Sandisk's Plan to Remove $300K of DRAM from a Single Storage Server
...and save over 10kW for AI factories using that type of server.
Patrick Kennedy's Axautik Group LLC and ServeTheHome Stack
We take the high-value insights that we gain from producing the largest server, AI, storage, and networking review site ServeTheHome and provide them in our Axautik Group LLC newsletter!
Live Last read · last published · next check
Latest posts
Helping Normalize NVIDIA Vera to AMD EPYC
This is a hotly requested topic, so we help cut through the framing
Qualcomm Dragonfly C1000 and HBC Are Coming
Some thoughts from discussions at Qualcomm Investor Day 2026
Everpure's Array Business Impacted by NAND, but Betting Its Future on a Huge Data Shift
This may be the bigger AI-SaaS shift
HPE is Taking a Diverging Path
At HPE Discover 2026, the stark contrast to Dell Tech World 2026 could not be ignored
Finding a Launch Date for Intel Xeon 7 Diamond Rapids
It is not as much of a mystery as you might think. We can narrow down the potential launch dates
I Attended 6 Big Keynotes Here Are My Thoughts
In the span of a few days, I sat through NVIDIA, Intel, Marvell, Qualcomm, Dell, and Supermicro keynotes
Here is Why the Qualcomm Cloud AI 100 is Not Everywhere
We compared it to common LPDDR options from AMD, NVIDIA, and Apple
The AI Boom is About to Reset Apple
You can see it in the company's changing lineup
The CPU Land Grab is Officially On
Today's Meta and AWS Graviton Deal along with Intel results paint a clear picture
The Low-Hanging AI Compute Fruit: A New PCIe GPU
The market needs this, but where is it?
Arm Dives into Data Center Chips
Here are some quick-reaction thoughts on the impact.
The Rise of Desk-Side AI: Why Employee-Scale Agentic Computing Is Inevitable
Where we are, and what needs to be solved (beyond just saying "security")
NVIDIA's Co-Packaged Optics Switch is Hinting at Something BIG
A fun note from a recent showcase of an engineering sample
Clearwater Questions after Intel's January 2026 Earnings Call
Highly touted, to almost hidden in just three months
The Memory Shortage is Having Another Impact on a Class of AI Workstations
A quick note asking about AMD Ryzen AI Max+ 395 Systems from CES 2026
10,000 Precision Welds and AI Clusters
Engineering for scale deployment is just different
This is Meta's Massive Plan to Recycle DDR4 Memory and Save Huge. Big wins for AMD and Marvell?
We found the prototypes. Every enterprise server vendor needs to offer this.
Intel Cancels its Mainstream Future Xeon: We Go Into the Impacts
A major shift in Intel Diamond Rapids is a sign of the times
With MLPerf Training 5.1 Configurations in Question We Look at the Server CPUs
Taking a CPU-centric look at the MLPerf Training benchmark