ConvNext is a Winner!
On Chimera QC Ultra it runs 28x faster than DSP+NPU
ConvNext is one of today’s leading new ML networks. Frankly, it just doesn’t run on most NPUs. So if your AI/ML solution relies on an NPU plus DSP for fallback, you can bet ConvNext needs to run on that DSP as a fallback operation. Our tests of expected performance on a DSP + NPU is dismal – less than one inference per second trying to run the ConvNext backbone if the novel Operators – GeLu and LayerNorm – are not present in the hardwired NPU accelerator and instead those Ops ran on the largest and most powerful DSP core on the market today.
This poor performance makes most of today’s DSP + NPU silicon solutions obsolete. Because that NPU is hardwired, not programmable, new networks cannot efficiently be run.
To read the full article, click here
Related Semiconductor IP
- Ultra Ethernet MAC & PCS 100G/200G/400G/800G
- Ethernet PCS 100G/200G/400G/800G/1.6T
- Ethernet MAC 100G/200G/400G/800G/1.6T
- Junction Over-Temperature Detector with Linear Centigrade-to-Voltage Output - X-FAB XT018
- Performance P570 Gen 3
Latest Blogs
- Inside the SiFive Performance™ P570 Gen 3: High Performance Efficiency for Next-Generation Consumer and Commercial Applications
- What the steam engine can teach us about modern chip design
- Automotive silicon in the era of AI, functional safety, and cybersecurity
- JPEG XS Officially Joins GenICam, The Machine Vision Standard Managed By EMVA
- Beyond PCIe Compliance: Why Stress Testing Is Crucial for Edge AI Deployments