요약하면
매출 가이던스 상향.
지금은 반도체보다 전력이 더 문제.
유럽과 캐나다로 확장 중.
추론과 트레이닝 사이에서 아직 큰 수요의 차이는 없음.
수요가 너무 많아, 데이터 센터 캐파를 늘리는 족족 기존 고객이 모두 흡수해버리는 구조.
계약이 끝난 사용된 낡은 GPU를 어떻게 처리하는가에 대해서 구체적 답변은 피함.(아마도 대규모 재투자가 필요한 부분이라 언급을 피한 것 같음.)
하이라이트
CoreWeave raised its full-year revenue guidance to $5.15-$5.35 billion, up from the previous range of $4.9-$5.1 billion.
As a result, we anticipate $500,000,000 in fully ramped annual run rate cost savings by the 2027, benefiting both the Core Scientific and CoreWeave shareholders directly.
The increase in G and A was largely driven by professional services. Adjusted operating income for Q2 was $200,000,000 compared to $85,000,000 in Q2 twenty twenty four.
There’s a lot of different pieces that are constrained. But, ultimately, the piece that is the most significant challenge right now is accessing powered shells that are capable of delivering the scale of infrastructure that our clients are requiring.
There are other jurisdictions that are going to be less welcoming to to tech coming out of The US, and that’s just the nature of the way the world is gonna unfold for this. We’ve had some success in Canada. We’re really excited about our partnership with Cohere up there. We think they they they’re doing a wonderful job and that that infrastructure is is really well positioned to be successful. We’ve done a a really good job expanding infrastructure across Europe.
And so we don’t see a real fluctuation in the economics associated with with inference or training.
The biggest challenge that we have is that every time we’re able to build capacity, it is immediately consumed by one of our existing or a new client that wants to expand their exposure to additional compute to be able to serve their models.
이하 가장 중요
Gregg Moskowitz : Great. And I’ll add my congratulations. In the Q2, can you give us a sense of how successfully you were able to repurpose older GPU clusters that had come off contract? Any changes today vis a vis how this was trending around the start of the year?
Mike Entrater, CEO, CoreWeave: Yeah. So thank you for the question. It’s a great question. So what we are seeing is we are seeing the infrastructure that is being delivered off of these contracts being recontracted out for additional term in order to be able to continue to deliver that compute largely for inference. And so we’re talking about the the h one hundreds.
