Summary Bitdeer demonstrates robust operational growth, with vertical integration driving hashrate, Bitcoin production, and AI cloud ARR expansion despite headline GAAP losses. Q1 revenue surged 170% YoY to $188.9 million. While self-mining revenue and fleet efficiency improved, non-cash charges and aggressive depreciation weighed on margins. AI Cloud ARR accelerated from ~$10 million in January to over $69 million post-Q1, with GPU utilization hitting 94%, signaling strong future revenue potential. BTDR’s current valuation lags AI-pivot peers, but sustained high GPU utilization and ARR conversion could drive a rerating as the market digests its vertical integration and AI pivot. Bitdeer ( BTDR ) is a Bitcoin miner (turned AI cloud provider) that has grown at an unprecedented pace since its founding in 2021. Results for the first fiscal quarter of FY26 were released last week Thursday. On going through the Q1 results , I find that Bitdeer’s vertical integration thesis remains valid, and with each quarter that passes, operational metrics show real growth that have been directly unlocked by the company’s vertical integration strategy. Last week's results showed some decline in financial numbers (which were mostly non-cash, and accounting-driven misses on the headline), but most of the non-cash charges that influenced key lines in the earnings were predictable for investors who follow Bitdeer closely, mainly because of the aggressive ramp of hashrate since the start of FY25 and the adoption of U.S. GAAP reporting. In the Q4 and full year FY25 earnings call released back in February, management had earlier mentioned this transition to GAAP accounting to officially take effect beginning in Q1 2026. Bitdeer - An Analysis of Q1 2026 Beyond the Headline Bitdeer's Q1 included both the good and the ugly (on the surface). Starting with the ugly, which made most of the Q1 headline, is the $159.5 million GAAP net loss reported against a $105.3 million net profit a year ago. Then the negative gross margin of -20.7%, which translates to $39 million gross loss, and adds more fuel to the bearish interpretation of the results. Depreciation and SBC led to gross margin compression (Bitdeer Q1 earnings report) The decline in gross income was largely driven by electricity costs in self-mining operations and the depreciation of previously deployed mining hardware, alongside stock-based compensation tied to self-mining activities. Management seems to have adopted a more aggressive depreciation assumptions especially with the launch of the SEALMINER A4 machines last month. Top line and adjusted EBITDA (Bitdeer Q1 earnings report ) On the more positive side Bitdeer continued its YoY top line expansion momentum, though on a sequential basis the top line declined. In Q4 there was a one-off batch delivery of SEALMINER sales which produced about $23.4 million in that quarter. Adjusted EBITDA came in at $14.4 million, looking materially healthier than both gross margin and net income because adjusted EBITDA added back a significant portion of the non-cash depreciation and stock-based compensation expenses that weighed on the cost of revenue line. Q1 revenue reached $188.9 million compared to $70.1 million in Q1 FY25, representing ~170% YoY growth. Self-mining revenue alone increased from $37.2 million a year ago to $146.9 million, thanks to the aggressive expansion of hashrate that started last year. Revenue from external hardware sales (Bitdeer Q1 earnings report ) In Q1, external miner hardware sales dropped to ~$3.7 million (one of the factors behind the sequential revenue decline) as Bitdeer shifted its internal inventory focus toward proprietary chips for self-mining expansion and ramping the cloud business (as the balance sheet analysis will show later in this piece). Bitcoin mining operating metrics (Bitdeer Q1 earnings report ) Average self-mining hashrate increased from 9.7 EH/s to 63.2 EH/s YoY, while total hashrate under management climbed to 78.1 EH/s from 24.2 EH/s. Bitcoin production rose from 350 BTC to 2,033 BTC YoY. And fleet efficiency also improved a lot, from 29.0 J/TH to 16.4 J/TH compared to a year ago. What these operational metrics point out is that while the headline GAAP loss expanded and gross margin took a non-cash hit, the metrics show that actual operations became stronger in Q1. Balance sheet (Bitdeer Q1 earnings report ) With the capitalization move back in February when Bitdeer issued $375 million in 5% convertible senior notes due 2032, cash was boosted in Q1 to $297.7 million (inclusive of restricted cash), plus $245 million in digital assets and digital assets receivables. That brings total near-term liquidity exposure to ~$542.7 million. While liquidity exists, Bitdeer's $1.9 billion in total borrowings shows that long-dated leverage still dominates the capital structure. The balance sheet also shows Prepayments and other assets dropping to $386.7 million in Q1 (down from $723.0 million from Q4 FY25), which management explicitly states, in the Q1 press release , is primarily driven by “delivery of raw materials procurement for SEALMINERs mass volume production.” Management’s explanation for the drop in prepayments is constructive because it means capital already spent in prior periods (mainly towards the last two quarters of FY25) is now converting into physical inputs. While prepayments dropped, the inventory line surged to $613 million from $252 million in Q4 FY25. This line is important in particular as this massive build-up shows the production pipeline moving deeper into the manufacturing and deployment stage. Management disclosed in the Q1 press release that inventories now include wafers, chips, work-in-progress, and finished SEALMINER units. That detail matters also, because it confirms the inventory increase is not being driven by stagnant finished goods sitting idle, but by active progression through Bitdeer’s vertically integrated supply chain. To that effect, the balance sheet shows capital moving through the stages of conversion. First through procurement and prepayments, then into chips and unfinished hardware under inventory, before ultimately becoming deployed mining and AI infrastructure under property, plant and equipment [PP&E]. PP&E also increased by $100 million, from $1.1 billion in Q4 FY25 to $1.2 billion in Q1 which management attributes mainly to deployment of SEALMINER units into Bitdeer’s datacenters alongside broader datacenter expansion activities. The inventory build therefore appears less like dead capital and more like future hashrate and future compute capacity waiting to be deployed into revenue-generating infrastructure. R&D expense (Bitdeer Q1 earnings report ) R&D expense also declined by ~66% to $20.2 million from $59.0 million a year ago. Bitdeer spent much of FY24 and early FY25 aggressively funding the design, and development stages of its proprietary SEALMINER hardware roadmap. The R&D decline now suggests that a large portion of the most capital-intensive chip development work may already be behind the company for now, with focus now shifted toward mass production, deployment, and monetization rather than early-stage engineering iteration. The timing of this decline also aligns closely with the broader balance sheet movements seen in Q1. As prepayments are converted into raw materials, inventories expanded to include wafers, chips, and finished SEALMINER units. PP&E also increased through active deployment into datacenters; the reduction in R&D reinforces the idea that Bitdeer is progressing into the infrastructure scaling phase of its vertical integration strategy. When all these balance sheet and operating movements are viewed together, you find that the slowed R&D spend, decline in prepayments, the surge in inventory, and the expansion in PP&E show a synchronized capital deployment cycle. While the above movements from the balance sheet and expense lines explain the financial flows of the vertical integration strategy for Bitdeer, the operational metrics I already mentioned early on is the clearest proof that ties all of Bitdeer's capital deployment activities together; the hashrate more than tripling from a year ago, Bitcoin production increasing over 6x from a year ago, fleet efficiency improving materially, and the AI cloud achieving scaled run rate growth within past few months. Management disclosed that AI Cloud ARR scaled from ~$10 million in January this year to about $43 million in March (end of Q1) before surpassing a $69 million run-rate post-Q1; while GPU utilization has climbed from around 41% to ~94% in the same period. Moving forward into Q2 and the rest of the fiscal year, I believe much of that ARR will translate into better recognized revenue acceleration across AI Cloud services, away from the $3.7 million in recognized AI Cloud revenue in Q1 if the current GPU utilization rate is maintained. Q1 ending with a $43 million ARR and that run-rate jumping to $69 million implies that Q2 would likely report a meaningful surge in AI cloud revenue and that could dominate the headline for the Q2 results. And for the longer term, there is support for a prolonged high utilization rate based on Bitdeer CEO’s own comment in the Q1 earnings call Q&A, that the AI cloud customer mix has shifted toward 3- to 5-year enterprise contracts. Phase 1 delivery of 180 MW at the Tydal, Norway, facility in Norway is another noteworthy catalyst for the AI segment in the near term, as completion is anticipated as early as the end of this year. The facility is designed to become Nvidia's Vera Rubin hosting platform (or colocation hub). I believe the current dip (following BTDR’s ~9% drop on Friday) with BTDR now trading around $12.85, presents one of the cleanest entries before the Q2 results are released sometime in mid-August. Bitcoin price swings would also determine if we would see more entry points, but on a fundamental setup basis for BTDR alone, I believe the current dip is an attractive accumulation zone. A forward catalyst for the mining segment is the fact that the more efficient SEALMINER A4 was officially launched last month. The miners operate at ~9.45 J/TH and these efficiency gains directly translate to better unit economics for the rest of the year compared to Q1. For Q2 and the rest of the fiscal year as the more efficient miners are deployed in Bitdeer's self mining, I also expect much less non-cash depreciation pressure tied to self mining compared to the $76.3 million depreciation charge in Q1 that mostly came from prior-generation mining machines that built Bitdeer's current hashrate base. Risk While the ARR numbers and GPU utilization rate have gone up impressively, the AI segment carries the risk of overpromising, as ARR does not always convert to revenue at the pace run-rate implies. The fact that we now know the ARR figures (~$43 million as of Q1 end and over $69 post Q1) going into Q2 gives an early insight to grade the AI Cloud revenue when Q2 results are released. Any shortfall between the $69 million April run-rate and actual Q2 AI Cloud revenue recognition will be immediately visible, and the market could react negatively if that happens. Besides ARR conversion, GPU utilization is another factor. Any drop in GPU utilization from the current 94% range would translate to weaker near-term revenue conversion, and could immediately weaken the AI thesis. The self-mining business remains heavily tied to Bitcoin's volatility. Besides volatility, Bitdeer has scaled the hashrate aggressively over the last year. While I've mentioned that depreciation could likely moderate in the coming quarters as the SEALMINER A4 machines are deployed, that is not a definitive conclusion, and only gets speculative. That argument rests on the hope that Bitdeer has undergone a one-time heavy depreciation cut in Q1 that resulted in that heavy $76 million non-cash depreciation charge. We do not know the exact mix of what is being depreciated or the precise deployment cadence of the fleet; so we can only assume so much from the outside. Takeaway I think Bitdeer's vertical integration thesis is being underappreciated and the AI cloud pivot is still being misunderstood by the market because it has not come with a multi-year deal announcement with a bluechip partner like TeraWulf ( WULF ) landed the $3.7 billion 10-year colocation agreement with Fluidstack, or IREN’s ( IREN ) $9.7 billion five-year AI cloud contract with Microsoft ( MSFT ), or Hut 8’s ( HUT ) $7 billion 15-year lease Google-back deal with Fluidstack. I believe that though Bitdeer's AI cloud pivot may not come with a one-off multi-billion anchor tenant deal yet, the GPU utilization reported in Q1 as well as the surge in ARR preludes the slow compounding revenue that builds up over time. Data by YCharts I also think the absence of that one-time multibillion anchor client currently makes Bitdeer lag its AI-pivot peers in price action over the past year. The preceding chart shows how peers have re-rated but BTDR seems to remain suppressed. Data by YCharts The price to sales multiple also shows BTDR has much room for a rerating compared to how much peers are being valued based on sales. The main metrics BTDR has to maintain for an upside rerating are the GPU utilization and ARR conversion. Both holding true would mean higher revenue, which means lower P/S, which would in turn support the rerating.
