91视频APP

Make new breakthroughs in drug discovery with IPU systems. Take cellular and molecular biology techniques to the next level to validate novel targets for drug discovery. Explore new AI approaches to drug combination modelling, reducing costs and speeding time to identify new treatments.

LabGenius develops next-generation antibody therapeutics for cancer and inflammatory diseases, using a combination of AI, synthetic biology and laboratory automation.  IPU systems halved the compute time needed to run crucial AI model training.

"Previously we used GPUs and it took us about a month to have a functioning model of all the proteins that are out there. With 91视频APP, we reduced the turnaround time to about two weeks, so we can experiment much more rapidly, and we can see the results quicker.鈥�

Dr Katya Putintseva, Machine Learning Advisor to LabGenius

Joint research from the University of Massachusetts Amherst & Facebook shows IPUs deliver a 7.5x speed-up for an Approximate Bayesian Computation (ABC) epidemiology model used in Covid-19 modelling.

鈥淭his hardware acceleration can give real tangible results in obtaining better quality results in much less time and could also be translated to many other scientific models which use simulation-based inference. We believe that this work would pave the way for more complex models and better quality inference in less time.鈥�

Sourabh Kulkarni, University of Massachusetts Amherst

Accelerate genome analysis and predict protein structure to understand disease better and bring tailored therapeutics to market faster, bringing the dream of precision medicine that much closer.

鈥淲ith IPUs we鈥檝e seen impressive throughput in some of our key areas of research. We are excited to be testing Mk2 architecture and continue exploring the benefits of this innovative chip architecture.鈥�

Chris Seymour, Director, Advanced Platform Development at Oxford Nanopore 

Advanced computer vision models give medical staff deeper understanding of diagnostic imaging to make faster decisions for diagnosis and treatment and ongoing patient care. 

Microsoft explored advanced computer vision model, EfficientNet, on the IPU for medical imaging. For pneumonia and Covid-19 X-ray imaging, researchers trained EfficientNet in just 40 minutes on IPU systems with higher accuracy than traditional vision models, compared to 3 hours on GPUs.