Improving experiment precision with machine learning

The challenge of noise in experiments

Experimentation is a central part of data-driven product development, yet in practice the results from experiments may be too imprecise to be of much help in improving decision-making. One possible response is to reduce statistical noise by simply running larger experiments. However, this is not always desirable, or even feasible. This raises the question of how we can make better use of the data we have and get sharper, more precise experimental estimates without having to enroll more people in the test.

In a collaboration between Meta’s Core Data Science and Experimentation Platform teams, we developed a new methodology for making progress on this problem, which both has formal statistical guarantees and is scalable enough to implement in practice. The work, described in detail in our NeurIPS paper, allows for general machine learning (ML) techniques to be used in conjunction with experimental data to substantially increase the precision of experimental estimates, relative to other existing methods.

How it works

Our algorithm, MLRATE (machine learning regression-adjusted average treatment effects), involves two main steps. First, we train a model predicting the experimental outcome of interest, given a set of pre-experimental covariates. Second, we use these predictions as a control variable in a linear regression. The coefficient on the treatment effect estimator is our variance-reduced average treatment effect estimator.

In the first step, we use sample splitting, so that predicted outcomes for each observation are generated by a model trained on data not including that observation. This allows us to use a broad class of ML methods in the first step, and gives us the flexibility to choose whichever model does the best job at predicting outcomes. The ML method in question may even be asymptotically biased, and not even converge to the truth in large samples, without affecting the validity of our estimator.

In the second step, we treat the predictions from the first step as a control variable in a linear regression. This form of linear regression adjustment is relatively common in the analysis of experimental data (e.g., Lin [2013], Deng et al. [2013]). The contribution of our paper is to show how this methodology can be generalized to accommodate control variables, which are themselves the output of a potentially complex ML algorithm.

Empirical results

To quantify the variance reduction gains one might expect from MLRATE in practice, we implemented it in A/A tests for a set of 48 outcome metrics commonly monitored in Meta experiments. Using either gradient-boosted decision trees or elastic net regression for the ML prediction step, we find that MLRATE has, on average, over 70 percent lower variance than the simple difference-in-means estimator for these metrics, and about 19 percent lower variance than the common univariate procedure, which adjusts only for pre-experiment values of the outcome.

Alternatively, to achieve the same precision as MLRATE, the conventional difference-in-means estimator would require sample sizes over five times as large on average across metrics, and the univariate linear regression procedure would require sample sizes about 1.6 times as large. The figure above displays the metric-level distribution of confidence interval widths relative to the univariate adjustment case. There is substantial heterogeneity in performance across metrics: For some, ML regression adjustment delivers only modest gains relative to univariate adjustment; for others, it drastically shrinks confidence intervals. This is natural given the variety of metrics in the analysis: Some, especially binary or discrete outcomes, may benefit more from more sophisticated predictive modeling, whereas for others, simple linear models may perform well.

Why MLRATE matters in practice

A couple of features of this methodology make it relatively straightforward to implement in practice. First, the formulas for calculating treatment effect estimators and confidence intervals are no more complex than they are in case of conventional linear regression adjustment. Second, most common off-the-shelf ML methods can be used for the prediction stage, as long as the covariates used are pre-experiment. Finally, MLRATE does not require an investment in ML modeling for each individual experiment to work well. Once predictive models have been trained for an outcome of interest, they can be used for many experiments, so the cost of the ML training does not scale with the number of experiments.

If you’re dealing with the problem of excessive noise in your experiments and you can construct good predictors of the outcome of interest, MLRATE may be a helpful new tool for variance reduction. Depending on the metric, it may even be the difference between experimentation being feasible or not. For more details, check out our NeurIPS paper.

The post Improving experiment precision with machine learning appeared first on Facebook Research.

Read More

More Efficient In-Context Learning with GLaM

Posted by Andrew M Dai and Nan Du, Research Scientists, Google Research, Brain Team

Large language models (e.g., GPT-3) have many significant capabilities, such as performing few-shot learning across a wide array of tasks, including reading comprehension and question answering with very few or no training examples. While these models can perform better by simply using more parameters, training and serving these large models can be very computationally intensive. Is it possible to train and use these models more efficiently?

In pursuit of that question, today we introduce the Generalist Language Model (GLaM), a trillion weight model that can be trained and served efficiently (in terms of computation and energy use) thanks to sparsity, and achieves competitive performance on multiple few-shot learning tasks. GLaM’s performance compares favorably to a dense language model, GPT-3 (175B) with significantly improved learning efficiency across 29 public NLP benchmarks in seven categories, spanning language completion, open-domain question answering, and natural language inference tasks.

Dataset
To build GLaM, we began by building a high-quality 1.6 trillion token dataset containing language usage representative of a wide range of downstream use-cases for the model. Web pages constitute the vast quantity of data in this unlabelled corpus, but their quality ranges from professional writing to low-quality comment and forum pages. We then developed a text quality filter that was trained on a collection of text from Wikipedia and books (both of which are generally higher quality sources) to determine the quality of the content for a webpage. Finally, we applied this filter to generate the final subset of webpages and combined this with books and Wikipedia to create the final training dataset.

Model and Architecture
GLaM is a mixture of experts (MoE) model, a type of model that can be thought of as having different submodels (or experts) that are each specialized for different inputs. The experts in each layer are controlled by a gating network that activates experts based on the input data. For each token (generally a word or part of a word), the gating network selects the two most appropriate experts to process the data. The full version of GLaM has 1.2T total parameters across 64 experts per MoE layer with 32 MoE layers in total, but only activates a subnetwork of 97B (8% of 1.2T) parameters per token prediction during inference.

The architecture of GLaM where each input token is dynamically routed to two selected expert networks out of 64 for prediction.

Similar to the GShard MoE Transformer, we replace the single feedforward network (the simplest layer of an artificial neural network, “Feedforward or FFN” in the blue boxes) of every other transformer layer with a MoE layer. This MoE layer has multiple experts, each a feedforward network with identical architecture but different weight parameters. Even though this MoE layer has many more parameters, the experts are sparsely activated, meaning that for a given input token, only two experts are used, giving the model more capacity while limiting computation. During training, each MoE layer’s gating network is trained to use its input to activate the best two experts for each token, which are then used for inference. For a MoE layer of E experts, this essentially provides a collection of E×(E-1) different feedforward network combinations (instead of one as in the classic Transformer architecture), leading to more computational flexibility.

The final learned representation of a token will be the weighted combination of the outputs from the two experts. This allows different experts to activate on different types of inputs. To enable scaling to larger models, each expert within the GLaM architecture can span multiple computational devices. We use the GSPMD compiler backend to solve the challenges in scaling the experts and train several variants (based on expert size and number of experts) of this architecture to understand the scaling effects of sparsely activated language models.

Evaluation
We use a zero-shot and one-shot setting where the tasks are never seen during training. The benchmarks for evaluation include (1) cloze and completion tasks [1,2,3]; (2) Open-domain question answering [4,5,6]; (3) Winograd-style tasks [7,8]; (4) commonsense reasoning [9,10,11]; (5) in-context reading comprehension [12,13,14,15,16]; (6) the SuperGLUE tasks; and (7) natural language inference [17]. In total, there are eight natural language generation tasks (NLG) where the generated phrases are evaluated against the ground truth targets via Exact Match (EM) accuracy and F1 measure, and 21 language understanding tasks (NLU) where the prediction from several options is chosen via conditional log-likelihood. Some tasks have variants and SuperGLUE consists of multiple tasks. Both EM accuracy and F1 are scaled from 0 to 100 across all our results and averaged for the NLG score below. The NLU score is an average of accuracy and F1 scores.

Results
GLaM reduces to a basic dense Transformer-based language model architecture when each MoE layer only has one expert. In all experiments, we adopt the notation of (base dense model size) / (number of experts per MoE layer) to describe the GLaM model. For example, 1B/64E represents the architecture of a 1B parameter dense model with every other layer replaced by a 64 expert MoE layer. In the following sections, we explore GLaM’s performance and scaling properties, including baseline dense models trained on the same datasets. Compared with the recently announced Megatron-Turing model, GLaM is on-par on the seven respective tasks if using a 5% margin, while using 5x less computation during inference.

Below, we show the 1.2T-parameter sparsely activated model (GLaM) achieved higher results on average and on more tasks than the 175B-parameter dense GPT-3 model while using less computation during inference.

Average score for GLaM and GPT-3 on NLG (left) and NLU (right) tasks (higher is better).

Below we show a summary of the performance on 29 benchmarks compared to the dense model (GPT-3, 175B). GLaM exceeds or is on-par with the performance of the dense model on almost 80% of zero-shot tasks and almost 90% of one-shot tasks.

Evaluation Higher (>+5%) On-par (within 5%) Lower (<-5%)
Zero-shot 13 11 5
One-shot 14 10 5

Moreover, while the full version of GLaM has 1.2T total parameters, it only activates a subnetwork of 97B parameters (8% of 1.2T) per token during inference.

GLaM (64B/64E) GPT-3 (175B)
Total Parameters 1.162T 0.175T
Activated Parameters 0.097T 0.175T

Scaling Behavior
GLaM has two ways to scale: 1) scale the number of experts per layer, where each expert is hosted within one computation device, or 2) scale the size of each expert to go beyond the limit of a single device. To evaluate the scaling properties, we compare the respective dense model (FFN layers instead of MoE layers) of similar FLOPS per token at inference time.

Average zero-shot and one-shot performance by increasing the size of each expert. The FLOPS per token prediction at inference time increases as the expert size grows.

As shown above, performance across tasks scales with the size of the experts. GLaM sparsely activated models also perform better than dense models for similar FLOPs during inference for generation tasks. For understanding tasks, we observed that they perform similarly at smaller scales, but sparsely activated models outperform at larger scales.

Data Efficiency
Training large language models is computationally intensive, so efficiency improvements are useful to reduce energy consumption.

Below we show the computation costs for the full version of GLaM.

Computation cost in GFLOPS both for inference, per token (left) and for training (right).

These compute costs show that GLaM uses more computation during training since it trains on more tokens, but uses significantly less computation during inference. We show comparisons using different numbers of tokens to train below.

We also evaluated the learning curves of our models compared to the dense baseline.

Average zero-shot and one-shot performance of sparsely-activated and dense models on eight generative tasks as more tokens are processed in training.
Average zero-shot and one-shot performance of sparsely-activated and dense models on 21 understanding tasks as more tokens are processed in training.

The results above show that sparsely activated models need to train with significantly less data than dense models to reach similar zero-shot and one-shot performance, and if the same amount of data is used, sparsely activated models perform significantly better.

Finally, we assessed the energy efficiency of GLaM.

Comparison of power consumption during training.

While GLaM uses more computation during training, thanks to the more efficient software implementation powered by GSPMD and the advantage of TPUv4, it uses less power to train than other models.

Conclusions
Our large-scale sparsely activated language model, GLaM, achieves competitive results on zero-shot and one-shot learning and is a more efficient model than prior monolithic dense counterparts. We also show quantitatively that a high-quality dataset is essential for large language models. We hope that our work will spark more research into compute-efficient language models.

Acknowledgements
We wish to thank Claire Cui, Zhifeng Chen, Yonghui Wu, Quoc Le, Macduff Hughes, Fernando Pereira, Zoubin Ghahramani‎ and Jeff Dean for their support and invaluable input. Special thanks to our collaborators: Yanping Huang, Simon Tong, Yanqi Zhou, Yuanzhong Xu, Dmitry Lepikhin, Orhan Firat, Maxim Krikun, Tao Wang, Noam Shazeer, Barret Zoph, Liam Fedus, Maarten Bosma, Kun Zhang, Emma Wang, David Patterson, Zongwei Zhou, Naveen Kumar, Adams Yu, Laurent Shafey, Jonathan Shen, Ben Lee, Anmol Gulati, David So, Marie Pellat, Kevin Robinson, Kathy Meier-Hellstern‎, Aakanksha Chowdhery, Sharan Narang, Erica Moreira and Eric Ni for helpful discussions and inspirations; and the larger Google Research team. We would also like to thank Tom Small for the animated figure used in this post.

Read More

Announcing the winners of the City-Scale 3D Map Making with Mapillary Metropolis request for proposals

In July 2021, Meta launched the Benchmarking City-Scale 3D Map Making with Mapillary Metropolis request for proposals (RFP). Today, we’re announcing the winners of this award.

VIEW RFP

Earlier this year, we introduced a novel, city-scale data set called Mapillary Metropolis, which was designed with the goal of creating a completely novel and complex benchmarking paradigm for training and testing computer vision algorithms in the context of semantic 3D map making.

For this RFP, we sought research proposals that leveraged Mapillary Metropolis to improve basic computer vision algorithms that use one or preferably multiple data modalities from our data set for improving semantic 3D building. We were particularly interested in the following areas:

  • City-scale 3D modeling from heterogeneous data sources
  • ML for object recognition, tracking, and dense labeling
  • Image-based matching, relocalization, and retrieval

The RFP attracted 29 proposals from 27 universities and institutions around the world. Thank you to everyone who took the time to submit a proposal, and congratulations to the winners.

Research award winners

Principal investigators are listed first unless otherwise noted.

Factorized, object-centric implicit representations for city-scale scenes
Jiajun Wu, Hong-Xing (Koven) Yu (Stanford University)

Multi-modal 6DOF visual relocalization in Mapillary Metropolis
Torsten Sattler, Zuzana Kukelova (Czech Technical University in Prague)

Neural feature fields for photorealistic scene synthesis
Andreas Geiger (University of Tübingen, Germany)

The post Announcing the winners of the City-Scale 3D Map Making with Mapillary Metropolis request for proposals appeared first on Facebook Research.

Read More

Machine-learning system flags remedies that might do more harm than good

Sepsis claims the lives of nearly 270,000 people in the U.S. each year. The unpredictable medical condition can progress rapidly, leading to a swift drop in blood pressure, tissue damage, multiple organ failure, and death.

Prompt interventions by medical professionals save lives, but some sepsis treatments can also contribute to a patient’s deterioration, so choosing the optimal therapy can be a difficult task. For instance, in the early hours of severe sepsis, administering too much fluid intravenously can increase a patient’s risk of death.

To help clinicians avoid remedies that may potentially contribute to a patient’s death, researchers at MIT and elsewhere have developed a machine-learning model that could be used to identify treatments that pose a higher risk than other options. Their model can also warn doctors when a septic patient is approaching a medical dead end — the point when the patient will most likely die no matter what treatment is used — so that they can intervene before it is too late.

When applied to a dataset of sepsis patients in a hospital intensive care unit, the researchers’ model indicated that about 12 percent of treatments given to patients who died were detrimental. The study also reveals that about 3 percent of patients who did not survive entered a medical dead end up to 48 hours before they died.

“We see that our model is almost eight hours ahead of a doctor’s recognition of a patient’s deterioration. This is powerful because in these really sensitive situations, every minute counts, and being aware of how the patient is evolving, and the risk of administering certain treatment at any given time, is really important,” says Taylor Killian, a graduate student in the Healthy ML group of the Computer Science and Artificial Intelligence Laboratory (CSAIL).

Joining Killian on the paper are his advisor, Assistant Professor Marzyeh Ghassemi, head of the Healthy ML group and senior author; lead author Mehdi Fatemi, a senior researcher at Microsoft Research; and Jayakumar Subramanian, a senior research scientist at Adobe India. The research is being presented at this week’s Conference on Neural Information Processing Systems.  

A dearth of data

This research project was spurred by a 2019 paper Fatemi wrote that explored the use of reinforcement learning in situations where it is too dangerous to explore arbitrary actions, which makes it difficult to generate enough data to effectively train algorithms. These situations, where more data cannot be proactively collected, are known as “offline” settings.

In reinforcement learning, the algorithm is trained through trial and error and learns to take actions that maximize its accumulation of reward. But in a health care setting, it is nearly impossible to generate enough data for these models to learn the optimal treatment, since it isn’t ethical to experiment with possible treatment strategies.

So, the researchers flipped reinforcement learning on its head. They used the limited data from a hospital ICU to train a reinforcement learning model to identify treatments to avoid, with the goal of keeping a patient from entering a medical dead end.

Learning what to avoid is a more statistically efficient approach that requires fewer data, Killian explains.

“When we think of dead ends in driving a car, we might think that is the end of the road, but you could probably classify every foot along that road toward the dead end as a dead end. As soon as you turn away from another route, you are in a dead end. So, that is the way we define a medical dead end: Once you’ve gone on a path where whatever decision you make, the patient will progress toward death,” Killian says.

“One core idea here is to decrease the probability of selecting each treatment in proportion to its chance of forcing the patient to enter a medical dead-end — a property that is called treatment security. This is a hard problem to solve as the data do not directly give us such an insight. Our theoretical results allowed us to recast this core idea as a reinforcement learning problem,” Fatemi says.

To develop their approach, called Dead-end Discovery (DeD), they created two copies of a neural network. The first neural network focuses only on negative outcomes — when a patient died — and the second network only focuses on positive outcomes — when a patient survived. Using two neural networks separately enabled the researchers to detect a risky treatment in one and then confirm it using the other.

They fed each neural network patient health statistics and a proposed treatment. The networks output an estimated value of that treatment and also evaluate the probability the patient will enter a medical dead end. The researchers compared those estimates to set thresholds to see if the situation raises any flags.

A yellow flag means that a patient is entering an area of concern while a red flag identifies a situation where it is very likely the patient will not recover.

Treatment matters

The researchers tested their model using a dataset of patients presumed to be septic from the Beth Israel Deaconess Medical Center intensive care unit. This dataset contains about 19,300 admissions with observations drawn from a 72-hour period centered around when the patients first manifest symptoms of sepsis. Their results confirmed that some patients in the dataset encountered medical dead ends.

The researchers also found that 20 to 40 percent of patients who did not survive raised at least one yellow flag prior to their death, and many raised that flag at least 48 hours before they died. The results also showed that, when comparing the trends of patients who survived versus patients who died, once a patient raises their first flag, there is a very sharp deviation in the value of administered treatments. The window of time around the first flag is a critical point when making treatment decisions.

“This helped us confirm that treatment matters and the treatment deviates in terms of how patients survive and how patients do not. We found that upward of 11 percent of suboptimal treatments could have potentially been avoided because there were better alternatives available to doctors at those times. This is a pretty substantial number, when you consider the worldwide volume of patients who have been septic in the hospital at any given time,” Killian says.

Ghassemi is also quick to point out that the model is intended to assist doctors, not replace them.

“Human clinicians are who we want making decisions about care, and advice about what treatment to avoid isn’t going to change that,” she says. “We can recognize risks and add relevant guardrails based on the outcomes of 19,000 patient treatments — that’s equivalent to a single caregiver seeing more than 50 septic patient outcomes every day for an entire year.”

Moving forward, the researchers also want to estimate causal relationships between treatment decisions and the evolution of patient health. They plan to continue enhancing the model so it can create uncertainty estimates around treatment values that would help doctors make more informed decisions. Another way to provide further validation of the model would be to apply it to data from other hospitals, which they hope to do in the future.

This research was supported in part by Microsoft Research, a Canadian Institute for Advanced Research Azrieli Global Scholar Chair, a Canada Research Council Chair, and a Natural Sciences and Engineering Research Council of Canada Discovery Grant.

Read More

Forging New Pathways: Boys & Girls Clubs Teens Take AI From Idea to Application

Building AI projects to aid Alzheimer’s patients and monitor pedestrian safety might not be the typical way teens spend their summer.

But that’s what a dozen teens with the Boys & Girls Clubs of Hudson County, in densely packed northeastern New Jersey, did as part of the AI Pathways Institute program.

They spent three weeks brainstorming, coding and traversing spinning robots while using NVIDIA Jetson Nano 2GB developer kits and Jetbot robotics toolkits to create projects that tapped into a multitude of real-world issues. The aim was “to leave our environment and society better than we found it,” said Gail Grant, teen tech coordinator at the nonprofit.

The AI Pathways Institute program in Hudson County is an outgrowth of a successful 2019 pilot program that NVIDIA and the Boys & Girls Clubs of Western Pennsylvania partnered on. Its goals are to introduce high school students to AI and machine learning through a three-week summer camp and provide them with hands-on experience with relevant projects.

NVIDIA and the Western Pennsylvania Clubs announced this spring a new three-year partnership to scale the program to more students through the development of the AI Pathways Toolkit. The toolkit strives to make it easy for other clubs and organizations to deliver hands-on AI and robotics education to youth.

The Hudson County club — which has worked for nearly 130 years to enrich the lives of young people from disadvantaged circumstances — is among the first to adopt it.

Participating students received a cash stipend, new laptop and certificate of achievement from Carnegie Mellon University, which had previously worked with the Western Pennsylvania club. Students who scored well on a post-program exam qualified themselves for future internships.

Emboldening Students in AI

Grant said the program’s goal was to provide students with a broader sense of what AI truly entails. Moreover, she hoped to embolden students to develop and pursue newfound interests in AI and technology by providing speaker events where STEM and AI professionals highlighted their career pathways.

Sekou Sy, a 16-year-old in the AI Pathways program, was surprised to learn about the breadth of robotics — and just how much could be done with AI.

“AI is used in so many objects and ideas that I didn’t think of before,” Sy said. “It’s not just in robots, but other areas like healthcare and the environment, all of which will expand even more in the future.”

Another student, Moureau Tillman (pictured above), was already familiar with AI and the Python programming language, but AI Pathways’ focus on hands-on learning allowed her to dive deeper into AI’s real-world impact.

“Seeing as I already worked with AI, I thought the course was going to be a re-learning experience,” Tillman said. “But I learned a lot of new things through working with robots and coming up with a project that would help people other than myself.”

AI Projects in Action

The program culminated in a presentation in which students explained how they used Python to program NVIDIA Jetson-based projects that would respond to prominent, practical issues.

One project, titled “Forget-Me-Not,” employed AI to help elders and Alzheimer’s patients by alerting them about tasks and appointments, as well as providing medicine or prescription reminders.

“When brainstorming for this project, I thought about what I struggled with myself,” said Sy, who worked on the project with two teammates. “My memory is not so good, and I lose things a lot. We came up with this app where you can put in your data, and it gets to know your daily routines. Then, it reminds you about just about everything you might need.”

Other groups created proposals for projects like “Ecobot,” a robot that roams beaches and discards trash; “Safety First,” which uses AI to monitor streets and call the police when a pedestrian is in danger; and “Fresh,” a device to be placed in users’ cars and monitor outdoor air pollution.

Although not all were able to see their projects to completion, students witnessed how AI could be applied to a variety of problems. Some were even inspired to pursue tech as a career.

“These projects made me think about real-life situations that people are dealing with,” Sy said. “I loved taking it one step further by incorporating the AI aspect, as well as seeing the impact that AI can have in solving these problems. In fact, AI is a field I’m really excited to pursue in the future.”

The Hudson County club plans to reproduce the program early next year. Moving forward, Grant hopes that it’ll be a recurring one that’s implemented throughout the year.

To learn more about entry-level education on AI and robotics, watch the on-demand NVIDIA GTC sessions “Begin Your AI Journey With NVIDIA Jetson Nano” (A31723) and “Getting Started With the Edge AI and Robotics Teaching Kit” (A31535).

The post Forging New Pathways: Boys & Girls Clubs Teens Take AI From Idea to Application appeared first on The Official NVIDIA Blog.

Read More

Latest GeForce NOW Upgrade Rolling Out With Ubisoft Connect Account Linking and Improved PC Gaming on Mac

Get into the game quicker with the latest GeForce NOW update starting to roll out this GFN Thursday.

Learn more about our latest app update — featuring Ubisoft Connect account linking for faster game launches — now rolling out to members, and the six new games joining the GeForce NOW library.

The update also improves the streaming experience on Mac. So, this GFN Thursday also takes a look at how GeForce NOW transforms Macs into powerful PC gaming rigs.

Streamline Your Sign-In

The newest GeForce NOW app update is rolling out to members on PC and Mac, beginning this week. Version 2.0.36 includes a new feature that links NVIDIA and Ubisoft accounts to enable faster Ubisoft Connect game launches by automatically logging into a user’s account. Skip the sign-in process and stream your favorite Ubisoft games this week.

The update also includes a fix for streaming at the correct aspect ratio on the Apple MacBook Pro M1 Max, as well as improvements to the countdown timer when viewed on the in-game overlay.

PC Gaming on a Mac? Here’s How It Works

It used to be a difficult choice: do you want to be a Mac user or a PC gamer? With GeForce NOW, you can have your Mac and PC game, too.

GeForce NOW transforms nearly any Mac into a high-end gaming rig, thanks to the power of the cloud. NVIDIA data centers do the heavy lifting, rendering games at full quality and streaming them down to Macbook Pro, Macbook Air, iMac and iOS. Get all the benefits of PC gaming, without leaving the Apple ecosystem.

For the full RTX 3080 experience, connect your Macbook or other laptop via Displayport to a gaming monitor. It’s a beautiful thing.

On GeForce NOW, you play the real PC versions of your games, without having to worry if something has been ported to Mac. Software compatibility for the new M1 Mac isn’t a problem either, since the native PC version of games streams straight from the cloud.

GeForce NOW also handles game saves for supported games, so members can play on their Macs, as well as any other supported device, without losing progress.

That means the next time your squad readies up in Apex Legends, you can join the fray from your iMac. Jealous of your friends building their ultimate Viking community in Valheim? With GeForce NOW, you can join them without leaving your Mac. Build your gaming library with weekly free games from the Epic Games Store, with offers like Dead By Daylight.

Plus, GeForce NOW RTX 3080 members can now play at native resolution on their M1 Macbook Air or Macbook Pro, at glorious 1600p. Stream with even longer sessions lengths — up to eight hours. And with RTX ON for both RTX 3080 and Priority members, experience games like Cyberpunk 2077 and Control with real-time ray tracing, without upgrading to a PC.

Playing PC games with GeForce NOW on a Mac is like having your cake and eating it, too.

Yes, Your Mac Can Run Crysis Now

Can it run Crysis Remastered? Yep. Get the game for free with a six-month Priority sign-up or GeForce NOW RTX 3080 order.

Ready for the ultimate battle on your Mac? For a limited time, get a copy of Crysis Remastered free with select GeForce NOW memberships. Purchase a six-month Priority membership, or the new GeForce NOW RTX 3080 membership, and get a free redeemable code for Crysis Remastered on the Epic Games Store. Terms and conditions apply.

Do Pass Go. Do Collect New Games.

Rediscover the Monopoly game you love, in a way you’ve never seen before, in Monopoly Madness.

GFN Thursday always means new games coming to the cloud. Six titles are being added to the GeForce NOW library this week, including two day-and-date releases:

We make every effort to launch games on GeForce NOW as close to their release as possible, but, in some instances, games may not be available immediately.

Also, we had hoped to add both Syberia: The World Before in December. However, these games have shifted their release dates to next year and will be coming to the cloud in the future.

Finally, in case you missed it – Fortnite flipped. Explore new locations, take on enemies with new weapons, and discover what’s new on the Island in Fortnite Chapter 3 – streaming now!

For the members who have experienced the magic of PC gaming on a Mac, we’ve got a question for you. Tell us on Twitter or down below in the comments.

The post Latest GeForce NOW Upgrade Rolling Out With Ubisoft Connect Account Linking and Improved PC Gaming on Mac appeared first on The Official NVIDIA Blog.

Read More

A tool to speed development of new solar cells

In the ongoing race to develop ever-better materials and configurations for solar cells, there are many variables that can be adjusted to try to improve performance, including material type, thickness, and geometric arrangement. Developing new solar cells has generally been a tedious process of making small changes to one of these parameters at a time. While computational simulators have made it possible to evaluate such changes without having to actually build each new variation for testing, the process remains slow.

Now, researchers at MIT and Google Brain have developed a system that makes it possible not just to evaluate one proposed design at a time, but to provide information about which changes will provide the desired improvements. This could greatly increase the rate for the discovery of new, improved configurations.

The new system, called a differentiable solar cell simulator, is described in a paper published today in the journal Computer Physics Communications, written by MIT junior Sean Mann, research scientist Giuseppe Romano of MIT’s Institute for Soldier Nanotechnologies, and four others at MIT and at Google Brain.

Traditional solar cell simulators, Romano explains, take the details of a solar cell configuration and produce as their output a predicted efficiency — that is, what percentage of the energy of incoming sunlight actually gets converted to an electric current. But this new simulator both predicts the efficiency and shows how much that output is affected by any one of the input parameters. “It tells you directly what happens to the efficiency if we make this layer a little bit thicker, or what happens to the efficiency if we for example change the property of the material,” he says.

In short, he says, “we didn’t discover a new device, but we developed a tool that will enable others to discover more quickly other higher performance devices.” Using this system, “we are decreasing the number of times that we need to run a simulator to give quicker access to a wider space of optimized structures.” In addition, he says, “our tool can identify a unique set of material parameters that has been hidden so far because it’s very complex to run those simulations.”

While traditional approaches use essentially a random search of possible variations, Mann says, with his tool “we can follow a trajectory of change because the simulator tells you what direction you want to be changing your device. That makes the process much faster because instead of exploring the entire space of opportunities, you can just follow a single path” that leads directly to improved performance.

Since advanced solar cells often are composed of multiple layers interlaced with conductive materials to carry electric charge from one to the other, this computational tool reveals how changing the relative thicknesses of these different layers will affect the device’s output. “This is very important because the thickness is critical. There is a strong interplay between light propagation and the thickness of each layer and the absorption of each layer,” Mann explains.

Other variables that can be evaluated include the amount of doping (the introduction of atoms of another element) that each layer receives, or the dielectric constant of insulating layers, or the bandgap, a measure of the energy levels of photons of light that can be captured by different materials used in the layers.

This simulator is now available as an open-source tool that can be used immediately to help guide research in this field, Romano says. “It is ready, and can be taken up by industry experts.” To make use of it, researchers would couple this device’s computations with an optimization algorithm, or even a machine learning system, to rapidly assess a wide variety of possible changes and home in quickly on the most promising alternatives.

At this point, the simulator is based on just a one-dimensional version of the solar cell, so the next step will be to expand its capabilities to include two- and three-dimensional configurations. But even this 1D version “can cover the majority of cells that are currently under production,” Romano says. Certain variations, such as so-called tandem cells using different materials, cannot yet be simulated directly by this tool, but “there are ways to approximate a tandem solar cell by simulating each of the individual cells,” Mann says.

The simulator is “end-to-end,” Romano says, meaning it computes the sensitivity of the efficiency, also taking into account light absorption. He adds: “An appealing future direction is composing our simulator with advanced existing differentiable light-propagation simulators, to achieve enhanced accuracy.”

Moving forward, Romano says, because this is an open-source code, “that means that once it’s up there, the community can contribute to it. And that’s why we are really excited.” Although this research group is “just a handful of people,” he says, now anyone working in the field can make their own enhancements and improvements to the code and introduce new capabilities.

“Differentiable physics is going to provide new capabilities for the simulations of engineered systems,” says Venkat Viswanathan, an associate professor of mechanical engineering at Carnegie Mellon University, who was not associated with this work. “The  differentiable solar cell simulator is an incredible example of differentiable physics, that can now provide new capabilities to optimize solar cell device performance,” he says, calling the study “an exciting step forward.”

In addition to Mann and Romano, the team included Eric Fadel and Steven Johnson at MIT, and Samuel Schoenholz and Ekin Cubuk at Google Brain. The work was supported in part by Eni S.p.A. and the MIT Energy Initiative, and the MIT Quest for Intelligence.

Read More

Simulating matter on the quantum scale with AI

Solving some of the major challenges of the 21st Century, such as producing clean electricity or developing high temperature superconductors, will require us to design new materials with specific properties. To do this on a computer requires the simulation of electrons, the subatomic particles that govern how atoms bond to form molecules and are also responsible for the flow of electricity in solids.Read More