Associate Professor at KTH Royal Institute of Technology
Joined November 2010
- Tweets149
- Following291
- Followers454
- Likes1.3K
Congratulations @WangChangjie13 on receiving the 2025 Applied Networking research prize for our NetConfEval work! 🏆
Now, let’s work on making LLMs also reliable and sustainable for code generation! 🚀
irtf.org/anrp/
Marco Chiesa retweeted
1/3
At CoNEXT ’24, @Skazza_94 presented our FAJITA paper for running stateful network functions at very high rates (~1.4 Tbps). PDF + video are at the @ERC_Research CoG ULTRA blog post below. Click below to see something perhaps unexpected in this work.
bit.ly/4g63His
Marco Chiesa retweeted
I am continuing to look for two more PhD students who wish to work with us on creating a dramatically more energy efficient and cheaper platform for AI inferencing (deadline Jan 17, 2025):
kth.se/lediga-jobb/768308?l=…
1/3
Marco Chiesa retweeted
At OvSCon'24, we will be presenting Emil Ståhl's master thesis work together with Red Hat on the potential of branch prediction to reduce cache misses in OVS.
Materials: lnkd.in/dHREkHd5
Presentation recording: lnkd.in/dCZ87se9
cc: @get_gianni_up , @marchiesa
We discovered that:
- Chasing a single pointer on a CPU may cost you 10 Mpps
- Many NF chains retrieve flow states based on different packet header fields, breaking the "exclusive memory access" guarantees of high-speed multi-core CPUs
In our CoNEXT'24 paper... (see comments)
We show that FAJITA can process up to 170Mpps with an NF consisting of a flow counter +a rate limiter and a load balancer. We then show how to improve by 50% the throughput of NF chains that match on different fields of the header, processing up to 90Mpps.
As usual, all code is available online!
This is joint work with @ghasemirahni
(first author) @alirezafarshin @Skazza_94 Gerald Q. Maguire Jr. @dmkostic
Paper: marchiesa.bitbucket.io/docs/…
Code: github.com/FAJITA-Packet-Pro…
Marco Chiesa retweeted
After all these OpenAI and Google announcements is there anything left for us academics to work on? Definitely! Three PhD student openings for working on a dramatically more energy efficient platform for AI inferencing: kth.se/lediga-jobb/720350?l=…
Do large language models facilitate network configuration? We investigate opportunities and challenges in operating network systems using LLMs in our accepted ACM CoNEXT'24 paper. Thanks, @WangChangjie13 @Skazza_94 @alirezafarshin @si_ferlin @dmkostic. More info in the comments.
Link to GitHub: github.com/NetConfEval/NetCo…
Link to HuggingFace: huggingface.co/datasets/NetC…
Marco Chiesa retweeted
I am looking for postdocs who want to work with us on fundamentally pushing the boundary of scientific knowledge when it comes to building Networked Systems for Machine Learning. #recruiting #ercfunded #postdoc #postdocposition #postdocjobs @ERC_Research
kth.varbi.com/en/what:job/jo…
Congrats to Mariano Scazzariello@Skazza_94 and Tommaso Caiazzi on their best paper award at Conext! The paper presentation is tomorrow (Wed) at 11:00
Marco Chiesa retweeted
We are looking for multiple postdocs who want to work with us on pushing the boundary of knowledge when it comes to building Networked Systems for Machine Learning. We offer a large degree of academic freedom and top experimental research infrastructure.
kth.varbi.com/en/what:job/jo…
