Peter Alvaro
University of California-Santa Cruz, CA·Computer Science
Large online services in finance, health and education process far more data than a single computer can hold. Today they usually split data across many machines, and each machine works mostly alone and coordinates by sending messages.
- Fields
Computer Science
Engineering
- Topics
distributed systems
operating systems
computer networks
- Grants
- 1 known grant, last ended Sep 2026
What they work on
- Design memory protocols that run over network fabrics so that many computers can reliably and quickly reach large pools of remote, disaggregated memory.
- Determine the appropriate role of computation-enhanced networks, meaning network hardware that can process data in transit, within a shared-memory distributed system.
- Provide fault tolerance and robustness for persistent memory, memory that keeps data without power, and develop programming models and long-lived data references for an operating system co-designed with the network.
More about the research
This "shared-nothing" design is becoming hard to change as applications evolve. It also struggles as computer hardware becomes more varied, and it forces programmers to decide where data lives and when it moves. The older alternative, "shared-everything," lets all machines use one common memory pool. That design was long considered impractical, but faster networks and new persistent memory, which keeps its contents when the power is off, may now make it possible.
Why it matters
If shared memory across a network becomes practical, programmers building large data services in industry, government and academia could write simpler software. The system itself would handle where data is placed and when it moves. This could make it easier to adapt data-heavy applications to new and more varied computing hardware.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Arvind Krishnamurthy
University of Washington, WA·Computer Science
Design NetFusion, a rack-level pool of programmable network interface cards (SmartNICs) that perform packet and network-function processing for many servers, with safe and fair sharing among applications.
Simon Peter
University of Washington, WA·Computer Science
Design a programmable memory controller that moves data among memory tiers connected by Compute Express Link based on usage patterns, module age and reliability.
Remzi Arpaci-Dusseau
University of Wisconsin-Madison, WI·Computer Science
Characterize dependencies across blockchain computing, storage and network resources to build a foundation for classifying faults that cause transaction data loss.
The school
University of California-Santa Cruz: 15 labs on PI Pages, $218M a year on research, 183 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.