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
- Network-on-Chip (NoC)
- 12-bit, 400 MSPS SAR ADC - TSMC 12nm FFC
- DVB-S2 Demodulator
- UCIe PHY (Die-to-Die) IP
- UCIe-S 64GT/s PHY IP
Latest Blogs
- Enabling End-to-End EDA Flow on Arm-Based Compute for Infrastructure Flexibility
- Real PPA improvements from analog IC migration
- Design specification: The cornerstone of an ASIC collaboration
- The importance of ADCs in low-power electrocardiography ASICs
- VESA Adaptive-Sync V2 Operation in DisplayPort VIP