Supermicro, Intel, and Micron Achieve Record-Breaking Results in Quantitative Trading Benchmark
Date: October 28, 2025 | Location: New York City
Super Micro Computer, Inc. (SMCI), a prominent provider of total IT solutions, has announced a groundbreaking achievement in collaboration with Intel and Micron as they set new records for the STAC-M3™ benchmark during the STAC Summit held in New York City. This collaboration highlights the commitment of these technology leaders to push the boundaries of performance in the financial services sector.
1. Understanding the STAC-M3 Benchmark
The STAC-M3 benchmark, managed by the STAC Benchmark™ Council, evaluates real-time quantitative trading capabilities using simulated market bid-ask and settled trade data across thousands of assets. It encompasses a full-stack approach, measuring query response times in multi-user environments, making it essential for algorithmic trading, risk management, and backtesting strategies utilized by banks, hedge funds, and trading exchanges.
The benchmark includes two primary testing sets: Antuco, which evaluates performance based on a smaller dataset, and Kanaga, which simulates larger datasets under real-world conditions. The recent tests revealed that Supermicro's configuration achieved remarkable results, setting new world records on 19 out of 24 Kanaga benchmarks, including all 10 user benchmarks for 50 and 100 users.
2. Record-Breaking Performance Metrics
Supermicro's dual-CPU Petascale servers demonstrated exceptional performance, completing the compute-intensive 100-user unpredictable interval statistics benchmark 36% faster than the previous record while utilizing 62% fewer CPU cores. The configuration comprised six 2U Supermicro storage servers, occupying a total of just 12U rack space, a significant reduction compared to previous record-holding configurations that required up to 44U. This innovative setup supported an unprecedented storage capacity of 1.6 PiB, utilizing Micron 9550 NVMe SSDs, which is over six times the capacity of the previous record.
Comments from Industry Leaders
Vik Malyala, SVP of Technology & AI at Supermicro, expressed pride in the collaboration with Intel, Micron, and KX Software, emphasizing that these record-breaking test results are a testament to their workload-optimized product design, further enabling their clients to gain a competitive edge in high-frequency trading.
Srini Krishna, Intel Fellow and Head of Data Center Products, highlighted the importance of managing large data volumes and maintaining low latency in high-frequency trading. He pointed out that the results showcased the Intel Xeon 6's ability to deliver record-setting performance, allowing trading teams to make timely decisions and assess risks effectively.
Alvaro Toledo, Vice President and General Manager for the Americas Core Data Center Business Unit at Micron, noted the critical role of speed in trading and how their NVMe SSDs contributed to turning tick data into actionable insights more rapidly than ever.
3. Technical Configuration of the Test
The successful test configuration included:
- Supermicro SSG-222B-NE3X24R Petascale all-flash storage servers: Six units, each outfitted with dual Intel Xeon 6767P CPUs (64 cores, 128 threads, turbo up to 3.9GHz).
- Memory: 16 Micron 128GB DDR5 RDIMMs, providing a total of 2.048 TB per system at 6400 MT/s.
- Storage: 24 Micron 9550 NVMe SSDs, each with a capacity of 12.8TB.
- Database Software: KX Software's kdb+ 4.1, optimized for real-time and historical data analytics.
4. Looking Ahead
The results of this benchmark will be formally presented at the STAC Summit, highlighting the collaborative efforts of Supermicro, Intel, Micron, and KX Software. These advancements not only underscore the technological capabilities of these companies but also the future of quantitative trading, where speed and data analysis are paramount.
Supermicro continues to solidify its position as a leader in Application-Optimized Total IT Solutions, paving the way for further innovations in AI, Cloud, and high-performance computing as they cater to an increasingly data-driven financial landscape.