Skip to Content

1,000 Cores on a Chip

Rapport’s Kilocore chip makes quick work of video processing.

Today’s hottest microprocessors for consumer PCs, Intel’s Core Duo and AMD’s Athlon 64 X2, combine two central processing units – or “cores” – on a single chip, where they can divide up tough jobs like encrypting data or processing high-definition video. Intel and AMD have begun to talk about “quad core” chips and even eight-core devices, which might be on the market by 2008. But in Redwood City, CA, there’s a small company called Rapport that’s already working on a 1,000-core chip.

Rapport’s debut chip, available later this year, has 256 processing elements. (Courtesy of Rapport)

Called the Kilocore1025 and expected to ship in mid-2007, the chip is designed to power handheld devices like game and media players. It includes an IBM Power PC CPU for general-purpose tasks but a whopping 1,024 additional “processing elements.” Each element handles only eight bits of data at a time, in contrast to 32 or 64 bits for today’s leading processors, and runs at the relatively low clock speed of 100 megahertz, far slower than the two to three gigahertz of today’s notebook and desktop PCs. But that means the Kilocore chip consumes only one-tenth as much power as Intel’s latest notebook PC chips.

And while the Kilocore’s individual cores run slowly, they can work in parallel to churn quickly through tasks like streaming video, says Fred Furtek, a lead chip architect for Rapport. That should translate into smooth video without the chops and hiccups that occur when your cell phone or PC can’t keep up with a video clip.

A debut chip with 256 processing elements will ship later this year. But before Rapport’s chips can go into next-generation mobile devices, the company must improve the software-development tools that let programmers take advantage of Kilocore’s parallel-processing architecture. “The graphics problem is wonderfully dividable,” says PC industry veteran Roger Kay, president of market research firm Endpoint Associates – so it’s relatively easy to handle with parallel processors. But without the right development tools, Kay says, Rapport will have difficulty selling its ideas to device makers.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.