BD
Bit Digital, Inc (BTBT)·Q4 2024 Earnings Summary
Executive Summary
- Bit Digital’s Q4 2024 marked a visible pivot to HPC: cloud services revenue reached $13.0M and accounted for ~50% of total Q4 revenue; digital asset mining contributed ~40% of Q4 revenue, and colocation added $1.4M from the October Enovum acquisition .
- Management highlighted a growing contracted run-rate in cloud services ($62M currently; ~$72M with DNA Fund starting late March), plus additional B200/H200 capacity and on‑demand initiatives that can further lift ARR in coming quarters .
- Guidance/timing shifted: the $100M HPC run-rate target moved from “by year‑end 2024” to “early 2025,” reflecting deliberate customer-led chip upgrades and capital discipline; Montreal II Tier‑3 data center targeted for mid‑2025, with Cerebras 5MW colocation also commencing mid‑2025 .
- Financing narrative is a catalyst: management intends to “turn off” the ATM at current valuations and pursue non‑dilutive mortgage/vendor financing, which could re-rate the stock away from “pure Bitcoin miner” perceptions as WhiteFiber (HPC) scales .
What Went Well and What Went Wrong
What Went Well
- Rapid HPC scale-up: “HPC revenue made up over 40% of full‑year revenue and more than half of Q4 revenue,” with Q4 cloud services revenue of $13.0M and nine active customers; run-rate ~$62M rising to ~$72M after DNA Fund begins .
- Strategic vertical integration: Enovum acquisition added Tier‑3 capacity, a 280+ MW expansion pipeline, and colocation revenue of $1.4M recognized from the acquisition date through year‑end; Montreal II retrofit is underway with a mid‑2025 go‑live .
- Blue‑chip validation: a 5MW, five-year colocation agreement with Cerebras confirms Bit Digital’s ability to deliver high‑density, customized HPC infrastructure; operations expected to commence mid‑2025 .
What Went Wrong
- Mining pressure: halving and rising network difficulty reduced bitcoin production YoY; mining margins compressed post‑halving (Q3 total gross margin was 32% vs. Q2 48%), with legacy fleet inefficiency a noted drag .
- Digital asset volatility hit quarterly results: Q3 adjusted EBITDA was −$21.8M, driven largely by a $21.9M unrealized loss on digital assets; this overshadowed underlying HPC gross profit momentum .
- Timing slip on HPC run-rate: the $100M HPC run-rate target—originally guided for YE24—shifted to early 2025 amid customer‑driven hardware upgrades to newer NVIDIA chips (B200/GB200) and prioritization of quality revenue .
Financial Results
Quarterly Headline Metrics
Segment Revenues by Quarter
Note: Q4 segment mix signposts (cloud ~50%, mining ~40%) were disclosed without a total revenue figure .
KPIs and Operating Metrics
Guidance Changes
Earnings Call Themes & Trends
Management Commentary
- “HPC revenue made up over 40% of full‑year revenue and more than half of Q4 revenue… Cloud services produced $13,000,000 of revenue in the fourth quarter” .
- “The Enovum acquisition… vertically integrated our data center operations… Montreal II… a five megawatt Tier three data center expected to go live in mid‑twenty twenty five… powered by 100% renewable hydroelectricity” .
- “We are firmly in the mix for blue chip deals… secured a five‑year colocation agreement to provide 5MW… with Cerebras… operations… mid‑2025” .
- “At these current levels, there is no desire to tap into the ATM… focus is on securing alternative financing options… mortgage financing… vendor financing and leasing structures” .
- “We believe that we are deeply misunderstood… the bigger our HPC business gets, the more our stock seems to trade like a pure play Bitcoin miner… eventually… valued properly” .
Q&A Highlights
- Cloud run-rate build: ~$62M current; ~$72M when DNA Fund begins; contracted basis “well past $100M” including data center business; B200 on‑demand could add ~$25M ARR; 908 H200 GPUs under evaluation for reserved contracts .
- Colocation revenue: ~$1.4M recognized since Enovum closing in Q4; Montreal II mid‑2025; Cerebras 5MW timing mid‑2025; additional U.S. capacity under LOI (24MW available, path to 48MW in 60–90 days, and +100MW by end‑2025) .
- Mining trajectory: redeploying/refreshing to S21/S21+; efficiency target ~22 J/TH; operational hash rate to ~2.5 EH/s by May; 3 EH/s requires ~6.6MW additional hosting and ~1,800 S21+ .
- Financing pivot: ATM off at current valuations; progressing on commercial mortgage term sheet; considering vendor/lease financing for GPUs to preserve capital efficiency .
- On‑demand platform: near‑term via third‑party partner; internal on‑demand platform targeted end‑2025/early‑2026 .
Estimates Context
- We attempted to retrieve S&P Global consensus EPS and revenue estimates for Q4 2024 and prior quarters, but the data was unavailable due to an API rate‑limit error; therefore, comparisons versus Wall Street consensus are not provided at this time. We will update if S&P Global data becomes accessible. [SPGI request error]
Key Takeaways for Investors
- HPC is now the core earnings driver: Q4 cloud services revenue of $13.0M, ~50% of Q4 revenue, with contracted and pipeline growth across reserved and on‑demand offerings; this should reduce earnings cyclicality versus mining .
- Vertical integration and metro‑area strategy (inference latency) creates defensible positioning; Enovum’s pipeline and the Cerebras win de‑risk multi‑MW expansions in 2025 .
- Financing shift is constructive: turning off ATM at current valuations and moving to mortgage/vendor financing could unlock non‑dilutive growth and improve investor perception .
- Mining becomes a “call option”: fleet optimization and efficiency gains matter more than EH/s scale; expect incremental improvements without capital crowd‑out from HPC .
- Near‑term catalysts: DNA Fund ramps late March, B200 on‑demand cluster launches in April, Montreal II mid‑2025, Cerebras deployment mid‑2025; execution could reframe the equity narrative away from crypto beta .
- Watch tariff/policy risk on data center equipment costs; management is pursuing supply chain diversification to mitigate potential impacts .
- If/when the market decouples BTBT from “pure miner” comps, the HPC business may warrant higher multiples; non‑dilutive financing and consistent run‑rate disclosures can accelerate that re‑rating .