Super Micro soars as blockbuster Q2 results restore faith in the AI server company

Sports-betting company DraftKings rose in after-market trading Thursday after it reported better-than-expected Q1 sales and earnings. Here’s the rough outline of the results:

• Q1 revenue of $1.65 billion vs. Wall Street’s $1.63 billion expectation, according to FactSet.

• Q1 earnings per share of $0.03 compared with a consensus estimate of $0.01.

• Q1 adjusted EBITDA of $167.9 million vs. $152.6 million expectation.

• Maintained previous full-year adjusted EBITDA guidance of $700 million to $900 million, compared with estimates of $791.4 million.

• Maintained previous full-year sales guidance of between $6.5 billion and $6.9 billion (midpoint $6.70 billion) and analysts’ estimates of $6.82 billion, according to FactSet.

Shares of traditional online sports gambling like DraftKings have struggled as prediction markets have emerged as a center of industry excitement.

The shift to such markets has been tricky for both DraftKings and rival FanDuel, the US leader in online sports betting — which have to manage pre-existing relationships with state gaming commissions that stand to be disrupted by prediction markets, which are regulated on the federal level by the CFTC.

DraftKings is down roughly 25% in 2026, while FanDuel parent Flutter Entertainment, which reported earnings yesterday, is down more than 50%.

CoreWeave is whipsawing in after-hours trading as investors digest whether its Q1 results can justify the 86% rally since late March.

In Q1, the neocloud firm reported:

• Revenue: $2.1 billion (estimate: $2 billion)

• Adjusted EBITDA: $1.2 billion (estimate: $1.1 billion)

While its revenue beat was only a little north of 5%, the figure surpassed all of the 32 analyst estimates compiled by Bloomberg.

As of March 31, CoreWeave’s revenue backlog was a whopping $99.4 billion, up from $66.8 billion in the prior quarter.

“We surpassed 1 GW of active power and believe we are well on our way to more than 8 GW by 2030, having positioned our capital structure to scale with the opportunity ahead," said CEO, co-founder, and Chairman Michael Intrator in a press release. “AI natives and enterprise customers are choosing CoreWeave because we sit between the models and the silicon, delivering the infrastructure, software, and expertise required to build and run AI at scale.”

At the end of the quarter, the company managed to close a unique debt deal backed by GPUs and what Meta is slated to pay for AI compute.

Since then, CoreWeave and its peers have been buoyed by a scramble for compute catalyzed by a seeming shortage for Anthropic, as the Claude developer aimed to beef up its footprint amid complaints around usage limits.

CoreWeave reached a multiyear deal with Anthropic to help power Claude, and also expanded its AI compute sales pact with Meta by $21 billion.

Retail favorite Rocket Lab rose late Thursday after reporting better-than-expected Q1 sales and offering up beat sales guidance for Q2.

Here’s how the company did:

• Q1 revenue of $200.3 million vs. Wall Street’s expectation for $189.7 million, according to FactSet.

• An adjusted loss per share of -$0.07 vs. the consensus estimate of a -$0.07 loss.

• Adjusted EBITDA of -$11.8 million vs. analyst expectations of -$26.3 million.

• Q2 sales guidance of between $225 million and $240 million ($232.5 million midpoint) vs. expectations for $205.3 million.

• Q2 guidance for an EBITDA loss of between -$20 million and -$26 million (-$23 million midpoint) vs. the -$14.5 million analysts were penciling in.

Rocket Lab shares have surged roughly 2,000% over the last two years, as the company capitalized on investor enthusiasm for space.

Over the last year, Rocket Lab also rode growing excitement about companies that plan to use their ability to place clusters of satellites into low-earth orbit, and then sell data services to earthlings below — essentially the business model of Elon Musk’s Starlink.

Though it’s privately held for now, Musk’s space behemoth — SpaceX — remains the key source of excitement around the sector, enthusiasm which will likely grow as SpaceX moves forward with plans for what’s likely to be the largest public offering ever.

Rabid space enthusiasm aside, Rocket Lab remains a money-losing company that’s burning a lot of cash, though Wall Street analysts think it could break even in 2027.

We’ll see. That projection hangs on the company’s ability to get its larger Neutron rocket into its commercial launch cycle sooner rather than later. And given that Neutron’s maiden launch — originally slated for 2025 — has been delayed to the fourth quarter of 2026, that’s by no means assured.

Opendoor Technologies delivered a set of better-than-expected Q1 results while touting that it’s just achieved a key profitability milestone.

In Q1, the online real estate company reported:

• Revenue: $720 million (estimate: $665.2 million)

• Adjusted EBITDA: -$31 million (estimate: -$33.5 million)

In the press release, the company said it is adjusted EBITDA profitable on a 12-month go-forward basis as of April 1.

For Q2, management offered mixed guidance. The company expects sales of about $900 million (estimate: $1.13 billion) with adjusted EBITDA roughly flat (estimate: -$4.66 million).

Under its new leadership, the online real estate company has redoubled its efforts on aggressive home-flipping and adopted a “default to AI approach,” including using the technology for home assessments and in closings.

“Our 4Q25 and January 2026 cash acquisition cohorts have the best combination of margin, margin stability, and resale velocity of any corresponding cohort in company history (excluding the COVID-era cohorts),” said CEO Kaz Nejatian in a press release.

Opendoor’s share price, one of the most interesting things in the stock market for a couple months in 2025, has been decidedly boring in 2026. Since late January, it’s traded in a range of roughly $4.30 to $5.60.