AI Summary
5 min readApple is planning a return to the server market in 2029, SK Hynix is in talks with Intel to manufacture memory chips in the U.S., and a new alliance of tech giants including NVIDIA, Google, and Anthropic is launching to tackle AI's massive energy needs by making the existing power grid more flexible. These three developments, covered in a recent podcast, highlight the accelerating race to build the physical infrastructure—chips, memory, and power—required to sustain the AI boom, even as debates about the pace of AI development continue.
Apple's Server Return: A 2029 Play with NVIDIA
Fifteen years after killing its Xserve line, Apple is getting back into the server business. The company is planning a new server offering built around four M8 Ultra chips, stitched together using NVIDIA's NVLink Fusion technology. This is notable because Apple and NVIDIA have had a rocky relationship for years, but Apple is now turning back to NVIDIA to power its revamped server hardware. The timeline is distant—the M8 generation is at least two silicon cycles away, likely arriving in 2029—but the strategic shift is clear.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Hard Fork AI
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (01:06) **Apple Re-enters Server Market with M8 Ultra & Nvidia** - Apple plans to launch a new server offering in 2029 using four M8 Ultra chips and Nvidia's NVLink fusion, marking a return 15 years after killing X Server.
- 2 (06:37) **SK Hynix in Talks with Intel to Make Memory Chips in US** - The South Korean memory giant would use Intel's US fabs for the first time, shifting HBM production for AI accelerators closer to US demand.
- 3 (08:42) **Nvidia, Google & Emerald AI Launch AI Energy Management Alliance** - The alliance, with 18 members including Anthropic and National Grid, aims to unlock 100 gigawatts from the existing US grid by making data centers grid-flexible.
- 4 (10:51) **The Data Center Power Bottleneck and Elon Musk's Workaround** - The host details how Musk built the Memphis data center using diesel generators instead of waiting for a grid connection, despite pollution lawsuits.
- 5 Standout Quotes
- 6 "The US power grid is only running at about fifty percent utilization on average... each 10% gain in grid utilization is going to lower electrical rates by about 3.4%."
- 7 "US data centers currently wait up to 10 years or more for a grid connection under some of the existing interconnection processes."
+ Full timestamped outline available in the app
Show Notes
In this episode, we explore significant advancements in AI hardware, including Apple's return to the server market and collaboration with Nvidia. Additionally, we discuss SK hynix's plan to produce memory chips in the U.S., as well as an initiative to optimize power generation for AI data centers.
Chapters
00:00 Introduction
00:51 Apple's Server Market Comeback
03:14 SK Hynix and Intel Partnership
03:57 AI Energy Management Alliance
06:02 Power Utilization Challenges
10:55 Conclusion and Recap
Show Links
More from this podcast
Hard Fork AI →