NASA Spends $50 Million to Develop Next-Gen Processor for Space Exploration

NASA has been using the same spaceflight computers for almost 30 years, but it won’t be much longer. The agency has awarded a $50 million contract to Arizona-based Microchip Technology Inc. to “architect, design, and deliver” a next-generation space-optimized processor. NASA expects the new chip, which will be used in future lunar and planetary missions, will be 100 times faster than the chips currently in use.

In an interview we conducted with Perseverance rover engineer Adam Steltzner, he told us the chip powering the robot was no faster than the CPU in a late 90s Mac computer — the phone in your pocket is an order of magnitude more powerful. However, you can’t just slap the latest Intel CPU in a spacecraft and call it a day. Space is a harsh environment with extreme temperatures and damaging radiation. Regular computer hardware tends not to last very long. The Ingenuity Mars helicopter is one notable exception. It runs on a Qualcomm smartphone SoC, and despite NASA’s low expectations, it’s still going strong on Mars.

Still, for mission-critical hardware, NASA needs the most robust designs it can get. That’s why Microchip Technology is designing a new chip known as the High-Performance Spaceflight Computing (HPSC) processor. This will mean a big improvement in the capabilities of future NASA missions, and not just in raw power. Microchip Technology promises “comprehensive Ethernet networking, advanced artificial intelligence/machine learning processing, and connectivity support.” At the same time, the processor will have the best fault tolerance, radiation hardening, and security architecture possible.

NASA began the bidding process for its new spaceflight processor last year. It specified that the new chip is capable of efficiently handling different workloads. Current spaceflight processors are designed to handle the most computationally intensive part of the mission; for example, the rocket sled landing sequence employed by Curiosity and Perseverance. However, this means the system has more power than it needs for day-to-day activities, and that loss of efficiency can be a big problem when power is in short supply. Microchip Technology says the HPSC processor will be able to scale performance up and down as needed and will even be able to completely shut off certain functionality to save power.

Microchip Technology has a goal of completing the HPSC processor within three years. The $50 million fixed-cost contract could end up being a steal, too. NASA used the last designs for decades, and this one could be similarly long-lived as the agency gears up for a new era of planetary exploration. Other government agencies could also find it useful in applications with similarly stringent requirements.

Continue reading

Report: Nvidia’s Next-Gen GPU Could Pack 18,432 CUDA Cores, 64TFLOPS

Nvidia's upcoming Ampere, which may be named after computing pioneer Ada Lovelace, could pack as many as 18,432 CUDA cores, a substantial increase over current top-end cards like the RTX 3090.

AMD Unveils Radeon RX 6000M GPUs for Next-Gen Gaming Notebooks

AMD has announced new Radeon Mobile GPUs.

For Next-Generation CPUs, Not Moving Data Is the New 1GHz

We've had the race to 0W. Next comes the race for zero-movement.

Alder Lake Extravaganza: Intel Unloads Details on its Next-Gen CPU

Intel's desktop CPUs have been stuck on 14nm for over six years now, and the cracks in that process node have been showing for at least two. The manufacturer was finally ready to share details on Alder Lake, Gracemont, and Golden Cove with us, and it had a lot to talk about.