cd /news/ai-chips/nvidias-cuda-13-4-integration-why-sh… · home › topics › ai-chips › article
[ARTICLE · art-143936] src=asiaai.fyi ↗ pub= topic=ai-chips verified=true sentiment=↑ positive

Nvidia’s CUDA 13.4 Integration: Why Shibaura Institute’s Scheme Speeds Up Legacy GPUs

NVIDIA has adopted the "Ozaki Scheme II" computational method, developed by Professor Katsuhiro Ozaki and his team at Shibaura Institute of Technology, into its CUDA toolkit 13.4, according to EE Times Japan. The scheme combines multiple low-precision AI calculations to achieve double-precision accuracy at higher speeds on existing NVIDIA Ampere-generation GPUs and newer, without requiring additional hardware circuits. The adoption speeds up legacy GPUs for scientific and engineering tasks that demand high numerical accuracy.

by read2 min views1 publishedOct 2, 2026
Nvidia’s CUDA 13.4 Integration: Why Shibaura Institute’s Scheme Speeds Up Legacy GPUs
Image: Asiaai (auto-discovered)

East Asian Technology Intelligence

Japan & China tech news — translated, contextualized, and delivered for Western readers.

Free. Unsubscribe anytime.

This story ran in Issue #114, alongside three other stories. Semiconductors & Hardware

What happened

NVIDIA has adopted the “Ozaki Scheme II” computational method, developed by Professor Katsuhiro Ozaki and his team at Shibaura Institute of Technology, into its latest CUDA toolkit 13.4. This new scheme accelerates double-precision calculations by combining multiple low-precision AI computations, making existing NVIDIA GPUs from the Ampere generation onwards significantly faster for scientific and engineering tasks without requiring new hardware.

Key details

  • NVIDIA’s latest CUDA toolkit 13.4 incorporates “Ozaki Scheme II” from Shibaura Institute of Technology.
  • Ozaki Scheme II combines multiple low-precision AI calculations to achieve double-precision accuracy at significantly higher speeds.
  • The method operates on existing NVIDIA Ampere generation GPUs and newer, without requiring additional hardware circuits.

What remains unclear

  • What is the performance overhead for Ozaki Scheme II compared to native double-precision calculations?
  • Are there specific scientific domains where Ozaki Scheme II offers a disproportionate advantage?
  • How widely will the CUDA toolkit 13.4 with Ozaki Scheme II be adopted by the broader scientific community compared to the initial Ozaki Scheme I?

Source: EE Times Japan · source published October 2, 2026

📊 Featured Chart

Source: NVIDIA

The adoption of Ozaki Scheme II leverages GPUs designed primarily for low-precision AI tasks to perform high-precision scientific computing more efficiently. This development enhances the utility of NVIDIA’s existing hardware base for a broader range of applications, particularly in fields requiring high numerical accuracy like scientific simulations and complex engineering. For Western readers, it highlights Japan’s continuing, if sometimes overlooked, contributions to fundamental computational methods in semiconductor design.

For the wider picture, see [Japan Semiconductor Ecosystem](https://asiaai.fyi/guides/japan-semiconductor-ecosystem/).

This story appeared in [AsiaAI.FYI Issue #114](https://asiaai.fyi/tsmc-qualcomm-exclusive-2nm-orders/).
	Get this in your inbox each week —
	[subscribe free](https://asiaaifyi.beehiiv.com/subscribe).
── more in #ai-chips 4 stories · sorted by recency
── more on @nvidia 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/nvidias-cuda-13-4-in…] indexed:0 read:2min 2026-10-02 · —