“I've collaborated in open source projects with, competed commercially against, and most recently worked with John as part of a tightly knit team. He helps define the archetype for effective community development and leadership. While a strong proponent of the promise of "open" he possesses a balanced, professional, and pragmatic view of where it's best placed and similarly how it's sustained. He's an impressive history of forming coherent and highly successful projects from amongst competing interests / agendas and business models and manages to balance maintaining an amicable dynamic with introducing tension to advance the state of the art. And ultimately, in a world in which "code wins," John's a winner. ”
Activity
-
Greeley, Colorado. That's where my gratitude in enterprise technology begins. I left the US Air Force and moved from Barksdale AFB/Shreveport, LA to…
Greeley, Colorado. That's where my gratitude in enterprise technology begins. I left the US Air Force and moved from Barksdale AFB/Shreveport, LA to…
Liked by John Griffith
-
Looking forward to moderating a panel discussion on microservices this Thursday, Sep 14th with panelists Daniel Bryant , CTO at SpectoLabs Ltd.…
Looking forward to moderating a panel discussion on microservices this Thursday, Sep 14th with panelists Daniel Bryant , CTO at SpectoLabs Ltd.…
Liked by John Griffith
Experience
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Education
-
-
-
-
-
-
Recommendations received
1 person has recommended John
Join now to viewOther similar profiles
Explore more posts
-
Epic Semi
As AI workloads scale from massive model training to real-time inference, compute performance means nothing without matching data throughput. With PCIe Gen5 and NVMe SSDs, we’re witnessing a true infrastructure shift not just incremental improvement. Moving from Gen3/4 to Gen5 isn’t a simple spec bump. It’s a strategic redesign of AI data center architecture where storage becomes a performance enabler, not a constraint. Read the full blog from Shahzad Ali to explore how Gen5 NVMe is transforming AI infrastructure. https://lnkd.in/dh4uSsi5 Naveed Sherwani Chloe Jian Ma Ahmad Alfaifi Andy Lu Collin Lee Shafy Eltoukhy #EpicSemi #AI #ArtificialIntelligence #DataCenter #Storage #NVMe #PCIeGen5 #GPU #TechLeadership #SSD
1 Comment -
Vroble.com
For advanced developers pushing the boundaries of system performance, **Processor Affinity** is a critical, often overlooked, optimization technique. By strategically binding processes or threads to designated CPU cores, engineers can significantly enhance cache utilization and minimize the performance overhead associated with frequent context switching. In high-throughput and latency-sensitive environments, such as real-time trading platforms or scientific simulations, implementing processor affinity can yield substantial gains, with some benchmarks showing up to a 29% improvement in throughput. This level of fine-grained control over CPU scheduling is instrumental in achieving predictable and peak performance. Consider integrating processor affinity into your performance tuning strategy for truly demanding workloads. It's a foundational principle for unlocking your hardware's full potential. #PerformanceEngineering #HPC #SystemsProgramming #DevOps #LowLatency #Optimization
-
Bob Bakh
AI teams are feeling the pressure right now. DRAM prices are climbing fast, NVMe lead times are stretching out years, and waiting on new hardware can slow everything down. But what if the answer isn’t buying more infrastructure but getting more out of what you already have? Next week, Steve McDowell (NAND Research) and WEKA’s Val Bercovici are sharing how teams are increasing GPU productivity, extending memory for demanding AI workloads, and working around hardware delays. If you’re building or scaling AI infrastructure, this should be a great session. 🗓️ Save your spot: http://spr.ly/6041heMjB
-
PyTorch
Thank you to everyone who joined "Inside Helion: Live Q&A with the Developers." In this clip, the team shares how Helion approaches kernel authoring by raising the level of abstraction and relying on autotuning to surface performant implementations across hardware. The discussion touches on portability, performance, and developer workflows. 🔗 Watch the full Live Q&A: https://lnkd.in/giJhbnA9 🔗 More from Jason Ansel at PyTorch Conference 2025: https://lnkd.in/g9VgD2AJ 🔗 Read the Helion blog: https://lnkd.in/g4ZRDsBh Featuring Jason Ansel, Oguz Ulgen, Will Feng, and Jongsok Choi. #PyTorch #Helion #OpenSourceAI #Performance
-
Larry Adams
For all my former RTM customers, there his help on the way in MariaDB 12.3+. There are other ways to get around write/update heavy workload like performing ETL's using SELECT INTO OUTFILE and following up with LOAD DATA INFILE to get around replication locking, but it's good to know that more solutions are on the way.
-
Vic Hung
I’ve been fortunate to work with leaders who truly value engineering fundamentals and clear technical thinking. One of my former managers, now retired, continues to write and share technical insights based on decades of real-world system experience. His articles reflect a deep system-level mindset — focusing on fundamentals, trade-offs, and how engineering decisions impact long-term product quality. For those interested in AI hardware, high-speed interconnects, and system architecture, his work is well worth following. https://lnkd.in/gW8AqUyY
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top contentOthers named John Griffith in United States
635 others named John Griffith in United States are on LinkedIn
See others named John Griffith