Many architects believe that major improvements in cost-energyperformance must now come from domain-specific hardware. This paper evaluates a custom ASIC-called a Tensor Processing Unit (TPU)-deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile responsetime requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X -30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X -80X higher. Moreover, using the GPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.

Many architects believe that major improvements in cost-energyperformance must now come from domain-specific hardware. This paper evaluates a custom ASIC-called a Tensor Processing Unit (TPU)-deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile responsetime requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X-30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X-80X higher. Moreover, using the GPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.

Background: mTORC1 plays an important role in the regulation of TOP mRNA translation. Results: LARP1 is a target of mTORC1 that associates with TOP mRNAs via their 5ЈTOP motif to repress their translation. Conclusion: LARP1 represses TOP mRNA translation downstream of mTORC1. Significance: We elucidate an important novel signaling pathway downstream of mTORC1 that controls the production of ribosomes and translation factors in eukaryotic cells.

The COVID-19 pandemic has challenged front-line clinical decision-making, leading to numerous published prognostic tools. However, few models have been prospectively validated and none report implementation in practice. Here, we use 3345 retrospective and 474 prospective hospitalizations to develop and validate a parsimonious model to identify patients with favorable outcomes within 96 h of a prediction, based on real-time lab values, vital signs, and oxygen support variables. In retrospective and prospective validation, the model achieves high average precision (88.6% 95% CI: [88.4–88.7] and 90.8% [90.8–90.8]) and discrimination (95.1% [95.1–95.2] and 86.8% [86.8–86.9]) respectively. We implemented and integrated the model into the EHR, achieving a positive predictive value of 93.3% with 41% sensitivity. Preliminary results suggest clinicians are adopting these scores into their clinical workflows.

Mutations in the Retinoblastoma (RB) tumour suppressor pathway are a hallmark of cancer and a prevalent feature of lung adenocarcinoma 1 , 2 , 3 . Despite being the first tumour suppressor to be identified, the molecular and cellular basis underlying selection for persistent RB loss in cancer remains unclear 4 – 6 . Methods that reactivate the RB pathway using inhibitors of cyclin-dependent kinases CDK4 and CDK6 are effective in some cancer types and currently under evaluation in lung adenocarcinoma 7 – 9 . Whether RB pathway reactivation will have therapeutic effects and if targeting CDK4/6 is sufficient to reactivate RB pathway activity in lung cancer is unknown. Here, we model RB loss during lung adenocarcinoma progression and pathway reactivation in established oncogenic KRAS-driven tumours in the mouse. We show that RB loss enables cancer cells to bypass two distinct barriers during tumour progression. First, RB loss abrogates the requirement for MAPK signal amplification during malignant progression. We identify CDK2-dependent phosphorylation of RB as an effector of MAPK signalling and critical mediator of resistance to CDK4/6 inhibition. Second, RB inactivation deregulates expression of cell state-determining factors, facilitates lineage infidelity, and accelerates the acquisition of metastatic competency. In contrast, reactivation of RB reprograms advanced tumours toward a less metastatic cell state, but is nevertheless unable to halt cancer cell proliferation and tumour growth due to adaptive rewiring of MAPK pathway signalling, which restores a CDK-dependent suppression of RB. Our study demonstrates the power of reversible gene perturbation approaches to identify molecular mechanisms of tumour progression, causal relationships between genes and the tumour suppressive programs they control, and critical determinants of successful therapy.