“This quarter represents continued meaningful progress in CleanSpark's evolution into a digital infrastructure and data center development company.”（CEO）
“Mining funds the platform, AI monetizes it. Together, they create a more balanced, durable business.”（CFO）
“We are building a platform, not a single strategy.”（CEO）
“We have designed our capital strategy to be flexible in order to take advantage of real-time opportunities in the marketplace.”（CFO）
“Our ability to expand within established grid relationships... When a prospective tenant asks what the growth path looks like, we can show them a real track record of unlocking additional scale.”（CEO）
Prepared Metrics
Metric
Value
Speaker/Context
季度营收
1.36亿美元
CFO，环比下降25%
挖矿产出
1,799 BTC
CFO，仅比上季度少22枚
毛利率
>40%
CFO，上季度47%
电力成本
0.052美元/千瓦时
CFO，上季度0.056美元/千瓦时
净亏损
3.78亿美元
CFO，含2.63亿美元非现金市值调整
调整后EBITDA
负2.41亿美元
CFO，上季度负2.95亿美元
流动性（现金+BTC）
约12亿美元
CFO，现金2.6亿美元+BTC 9.25亿美元
比特币持仓
13,561 BTC
CFO，季末后价值约11亿美元
信贷额度未使用
4亿美元
CFO
DAM净现金回报（季度）
400万美元
CFO，FY2026迄今1,720万美元
已签约容量
1.8吉瓦
CEO
桑德斯维尔容量
250兆瓦
CEO，全部通电
Sealy获批容量
285兆瓦
CEO，约200兆瓦2027年上半年通电
Brazoria首期获批容量
300兆瓦
CEO，已获ERCOT批准
Q&A Batch (1-5 of 9)
Q1 — Nick Giles
Topic: Asset conversion potential for AI/HPC and capital deployment at Sandersville
Key points:
Washington site: 86 MW energized, line study underway for significant expansion; Jackson, TN: 60 MW acquired; Cheyenne, WY: 110 MW, fence-line neighbor with hyperscaler.
Sandersville: ~$200M deployed (including miners); additional acreage acquired adjacent; current investment is "millions, not tens of millions" (clearing trees, moving dirt) until lease signed.
Portfolio approach: 60–70% of data center build occurs in factory, reducing time-to-market; introductions via third-party vendors to hyperscale tenants for multi-site deployment.
Mgmt stance: Bullish — multiple sites with expansion potential; capital discipline before lease signing; portfolio approach seen as "frosting on the cake."
Q2 — John Todaro
Topic: Lead tenant progress and Bitcoin hashrate outlook
Key points:
Lead tenant conversations remain consistent with prior framing (IG hyperscale); no new update.
Bitcoin hashrate: current fleet efficiency 16 J/TH; new immersion miners (from BITMAIN) will drop efficiency further, increasing margins; pods are movable for AI strategy.
Hashrate trend: Bitcoin production recently broke 23; expects trend to 55 EH/s through end of year (depends on network difficulty and Bitcoin price).
Mgmt stance: Neutral on tenant update (no change); bullish on mining efficiency and hashrate growth.
Q3 — Brett Knoblauch
Topic: Portfolio tenant approach and co-location of Bitcoin mining with AI
Key points:
Portfolio approach does not change timing for first lease signing; assets have independent value; "first prize" is right credit-quality tenant with portfolio interest.
Co-location strategy: Bighorn mining allows monetizing near 100% of baseload; AI builds take 18+ months; energized sites can host mining pods while waiting for AI build.
Patent: control module allocates power based on criteria; peak PUE at Sandersville ~1.4, average annual ~1.15, leaving significant MW available 93+% of year; Bitcoin mining can utilize spare capacity, subsidizing utility pricing.
Mgmt stance: Bullish — co-location is innovative and viable; portfolio approach is additive, not delaying.
Q4 — Paul Golding
Topic: Brazoria 300 MW review process and incremental land needs
Key points:
Brazoria: first 300 MW fully approved; second 300 MW in "deeply progressed posture"; process leverages prior ERCOT/CenterPoint experience from Sealy (closed 60 days prior).
Land: Texas parcels sufficient for planned AI campuses; Sandersville needed additional 122 acres (city assisted with easement for power); other sites may need more land, evaluated site-by-site.
Community relationships: city of Sandersville proactively helped secure land and right-of-way; consistent positive reception across Georgia and Tennessee.
Mgmt stance: Bullish — strong track record in multiple power markets; land acquisition low-friction due to community relationships.
Q5 — Brian Dobson
Topic: Vision for CleanSpark by 2030 and AI build-out dynamics
Key points:
AI build-out differs from 1990s Internet (speculative fiber); AI leverages existing distribution rails (laptops, phones, connectivity), enabling faster proliferation.
Duke University study: 22 largest U.S. power grids have 76–125 GW of headroom if curtailed 0.5%–1.5% of time (~115 hours/year); Bitcoin mining as rapidly interruptible load unlocks this capacity.
Demand: tenants asking for 60 MW+ between now and 2027; sorting process focuses on financial capability and project financeability.
Mgmt stance: Bullish — every economically viable megawatt will shift to intelligence; small sites (10–50 MW) to gigawatt campuses all valuable; company positioned to bring infrastructure to underserved communities.