Super Edge Medical SoC (SEMC)

By FastStream Technologies

Overview:

Post Covid 19, the biggest bet for revival of the industry is on 5G proliferation across the world. It is widely expected that 5G’s Enhanced Mobile broadband (eMMB) with speeds as high as 20X of 4G speed, Ultra reliable and Low Latency Communication ( URLLC) and massive Machine type connectivity (mMTC) will transform the world. One of the most important use cases for 5G is its use in Medical field, covering diverse applications like remote Surgery, remote patient monitoring etc with immense business potential. For example, Telemedicine market is expected to grow at a compound annual growth rate of 16.5% from 2017 to 2023, parallel with the emergence and roll-out of 5G.

The healthcare industry itself has technologically got transformed over the years and now generates massive amounts of data each day. A single patient can generate hundreds of gigabytes of data each day, from patient medical records to the large image files generated by MRI, CAT, or PET scans. For example, depending on detector area and pixel size, digital mammograms may have an image matrix of up to 4,800 × 6,000 pixels with a file size of more than 50 MBytes.

In parallel, Artificial Intelligence and Machine Learning (AI/ML) are transforming completely the medical diagnostic field. Medical imaging is poised to be used heavily in early diagnosis, detection and treatment of diseases. For example, early detection of pancreatic cancer is challenging because cancer-specific symptoms occur only at an advanced stage, and a reliable screening tool to identify high-risk patients is lacking. Researchers are trying to find a solution by developing an artificial neural network (ANN), training it and testing it with existing health data. We are certainly seeing emergence of AI assisted work flows in health industry heavily relying on Medical Imaging and AI.

The Super Edge Medical SoC (SEMC) being designed by FastStream Technologies, India is to cater to this specific requirement of the Healthcare industry. OEMs and System Designers will be able to leverage this fully integrated Edge SoC to create solutions which are not possible today. One of the major differentiator for this SoC solution is that it can operate with full 5G connectivity and if needed, it can operate in full fledged Edge Mode with Edge Computing, Edge Inference, Edge Storage and even private Edge Cloud.

Major blocks of the SEMC:

1- Edge Compute Engine : 8 X Cortex-A78 CPU cores, running up to 3.0 GHz – Medical equipments have life of 8-10 years and we have incorporated enough compute power to take care of present and future needs. This is also needed for running large machine learning models of future.

RAM- 32 GB , Four 72-bit (64-bit + ECC) 3.2 GT/s LP DDR4 SDRAM memory controllers with ECC

4 X8 GB 2666 MHz- Module Support

2- 4K/8K Video Processing: Real Time processing of 4K/8K Video (up to 7680X4320) to be rendered on huge 98 inch multiple displays is made possible by 4 X Mali-G78 Second-generation premium GPU based on the Mali Valhall architecture.

3- Edge Inference Processor : 1 X ARM NPU ETHOS-N78 Highly Scalable and Efficient Second Generation ML Inference processor

4- Private Edge Cloud for Super Storage : 32 SerDes lanes, PCIe Gen4 supporting data rates up to 16 Gbps. Four PCIe Gen 4.0 8-lane controllers .Using this PCIe Gen4 connectivity , the motherboard will have 2 connectors for 2 X 8TB NVMe PCIe Gen 4 M.2 2280 Internal High Performance Solid State Drive (SSD). Till availability of SSD drives in this configuration, lower capacity SSD up to 2TB can be used.

5- Support for 96 inch UHD (Ultra High Definition) Monitors: Remote surgeries will need high definition real time videos to be projected on large UHD monitors through 4X HDMI 2.0 ports at 120 Hz

6- 5G – 5G connectivity will be achieved by using external 3^rd party modems like Quectel RM500Q-GL. It is a 5G module optimized for Edge applications.

7- Security: Trust Zone technology within Cortex A-78 is used to run trusted boot and trusted OS to create a Trusted Execution Environment (TEE). Typical use cases include the protection of Authentication, cryptography, payment, key material and Digital Rights management (DRM) .Edge cloud security is in-built into SSDs through. HW:AES 128/256 bit (XTS, CTR, CBC, ECB mode), SHA 160/256/512, RSA 2048 SW:TCG Support & Opal 2.0, Pyrite, Sanitize and Crypto Erase

8- Other Interfaces : 2 X MIPI CSI-2,SD/eMMC , 2X USB 3.0 controllers with integrated PHY, 4 X I2C , 4 X UART , 24X GPIO,3 X HDMI , Wi-Fi , BT-5.0 ( Medical Device direct connectivity)

9- Operating System : Linux

Block Diagram:

Remote Surgery Use Model envisaged to be used with SEMC:

5G makes this possible by cutting latency to almost instantaneous, 2 ms between devices, allowing surgeons to conduct procedures from thousands of miles away as if they were right next to the patient. All medical records of the patient are stored in the local Edge Cloud of the Surgeon in 16TB storage provided by SSDs. They are retrieved instantly by PCIe Gen 4 connectivity of SSDs and displayed on large 98 inch screen. Real time data of all the medical procedure monitoring devices at the patients end are displayed on UHD 98 inch monitors of the surgeon.

In future, Mixed Reality (MR) and Robotics surgery will also be feasible through the same platform with entire procedure to be done by the remote surgeon only.

During medical procedure, the high resolution medical images obtained can be fed to the already trained Machine Learning models stored in Edge Cloud and inference drawn quickly using powerfulETHOS .

If you wish to download a copy of this white paper, click here