Stefan Abi-Karam
@stefanabikaram.com
Graduate Student (@SharcLab) + Research Faculty at @GeorgiaTech 🐝
Working on digital hardware design + AI
https://stefanabikaram.com/
Working on digital hardware design + AI
https://stefanabikaram.com/
Reposted by Stefan Abi-Karam
Our next #YUG will be with Matt Young, talking about triple modular redundancy.
Join us at 18:00 CET / 22:30 IST / 09:00 PT on Thursday 9th October.
Sign up to our mailing list to get a reminder before the event: blog.yosyshq.com/newsletter/
Join us at 18:00 CET / 22:30 IST / 09:00 PT on Thursday 9th October.
Sign up to our mailing list to get a reminder before the event: blog.yosyshq.com/newsletter/
September 29, 2025 at 10:54 AM
Our next #YUG will be with Matt Young, talking about triple modular redundancy.
Join us at 18:00 CET / 22:30 IST / 09:00 PT on Thursday 9th October.
Sign up to our mailing list to get a reminder before the event: blog.yosyshq.com/newsletter/
Join us at 18:00 CET / 22:30 IST / 09:00 PT on Thursday 9th October.
Sign up to our mailing list to get a reminder before the event: blog.yosyshq.com/newsletter/
Reposted by Stefan Abi-Karam
I will work on exploring applications of Aristotle to the formal verification of hardware. This job is a perfect intersection of hardware design and verification, functional programming, formal methods and machine learning, bringing together several threads of my career so far.
September 29, 2025 at 7:05 PM
I will work on exploring applications of Aristotle to the formal verification of hardware. This job is a perfect intersection of hardware design and verification, functional programming, formal methods and machine learning, bringing together several threads of my career so far.
"High-Level Synthesis Synthesis"
stefanabikaram.com/writing/hls-...
Some fun writing on hardware design pedantry
stefanabikaram.com/writing/hls-...
Some fun writing on hardware design pedantry
Stefan Abi-Karam
Personal webiste for Stefan Abi-Karam.
stefanabikaram.com
September 29, 2025 at 3:41 AM
"High-Level Synthesis Synthesis"
stefanabikaram.com/writing/hls-...
Some fun writing on hardware design pedantry
stefanabikaram.com/writing/hls-...
Some fun writing on hardware design pedantry
Reposted by Stefan Abi-Karam
[New Blog Post] Proving the Infinitude of Primes in Knuckledragger #python #logic www.philipzucker.com/knuckle_prim...
Proving the Infinitude of Primes in Knuckledragger
Cody challenged me to prove that there are an infinitely many primes in Knuckledragger, saying it’s the minimum thing to do to demonstrate you have a proof assistant. https://en.wikipedia.org/wiki/Euc...
www.philipzucker.com
September 28, 2025 at 9:05 PM
[New Blog Post] Proving the Infinitude of Primes in Knuckledragger #python #logic www.philipzucker.com/knuckle_prim...
Reposted by Stefan Abi-Karam
i built a pages server for Codeberg that's way cheaper to operate than the existing one, please enjoy grebedoc.dev/
Grbrmlnd
grebedoc.dev
September 15, 2025 at 2:51 PM
i built a pages server for Codeberg that's way cheaper to operate than the existing one, please enjoy grebedoc.dev/
Reposted by Stefan Abi-Karam
I got inspired at orconf, so now play.spade-lang.org supports a full in-browser flow for submitting to tinytapeout.com/
Still experimental, so if you try it and run into any issues, let me know :)
Still experimental, so if you try it and run into any issues, let me know :)
September 20, 2025 at 9:12 AM
I got inspired at orconf, so now play.spade-lang.org supports a full in-browser flow for submitting to tinytapeout.com/
Still experimental, so if you try it and run into any issues, let me know :)
Still experimental, so if you try it and run into any issues, let me know :)
Reposted by Stefan Abi-Karam
A lovely Xilinx 7-series FPGA chip layout of a sorting network produced from a dependently typed DSL in Haskell. More stuff like this during my keynote talk at ICFP 2025 in October in Singapore. icfp25.sigplan.org/details/icfp...
September 22, 2025 at 2:25 AM
A lovely Xilinx 7-series FPGA chip layout of a sorting network produced from a dependently typed DSL in Haskell. More stuff like this during my keynote talk at ICFP 2025 in October in Singapore. icfp25.sigplan.org/details/icfp...
Reposted by Stefan Abi-Karam
I did some more hacking to produce a totally floorplanned 64-input sorter circuit on a Xilinx XC7A200T FPGA, on the LHS picture shown as a tight rectangular block, on the RHS a close-up that shows the butterfly wiring pattern.
September 23, 2025 at 11:25 PM
I did some more hacking to produce a totally floorplanned 64-input sorter circuit on a Xilinx XC7A200T FPGA, on the LHS picture shown as a tight rectangular block, on the RHS a close-up that shows the butterfly wiring pattern.
Reposted by Stefan Abi-Karam
Check out the fantastic open source hardware hacker content from this weekend’s ORCONF’25!
fossi-foundation.org/orconf/2025?...
HT/thanks @mattvenn.net
fossi-foundation.org/orconf/2025?...
HT/thanks @mattvenn.net
ORConf 2025: September 12-14, 2025 in Valencia, Spain
The ORConf conference is a weekend of presentations and networking
dedicated to free and open source silicon. It's an event for the open
source semiconductor community, and is run by volunteers from t...
fossi-foundation.org
September 16, 2025 at 2:29 PM
Check out the fantastic open source hardware hacker content from this weekend’s ORCONF’25!
fossi-foundation.org/orconf/2025?...
HT/thanks @mattvenn.net
fossi-foundation.org/orconf/2025?...
HT/thanks @mattvenn.net
Shout out to all the CGRA folks, all jokes are in good fun
September 8, 2025 at 1:43 AM
Shout out to all the CGRA folks, all jokes are in good fun
Reposted by Stefan Abi-Karam
new preprint, tl;dr:
• LLMs match or exceed SOTA strategies on chemical reaction optimizations.
• we built the Iron Mind platform and we hope that it can serve as a new benchmark for both reaction optimizers and foundation models.
a thread [1/5]
• LLMs match or exceed SOTA strategies on chemical reaction optimizations.
• we built the Iron Mind platform and we hope that it can serve as a new benchmark for both reaction optimizers and foundation models.
a thread [1/5]
September 6, 2025 at 5:59 PM
new preprint, tl;dr:
• LLMs match or exceed SOTA strategies on chemical reaction optimizations.
• we built the Iron Mind platform and we hope that it can serve as a new benchmark for both reaction optimizers and foundation models.
a thread [1/5]
• LLMs match or exceed SOTA strategies on chemical reaction optimizations.
• we built the Iron Mind platform and we hope that it can serve as a new benchmark for both reaction optimizers and foundation models.
a thread [1/5]
Reposted by Stefan Abi-Karam
Lindsey Kuper's* group has produced this fantastic zine on choreographic programming that folks should definitely check out: decomposition.al/blog/2024/12...
(* can't seem to find Lindsey here but please tag if you know the handle)
(* can't seem to find Lindsey here but please tag if you know the handle)
“Communicating Chorrectly with a Choreography” is out!
decomposition.al
December 6, 2024 at 4:16 AM
Lindsey Kuper's* group has produced this fantastic zine on choreographic programming that folks should definitely check out: decomposition.al/blog/2024/12...
(* can't seem to find Lindsey here but please tag if you know the handle)
(* can't seem to find Lindsey here but please tag if you know the handle)
Reposted by Stefan Abi-Karam
Today at ASPLOS Chengsong Tan will present our work with @wicko3.bsky.social on mechanical formalisation of CXL.cache - check out the paper! www.doc.ic.ac.uk/~afd/papers/... #CXL
April 2, 2025 at 8:56 AM
Today at ASPLOS Chengsong Tan will present our work with @wicko3.bsky.social on mechanical formalisation of CXL.cache - check out the paper! www.doc.ic.ac.uk/~afd/papers/... #CXL
Reposted by Stefan Abi-Karam
This study of deep learning optimizers found that previously claimed speedups of alternative optimizers over AdamW in language model pretraining were often inflated due to methodological shortcomings like unequal hyperparameter tuning and limited evaluation.
September 4, 2025 at 12:10 AM
This study of deep learning optimizers found that previously claimed speedups of alternative optimizers over AdamW in language model pretraining were often inflated due to methodological shortcomings like unequal hyperparameter tuning and limited evaluation.
Reposted by Stefan Abi-Karam
Benchmarking the variety of different proposed LLM optimizers: Muon, AdEMAMix, ... all in the same setting, tuned, with varying model size, batch size, and training duration!
They find that AdEMAMix is SOTA, the importance of weight decay, or how Muon “prefers” WSD, and more...
They find that AdEMAMix is SOTA, the importance of weight decay, or how Muon “prefers” WSD, and more...
September 4, 2025 at 12:03 AM
Benchmarking the variety of different proposed LLM optimizers: Muon, AdEMAMix, ... all in the same setting, tuned, with varying model size, batch size, and training duration!
They find that AdEMAMix is SOTA, the importance of weight decay, or how Muon “prefers” WSD, and more...
They find that AdEMAMix is SOTA, the importance of weight decay, or how Muon “prefers” WSD, and more...
Reposted by Stefan Abi-Karam
Paper: Benchmarking Optimizers for Large Language Model Pretraining
( arxiv.org/abs/2509.01440 )
Repo: github.com/epfml/llm-op...
( arxiv.org/abs/2509.01440 )
Repo: github.com/epfml/llm-op...
Benchmarking Optimizers for Large Language Model Pretraining
The recent development of Large Language Models (LLMs) has been accompanied by an effervescence of novel ideas and methods to better optimize the loss of deep learning models. Claims from those method...
arxiv.org
September 4, 2025 at 12:03 AM
Paper: Benchmarking Optimizers for Large Language Model Pretraining
( arxiv.org/abs/2509.01440 )
Repo: github.com/epfml/llm-op...
( arxiv.org/abs/2509.01440 )
Repo: github.com/epfml/llm-op...
I have yet to read “Computer Architecture: A Quantitative Approach”
September 4, 2025 at 2:38 AM
I have yet to read “Computer Architecture: A Quantitative Approach”
Reposted by Stefan Abi-Karam
We were cooked the moment operating system UIs stopped shipping looking like this
September 2, 2025 at 10:37 PM
We were cooked the moment operating system UIs stopped shipping looking like this
Reposted by Stefan Abi-Karam
Reposted by Stefan Abi-Karam
Over the weekend I released the next part of Isle #FPGA Computer covering bitmap graphics.
projectf.io/isle/bitmap-...
I've added more docs and tests, but I'm still finding the right approach to the Isle blog and working on diagrams.
projectf.io/isle/bitmap-...
I've added more docs and tests, but I'm still finding the right approach to the Isle blog and working on diagrams.
September 1, 2025 at 9:43 AM
Over the weekend I released the next part of Isle #FPGA Computer covering bitmap graphics.
projectf.io/isle/bitmap-...
I've added more docs and tests, but I'm still finding the right approach to the Isle blog and working on diagrams.
projectf.io/isle/bitmap-...
I've added more docs and tests, but I'm still finding the right approach to the Isle blog and working on diagrams.
Reposted by Stefan Abi-Karam
This Friday I successfully defended my PhD 🎉 If you want to watch a recording of the defense there is a recording here vimeo.com/1114208379/89dcd4f302
Improved Tooling for Digital Hardware Development - Frans Skarman
This is "Improved Tooling for Digital Hardware Development - Frans Skarman" by Linköping University on Vimeo, the home for high quality videos and…
vimeo.com
August 31, 2025 at 6:54 AM
This Friday I successfully defended my PhD 🎉 If you want to watch a recording of the defense there is a recording here vimeo.com/1114208379/89dcd4f302
Reposted by Stefan Abi-Karam
Then and now, Prof. Cong and students trailblaze new design automation research for mainstream high level synthesis tools and myriad applications.
Chuck Thacker made beautiful computers with FPGAs, so this award seems particularly apt.
web.mit.edu/6.173/www/cu...
www.cl.cam.ac.uk/~swm11/examp...
Chuck Thacker made beautiful computers with FPGAs, so this award seems particularly apt.
web.mit.edu/6.173/www/cu...
www.cl.cam.ac.uk/~swm11/examp...
April 9, 2025 at 3:52 PM
Then and now, Prof. Cong and students trailblaze new design automation research for mainstream high level synthesis tools and myriad applications.
Chuck Thacker made beautiful computers with FPGAs, so this award seems particularly apt.
web.mit.edu/6.173/www/cu...
www.cl.cam.ac.uk/~swm11/examp...
Chuck Thacker made beautiful computers with FPGAs, so this award seems particularly apt.
web.mit.edu/6.173/www/cu...
www.cl.cam.ac.uk/~swm11/examp...
Reposted by Stefan Abi-Karam
arxiv 📄 Evaluating Large Language Models for Automatic Register Transfer Logic
Generation via High-Level Synthesis
http://arxiv.org/abs/2408.02793v1
The ever-growing popularity of large language models (LLMs) has resulted in
their increasing adoption for hardware design and verification. Prio...
Generation via High-Level Synthesis
http://arxiv.org/abs/2408.02793v1
The ever-growing popularity of large language models (LLMs) has resulted in
their increasing adoption for hardware design and verification. Prio...
December 6, 2024 at 6:53 PM
arxiv 📄 Evaluating Large Language Models for Automatic Register Transfer Logic
Generation via High-Level Synthesis
http://arxiv.org/abs/2408.02793v1
The ever-growing popularity of large language models (LLMs) has resulted in
their increasing adoption for hardware design and verification. Prio...
Generation via High-Level Synthesis
http://arxiv.org/abs/2408.02793v1
The ever-growing popularity of large language models (LLMs) has resulted in
their increasing adoption for hardware design and verification. Prio...
Reposted by Stefan Abi-Karam
Zedong Peng, Zeju Li, Mingzhe Gao, Qiang Xu, Chen Zhang, Jieru Zhao: ForgeHLS: A Large-Scale, Open-Source Dataset for High-Level Synthesis https://arxiv.org/abs/2507.03255 https://arxiv.org/pdf/2507.03255 https://arxiv.org/html/2507.03255
July 8, 2025 at 6:29 AM
Zedong Peng, Zeju Li, Mingzhe Gao, Qiang Xu, Chen Zhang, Jieru Zhao: ForgeHLS: A Large-Scale, Open-Source Dataset for High-Level Synthesis https://arxiv.org/abs/2507.03255 https://arxiv.org/pdf/2507.03255 https://arxiv.org/html/2507.03255
Reposted by Stefan Abi-Karam
Huge congrats to Yann Herklotz (@ymherklotz) for successfully defending his PhD thesis today, all about his proven-in-Coq high-level synthesis tool. And enormous thanks to George Constantinides (@gconstantinides) and Xavier Leroy for their thoughtful and thorough examining.
November 16, 2024 at 9:57 AM
Huge congrats to Yann Herklotz (@ymherklotz) for successfully defending his PhD thesis today, all about his proven-in-Coq high-level synthesis tool. And enormous thanks to George Constantinides (@gconstantinides) and Xavier Leroy for their thoughtful and thorough examining.