NFDG

About

NFDG is a venture capital fund founded in 2023 by Nat Friedman and Daniel Gross, with $1.1 billion in committed capital for their debut fund ¹². The fund’s name derives from the founders’ initials ¹.

Nat Friedman (born 1977) is an American technology executive who studied Computer Science and Mathematics at MIT, graduating in 1999 ³. He co-founded Xamarin and later served as CEO of GitHub, overseeing its growth following Microsoft’s acquisition ³⁴. Daniel Gross (born 1991 in Jerusalem) was accepted into Y Combinator’s Winter 2010 batch at age 17 — the youngest founder ever admitted at that time ⁵. He founded Greplin (later Cue), which was acquired by Apple for an estimated $40-60 million in 2013 ⁵. Gross then served as Director of AI and Search at Apple from 2013-2017, followed by a stint as a Y Combinator partner starting in 2017 ⁵.

NFDG invested between $1 million and $100 million per deal, exclusively focused on AI companies — AI-enabled products, AI infrastructure, and AI applications ¹². The fund launched at the onset of the generative AI boom, positioning it to capture early positions in what became the most heavily funded technology wave in venture history.

With approximately 50% of the fund deployed, NFDG achieved approximately 4x returns (~$550M invested reaching ~$2.2B in portfolio value) within two years ¹². The fund’s crown jewel was Safe Superintelligence Inc. (SSI), co-founded by Gross alongside Ilya Sutskever, which grew from a $5 billion to a $32 billion valuation ¹⁵.

In June-July 2025, Meta Platforms agreed to partially acquire NFDG’s holdings — potentially more than $1 billion worth — providing liquidity to the fund’s limited partners at approximately 4x returns within two years ¹²⁴. Both Friedman and Gross joined Meta: Friedman became Vice President of Product and Applied Research, and Gross joined Meta Superintelligence Labs ²⁴⁵.

Stated Thesis

NFDG positioned itself exclusively as an AI-focused venture fund from inception. Gross has stated: “We want to democratize AI. We want to level the playing field for startups to ensure that innovation doesn’t get locked up in large companies” ⁵.

The fund was structured to provide portfolio companies with more than capital. NFDG built the Andromeda Cluster, its own supercomputer, initially with 2,512 NVIDIA H100 GPUs at an approximate cost of $100 million, expanding to over 4,000 GPUs by 2024 ⁵. This cluster could train a 65-billion-parameter model in approximately 10 days and was offered to portfolio companies at below-market rates ⁵.

NFDG also operated an AI Grant program (through aigrant.org) providing startups with $250,000 in funding plus $250,000 in Microsoft Azure cloud credits, using no-cap, no-discount MFN SAFEs ⁵⁶. By 2024, the AI Grant program had backed approximately 60 companies ⁵.

Inferred Thesis

Based on NFDG’s verified portfolio data:

Stage distribution: NFDG invested across stages from seed to growth, with check sizes ranging from $1 million to $100 million per deal ¹. The AI Grant program operated at a separate, smaller scale ($250K checks) ⁵⁶.

Sector focus: Exclusively AI. Every known NFDG investment is an AI company or AI-adjacent infrastructure company. This makes NFDG one of the most sector-concentrated funds of its size ever raised.

Portfolio concentration: With ~$550M deployed across the full portfolio, and SSI alone reaching $32B valuation, the fund’s returns were heavily driven by a single position. This is consistent with a concentrated, conviction-driven investment approach rather than broad diversification.

Infrastructure as investment thesis: Building a $100M+ GPU cluster for portfolio companies was unprecedented for a venture fund. This signaled a thesis that compute access would be a critical competitive advantage for AI startups, particularly during the 2023-2024 GPU shortage ⁵.

Speed of deployment and exit: The fund achieved 4x returns in approximately two years, then both GPs departed to join Meta. This is an unusually short lifecycle for a venture fund and raises questions about long-term fund management ¹.

Notable pattern — founder-operator to investor and back: Both Friedman and Gross were operators before becoming investors, and then returned to operating roles at Meta. This suggests NFDG may have functioned partly as a vehicle for the founders to build conviction about AI’s trajectory before committing to an operating role at a major company.

Portfolio

This table represents the publicly known NFDG portfolio. The fund deployed approximately $550M of its $1.1B commitment ¹.

In Their Own Words

“We want to democratize AI. We want to level the playing field for startups to ensure that innovation doesn’t get locked up in large companies.” — Daniel Gross, Co-Founder ⁵

“I expect miracles to follow.” — Daniel Gross, upon leaving SSI for Meta, June 2025 ⁵

What Founders Say

No independently sourced founder testimonials found.

Sources