Urgent.News

What's breaking now, across thousands of outlets.

Tech

Beyond Nvidia: How Singapore is earning its place in the semiconductor value chain

It occupies the part which AI has made expensive to get wrong – this is starting to be reflected by the listed...

Singapore is establishing itself as a significant player in the semiconductor value chain, despite not having listed companies in the sector. The government has invested heavily in creating a physical ecosystem through its semiconductor industry, with projects like Silicon Box and VSMC building advanced manufacturing facilities. Despite this, local equity markets have not yet appreciated the value of these investments.

However, Singapore's listed semiconductor companies are showing strong financial performance. AEM reported a 904% increase in net profit year-over-year and a significant boost in revenue guidance. Similarly, UMS Integration posted substantial revenue and profit growth. These companies are not directly competing with industry giants like Nvidia, but rather are part of a more specialized ecosystem focused on advanced packaging, testing, interconnect, and thermal management.

The shift towards these downstream areas is driven by the increasing complexity and cost of semiconductor manufacturing, making the value downstream higher. With more performance gains coming from improved packaging, testing, and cooling, Singapore's expertise in these areas becomes more critical. This shift benefits Singapore, especially in applications that require high power density, such as AI systems.

As these systems become more valuable, the risk of failure also increases, making the role of suppliers even more crucial. Suppliers that can improve yield, reliability, throughput, power efficiency, or performance will have stronger economics and more defensible positions. Singapore's physical advantages, such as its established semiconductor ecosystem, have attracted major players like Micron and Applied Materials, creating a mutually reinforcing cluster of specialized suppliers and talent.

The government's commitment to the sector is further evidenced by the recent S$800 million investment in a semiconductor program focused on advanced packaging and photonics. The presence of large-scale facilities like Micron's US$24 billion wafer fab and UMC's US$5 billion production plant underscores Singapore's growing importance in the global semiconductor landscape.

As more companies move into more technically demanding and critical areas, Singapore is well-positioned to maintain its advantage by leveraging its existing ecosystem and infrastructure.

Written by urgent.news from The Business Times - Companies & Markets's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at businesstimes.com.sg →

More in Tech

API Rate Limiting: What Actually Breaks When You Get It Wrong

Most teams add rate limiting to their API as an afterthought — usually right after something has already gone wrong. A scraper hammers an endpoint, a client integration goes into a retry loop, or a…

  • Simply setting a single request limit is insufficient as all requests don't have equal costs.
  • Centralize rate limit counters across multiple instances to prevent overloading the system.

At the edge, the number that matters is memory - not throughput (specially in Ramageddon)

We rebuilt LF Edge eKuiper in Rust and ran it against eKuiper, Telegraf and Redpanda Connect on five real MQTT workloads — one core, 1 GB of memory, and output checked message-by-message.

  • Memory usage, not throughput, is key metric in resource-constrained environments
  • Rekuiper, Rust-based engine, maintained memory usage between 5 and 10 MB
  • Design choices include bounded queues, incremental window aggregation, offline sink cache

My Extraction Score Was 0.08 and the Model Was Innocent: Rebuilding the Ruler

Update — v0.3.1 released. CauterRule is now live on GitHub and PyPI . It turns repeated agent failures into permanent standing rules — extract, replay-test, promote.

  • CauterRule tool evaluates cloud models across 40 corpora and 4,742 trajectory-runs
  • Extraction-accuracy metric initially 0.08, later adjusted to focus on trigger portion
  • Adjusted metric shows agreement scores from 0.74 to 0.93, accurate performance representation

More from Sunday 13 September →