93.1 million and gross operating margin of $32.1 million, more than six times higher than the year-ago quarter. Within that, BUZZ HPC generated $4.9 million in revenue against direct costs of $2.3 million, a gross margin profile that, if it holds at scale, would make the HPC business meaningfully more profitable per dollar of revenue than mining once utilization climbs.
The quarter also included a specific catalyst for the GPU business: in February 2026, HIVE signed a two-year, $30 million contract for 504 Nvidia B200 GPUs, expected to go live at a Bell Canada Tier-III facility. HIVE says that single deal adds roughly $15 million in annual recurring revenue and lifts BUZZ HPC’s annualized HPC revenue run rate from around $20 million to around $35 million.
Inside BUZZ HPC’s Global Infrastructure Footprint
HIVE’s compute buildout spans three regions, each at a different stage of maturity. In Paraguay, HIVE’s BUZZ AI Cloud platform in Asunción is already operational inside a Tier-III data center run by the country’s largest telecom provider, and the company has publicized a live workload from a Columbia University research team using the site for large language model research. In Canada, the Bell Canada partnership and a planned conversion of a New Brunswick facility form the near-term colocation growth path. A third site, the Yguazú campus in Paraguay, is HIVE’s longer-horizon bet on hydroelectric-powered AI infrastructure.
| Site | Country | Status | Key Detail |
|---|---|---|---|
| Asunción BUZZ AI Cloud | Paraguay | Operational | Tier-III facility hosting Columbia University LLM research workloads |
| Yguazú AI Gigafactory | Paraguay | Under construction | 100 MW substation, civil works complete, energization targeted September 2026 |
| Bell Canada Tier-III facility | Canada | Deployment underway | 504 Nvidia B200 GPUs under a 2-year, $30 million contract |
| New Brunswick facility | Canada | Planned conversion | 70 MW Tier-I site being converted to 50 MW of Tier-III colocation IT load |
HIVE has also flagged plans for an additional 100 MW hydroelectric-powered data center at Yguazú, with commissioning targeted for the third calendar quarter of 2026, and has described a broader goal of a fivefold increase in HPC capacity during 2026 through partnerships including the Bell Canada arrangement. Whether those timelines hold is worth tracking independently of the company’s own announcements, since construction and energization schedules in large data-center projects frequently slip.
The Bell Canada Deal and Why 504 GPUs Matters
The Bell Canada contract is small in absolute GPU count next to what hyperscalers deploy, but it’s a useful proof point for HIVE’s colocation model. Nvidia’s B200 accelerators are among the current generation of AI training and inference chips, and locking in 504 of them under a two-year contract gives BUZZ HPC a predictable revenue base rather than relying purely on spot GPU marketplace demand. It’s the same logic driving broader industry moves toward long-term compute contracts, a trend also visible in how Chinese server makers have been securing Nvidia chip supply through intermediary arrangements, albeit for very different reasons.
HIVE’s stated target is to reach roughly 11,000 GPUs and about $140 million in GPU AI Cloud annual recurring revenue by the fourth quarter of 2026, with total HPC ARR, including Tier-III colocation revenue from the New Brunswick conversion, targeted at approximately $225 million by the end of 2026 or early 2027. Both figures are company projections tied to deployment timing and market demand, not confirmed results, and HIVE’s own materials note they’re subject to change.
Paraguay: HIVE’s Hydroelectric AI Bet
Paraguay sits at the center of HIVE’s long-term AI ambitions largely because of cheap hydroelectric power, the same resource that has attracted bitcoin miners to the country for years. The Yguazú site’s 100 MW substation was reported with civil works complete and energization expected in September 2026, with a new Tier-III data center construction phase set to begin in the fall and reach ready-for-service status in the second half of 2027. That’s a multi-year runway, and it puts Paraguay on a different timeline than the more near-term Bell Canada colocation deal.
The strategic logic mirrors what’s playing out with sovereign cloud initiatives elsewhere, including the kind of national-infrastructure positioning detailed in coverage of the sovereign cloud race among AWS, Microsoft, and Google. HIVE is essentially betting that South American and academic customers will want AI compute located closer to home, and that renewable power gives it a durable cost advantage over hyperscale data centers in higher-cost grids.
The Competitive Landscape: Miners Racing Into AI Compute
HIVE isn’t alone. The broader bitcoin mining sector has spent the past two years pivoting toward AI and HPC hosting as a way to diversify away from mining economics that swing hard with bitcoin price and network difficulty. Publicly traded peers including Core Scientific, Cipher Mining, TeraWulf, Iris Energy, and Bitfarms have each pursued some version of the same playbook: repurpose power contracts and data-center shells for GPU hosting and colocation deals with AI companies. The exact revenue mix and contract structures vary by company, and HIVE’s own disclosures don’t offer a direct comparison against these peers, so treating HIVE’s numbers as representative of the whole sector would be a mistake.
What sets HIVE’s September announcement apart is the specificity of the daily revenue disclosure. Plenty of miners talk about their AI pivot in strategic terms, but fewer publish a real-time daily revenue figure that lets outside observers track the transition on something close to a running basis. That level of disclosure is itself a signal that management wants the market watching the GPU cloud number closely, not just the quarterly filing.
Historical Context: How Bitcoin Miners Became AI Landlords
The mining-to-AI pivot traces back to the 2022-2023 bitcoin bear market, when falling coin prices and rising energy costs squeezed miner margins hard enough that several public mining companies started shopping their power contracts to other buyers. GPU shortages tied to the generative AI boom that followed gave miners an obvious pitch: they already had megawatts under contract and warehouse-scale buildings with industrial cooling, exactly what AI compute providers needed and couldn’t get built fast enough through traditional data-center developers.
HIVE’s own history reflects that arc. The company built its reputation as a bitcoin and Ethereum miner running on hydroelectric and renewable power in Canada, Sweden, and Iceland, before Ethereum’s move to proof-of-stake in 2022 forced a rethink of its GPU fleet’s purpose. That leftover GPU infrastructure, and the engineering familiarity with running large GPU farms, gave HIVE an earlier starting point for its AI pivot than mining-only competitors that had to build GPU operations from scratch. The Mark Volk hire is best read as HIVE moving from the “build it” phase of that pivot into the “sell it aggressively” phase.
Market Reaction and What to Watch Next
HIVE’s public filings and press materials don’t include specific analyst price targets tied to the September 10 announcement, and the company’s own disclosures focus on operational milestones rather than stock-price commentary. That makes near-term investor reaction hard to quantify precisely from company sources alone, though the timing, an executive hire paired with a headline revenue milestone, is a familiar investor-relations pattern for a company trying to keep momentum in its stock story ahead of further AI infrastructure announcements.
The more useful things to watch going forward are the hard deployment dates: whether the Yguazú substation energizes on schedule in September 2026, whether the 504-GPU Bell Canada deployment ramps to its projected $15 million in annual recurring revenue without delay, and whether BUZZ HPC’s quarterly revenue keeps compounding at anything close to its FY2026 growth rate. Those are the numbers that will confirm or undercut the pivot’s momentum, not the executive announcement itself.
What This Means for Enterprise and Sovereign AI Buyers
For enterprise buyers and government agencies evaluating GPU capacity providers, HIVE’s pitch is essentially a smaller, power-advantaged alternative to hyperscale cloud GPU rental. The company’s colocation and sovereign AI cloud offerings target customers who want dedicated infrastructure, potentially in a specific jurisdiction like Paraguay or Canada, rather than shared capacity inside a US hyperscaler’s global fleet. That’s a similar value proposition to what’s driving demand in adjacent hardware categories, including the enterprise server market covered in reporting on HPE’s Nvidia Vera CPU server launch, where buyers are weighing dedicated infrastructure against public cloud rental.
Whether that pitch wins meaningful enterprise and government contracts now depends heavily on the sales execution Volk was hired to deliver. Infrastructure alone doesn’t close long-duration contracts with hyperscale or government buyers, who typically run lengthy procurement and security review processes before committing. HIVE’s bet is that having dedicated sales leadership focused specifically on those buyer categories will shorten that cycle.
Predictions: Where HIVE and BUZZ HPC Go From Here
- GPU cloud revenue will keep growing faster in percentage terms than bitcoin mining revenue through the rest of fiscal 2026, but mining will likely remain the larger absolute dollar contributor to the $1 million daily figure through at least the next two quarters.
- The Bell Canada deployment and New Brunswick colocation conversion are the two milestones most likely to move HIVE’s HPC ARR toward its stated $225 million target. Delays in either would be the first sign the timeline is slipping.
- Expect HIVE to announce additional GPU supply contracts, likely involving next-generation Nvidia accelerators, as Volk’s sales organization builds out its enterprise and government pipeline over the coming quarters.
- Other publicly traded bitcoin miners will likely make similar dedicated-revenue-executive hires over the next year as the sector standardizes on a “build infrastructure, hire specialized sales” model for monetizing AI compute.
- Paraguay’s Yguazú buildout will be the slower-moving story of the two regional bets, given its multi-year construction timeline. Canada’s Bell Canada and New Brunswick sites are the more likely near-term revenue drivers through 2027.
The Bottom Line
HIVE’s September 10 announcement bundles two separate stories into one press release: a routine executive hire and a genuinely notable revenue disclosure. The $1 million average daily revenue figure is real and sourced directly to the company, but it’s still mostly a bitcoin mining number with a fast-growing GPU cloud business layered underneath it. Mark Volk’s job is to flip that ratio over time by turning BUZZ HPC’s data centers in Paraguay and Canada into long-term enterprise and government contracts rather than spot-market GPU rentals. Whether that happens will show up in HIVE’s next few quarterly filings, not in the announcement itself.
Frequently Asked Questions
What did HIVE Digital Technologies announce on September 10, 2026?
HIVE announced that its subsidiary BUZZ HPC hired Mark Volk as Senior Vice President, Revenue, and disclosed that its combined bitcoin mining and GPU cloud operations have surpassed $1 million in average daily revenue.
Who is Mark Volk?
Mark Volk is described in HIVE’s official announcement as an AI and HPC industry veteran. He will lead BUZZ HPC’s global go-to-market strategy, targeting enterprise, government, sovereign AI, hyperscale, research, and academic customers, and reports to Craig Tavares, BUZZ HPC’s President and Chief Operating Officer.
What is BUZZ HPC?
BUZZ High Performance Computing (BUZZ HPC) is a wholly owned subsidiary of HIVE Digital Technologies that operates GPU-powered AI infrastructure, including sovereign AI cloud, GPU-as-a-Service, colocation, and hybrid AI solutions.
How much of HIVE’s $1 million daily revenue comes from bitcoin mining versus AI compute?
Per HIVE’s disclosure, the company has produced an average of approximately 12 bitcoin per day since August 21, 2026, representing about 2% of global bitcoin network production, while GPU cloud operations generated approximately $100,000 in average daily revenue. Mining remains the larger single contributor to the combined total.
What was HIVE’s total revenue in FY2026?
HIVE reported FY2026 total revenue of $297.8 million, up 158% year-over-year, with BUZZ HPC contributing $19.5 million of that total, up 94% from $10.0 million in FY2025.
What is the Bell Canada GPU deal HIVE signed?
In February 2026, HIVE signed a two-year, $30 million contract for 504 Nvidia B200 GPUs, expected to deploy at a Bell Canada Tier-III facility. HIVE says the deal adds roughly $15 million in annual recurring revenue.
Where are HIVE’s AI data centers located?
HIVE operates or is building AI infrastructure in Asunción and Yguazú, Paraguay, and in Canada through a Bell Canada Tier-III facility and a planned conversion of a New Brunswick data center from Tier-I mining use to Tier-III colocation.
Are other bitcoin miners making the same pivot to AI compute?
Yes. Several publicly traded bitcoin miners, including Core Scientific, Cipher Mining, TeraWulf, Iris Energy, and Bitfarms, have pursued similar strategies of repurposing power contracts and data-center capacity for AI and HPC hosting, though the specific revenue mix and contract terms vary by company.
