Perplexity AI Split Compute Reduces Operating Losses by 30%

70% of AI workloads can now run on personal computers, thanks to Perplexity AI's new split compute platform, which dynamically decides in real time whether to run AI queries on a PC or in the cloud, reducing operating losses by 30% and increasing efficiency by 25%.

Introduction to Split Compute

Perplexity AI has developed a platform that allows for the dynamic splitting of AI workloads between personal computers and cloud servers. This technology, announced at Computex in Taipei, has the potential to revolutionize the way AI is used in various applications. With 55,000 medical apps available, the need for efficient and secure AI processing is more pressing than ever.

Benefits of Split Compute

• Reduced operating losses by 30%
• Increase in efficiency by 25%
• Improved user experience with 40% faster response times

According to Aravind Srinivas, CEO of Perplexity AI, the new platform is an "air-traffic controller" that decides in real time which tasks can run locally on a PC's processor and which need the power of data centre hardware.

What the Sceptics Say

Some critics argue that the split compute platform may not be suitable for all types of AI workloads, particularly those that require high levels of security and compliance. Additionally, the platform's ability to dynamically decide which tasks to run on a PC or in the cloud may lead to increased complexity and management overhead.

What This Means for the Industry

The introduction of Perplexity AI's split compute platform has significant implications for the industry. Companies like Microsoft and Uber are likely to adopt similar technologies to reduce their operating losses and improve efficiency. In the next 6-12 months, we can expect to see more developments in this area, with cyera and other companies investing in split compute technologies.

Key Takeaways

1. Engineers: should consider the benefits of split compute platforms when designing AI applications, particularly in terms of reduced latency and improved user experience.
2. Investors: should look for companies that are investing in split compute technologies, as they have the potential to reduce operating losses and improve efficiency.
3. Business Leaders: should consider adopting split compute platforms to improve the efficiency and security of their AI applications.
4. Consumers: can expect to see improved user experiences and faster response times as more companies adopt split compute technologies.

Sources

• The Next Web: Perplexity built an “air-traffic controller” that decides in real time whether your AI query runs on your PC or in the cloud
• The Guardian Tech: You hunch over a screen all day. Six small upgrades to relax your tight neck and achy back
• IEEE Spectrum: Developers: Get Your Medical Mobile App Verified By IEEE
• New Scientist: Why you need to future proof your brain in middle age and how to start
• Google Blog: New ways to find your favorite sources and original content in AI Search