This year marks Google I/O’s 15th anniversary. To celebrate, we’re answering the question “What does I/O mean?”
The Google Pay button is your customer’s entry point to a swift and familiar checkout with Google Pay. Here are some updates we are bringing to the button to make it easier for you to use it and customize it based on your checkout design, while creating a more consistent and informative experience to help your customers fly through your checkout flows more easily.
Previously, you told us about the importance of a consistent checkout experience for your business. We gave the Google Pay button a fresh new look, applying the latest Material 3 design principles. The new Google Pay button comes in two versions that make it look great on dark and light themed applications. We have also added a subtle stroke that makes the button stand out in low contrast interfaces. In addition, we are introducing new customization capabilities that make it easy to adjust the button shape and corner roundness and help you create more consistent checkout experiences.
 |
| Figure 1: The new Google Pay button view for Android can be customized to make it more consistent with your checkout experience. |
We also improved the Google Pay API for Android, introducing a new button view that simplifies the integration. Now, you can configure the Google Pay button in a more familiar way and add it directly to your XML layout. You can also configure the button programmatically if you prefer. This view lets you configure properties like the button theme and corner radius, and includes graphics and translations so you don't have to download and configure them manually. Simply add the view to your interface, adjust it and that's it.
The new button API is available today in beta. Check out the updated tutorial for Android to start using the new button view.
 |
| Figure 2: An example implementation of how you can add and configure the new button view directly to your XML layout on Android. |
Earlier, we introduced the dynamic Google Pay button for Web, which previewed the card network and last four digits for the customer’s last used form of payment directly on the button. Since then, we have seen great results, with this additional context driving conversion improvements and increases of up to 40% in the number of transactions.
Later this quarter, you’ll be able to configure the new button view for Android to show your users additional information about the last card they used to complete a payment with Google Pay. We are also working to offer the Google Pay button as a UI element in Jetpack Compose, to help you build your UIs more rapidly using intuitive Kotlin APIs.
 |
| Figure 3: An example of how the dynamic version of the new Google Pay button view will look on Android. |
Check out the documentation to start using the new Google Pay button view on Android. Also, check out our own button integration in the sample open source application in GitHub.
The Stable and extended stable channel has been updated to 113.0.5672.92/.93 Windows and 113.0.5672.92 for Mac and Linux which will roll out over the coming days/weeks. A full list of changes in this build is available in the log.
Interested in switching release channels? Find out how here. If you find a new issue, please let us know by filing a bug. The community help forum is also a great place to reach out for help or learn about common issues.
Infrastructure fragmentation in Machine Learning (ML) across frameworks, compilers, and runtimes makes developing new hardware and toolchains challenging. This inhibits the industry’s ability to quickly productionize ML-driven advancements. To simplify the growing complexity of ML workload execution across hardware and frameworks, we are excited to introduce PJRT and open source it as part of the recently available OpenXLA Project.
PJRT (used in conjunction with OpenXLA’s StableHLO) provides a hardware- and framework-independent interface for compilers and runtimes. It simplifies the integration of hardware with frameworks, accelerating framework coverage for the hardware, and thus hardware targetability for workload execution.
PJRT is the primary interface for TensorFlow and JAX and fully supported for PyTorch, and is well integrated with the OpenXLA ecosystem to execute workloads on TPU, GPU, and CPU. It is also the default runtime execution path for most of Google’s internal production workloads. The toolchain-independent architecture of PJRT allows it to be leveraged by any hardware, framework, or compiler, with extensibility for unique features. With this open-source release, we're excited to allow anyone to begin leveraging PJRT for their own devices.
We are entering a world of ambient experiences where intelligent apps and devices surround us, from edge to the cloud, in a range of environments and scales. ML workload execution currently supports a combinatorial matrix of hardware, frameworks, and workflows, mostly through tight vertical integrations. Examples of such vertical integrations include specific kernels for TPU versus GPU, specific toolchains to train and serve in TensorFlow versus PyTorch. These bespoke 1:1 integrations are perfectly valid solutions but promote lock-in, inhibit innovation, and are expensive to maintain. This problem of a fragmented software stack is compounded over time as different computing hardware needs to be supported.
The workflow to enable this vision with PJRT is as follows (shown in Figure 1):
The PJRT API will be backward compatible. The plugin would not need to change often and would be able to do version-checking for features.
 |
| Figure 1: To target specific hardware, provide an implementation of the PJRT API to package a compiler and runtime plugin that can be called by the framework. |
As a foundational pillar of the OpenXLA Project, PJRT is well-integrated with projects within the OpenXLA Project including StableHLO and the OpenXLA compilers (XLA, IREE). It is the primary interface for TensorFlow and JAX and fully supported for PyTorch through PyTorch/XLA. It provides the hardware interface layer in solving the combinatorial framework x hardware ML infrastructure fragmentation (see Figure 2).
 |
| Figure 2: PJRT provides the hardware interface layer in solving the combinatorial framework x hardware ML infrastructure fragmentation, well-integrated with OpenXLA. |
Industry partners such as Intel and others have already adopted PJRT.
Intel is leveraging PJRT in Intel® Extension for TensorFlow to provide the Intel GPU backend for TensorFlow and JAX. This implementation is based on the PJRT plugin mechanism (see RFC). Check out how this greatly simplifies the framework and hardware integration with this example of executing a JAX program on Intel GPU.
"At Intel, we share Google's vision of modular interfaces to make integration easier and enable faster, framework-independent development. Similar in design to the PluggableDevice mechanism, PJRT is a pluggable interface that allows us to easily compile and execute XLA's High Level Operations on Intel devices. Its simple design allowed us to quickly integrate it into our systems and start running JAX workloads on Intel® GPUs within just a few months. PJRT enables us to more efficiently deliver hardware acceleration and oneAPI-powered AI software optimizations to developers using a wide range of AI Frameworks." - Wei Li, VP and GM, Artificial Intelligence and Analytics, Intel.
Allen Hutchison, Andrew Leaver, Chuanhao Zhuge, Jack Cao, Jacques Pienaar, Jieying Luo, Penporn Koanantakool, Peter Hawkins, Robert Hundt, Russell Power, Sagarika Chalasani, Skye Wanderman-Milne, Stella Laurenzo, Will Cromar, Xiao Yu.
By Aman Verma, Product Manager, Machine Learning Infrastructure
Testing is an integral part of software engineering, especially in the context of payments, where small glitches can have significant implications for your business.
We previously introduced a set of test cards for you to use with the Google Pay API in TEST mode. These cards allow you to build simple test cases to verify that your Google Pay integration operates as expected. As much as this was a great start, a few predefined cards only let you run a limited number of happy path test scenarios, confined to the domain of your applications.
Today, we are introducing PSP test cards, an upgrade to Google Pay’s test suite that lets you use test cards from your favorite payment processors to build end-to-end test scenarios, enabling additional testing strategies, both manual and automated.
 |
| Figure 1: Test cards from your payment processor appear in Google Pay’s payment sheet when using TEST mode. |
When you select a card, the result is returned back to your application via the API, so you can use it to validate comprehensive payment flows end-to-end, including relaying payment information to your backend to complete the order with your processor. These test cards allow you to verify the behavior of your application against diverse payment outcomes, including successful and failed transactions due to fraud, declines, insufficient funds and more.
This upgrade also supports test automation, so you can write end-to-end UI tests using familiar tools like UIAutomator and Espresso on Android, and include them in your CI/CD flows to further strengthen your checkout experiences.
The new generation of Google Pay’s test suite is currently in beta, with web support coming later this year. You’ll be able to use test cards on Android for 5 of the most widely used PSPs – Stripe, Adyen, Braintree, WorldPay and Checkout.com, and we’ll continue to add test cards from your favorite PSPs.
Improved testing capabilities have been one of the most frequent requests from the developer community. From Google Pay, we are committed to providing you with the tools you need to harden your payment flows and improve your checkout performance.
 |
| Figure 2: With the upgraded test suite you can run end-to-end automated tests for successful and failed payment flows. |
Take a look at the documentation to start enhancing your payments tests. Also, check out the sample test suite in the Google Pay demo open source application.
These startups are using AI to decarbonize buildings, create sustainable agriculture, protect biodiversity, and remove carbon from the atmosphere.
En honor a la Semana de Apreciación al Maestro, la Maestra del Año 2023, Rebecka Peterson, comparte más sobre su historia y cómo Google está comprometido con el futuro d…
In honor of Teacher Appreciation Week, 2023 Teacher of the Year Rebecka Peterson shares more about their story and how Google is committed to the future of education.
Posted by Jose Ugia – Developer Relations Engineer
Posted by Jose Ugia – Developer Relations Engineer