Parse notes
Sponsored
Manual testing on personal devices is TOO SLOW AND LIMITED. Teams CUT RELEASES A WEEK EARLY just to test before submitting them to app stores— and without broad device coverage, issues slip through.
Nvidia has announced new chips for building and deploying artificial intelligence models. Blackwell Ultra will allow cloud providers to make as much as 50 times the revenue as chips from the Hopper generation - it will ship in the second half of this year. Vera Rubin, which is expected to ship in the second half of 2026, is a system with two main components: a CPU called Vera and a new GPU design called Rubin. The custom Vera design will be twice as fast as the CPU used in last year's Grace Blackwell chips - it will enable Rubin to manage 50 petaflops while doing inference, more than double the capacity of the company's current Blackwell chips.
Google is acquiring Wiz for $32 billion. Wiz will continue to be positioned as a multicloud offering. The acquisition brings a massive customer list to Google, so the companies need to ensure existing customers don't start shopping around for another security provider. Many customers chose Wiz due to its ability to support multiple cloud platforms - cutting off that ability risks alienating those users. Google Cloud trails significantly behind its competitors with a 12% share of the global cloud market.
Nvidia and Google DeepMind's Newton is a physics engine that can simulate robotic movements in real-world settings. It will be used by Disney to power its next-generation entertainment robots - a video of one of these robots is available in the article. Newton helps robots be more expressive and learn how to handle complex tasks with precision. It is highly customizable - developers can use it to program robotic interactions with a range of deformable objects. Newton will be compatible with Google DeepMind's ecosystem of robotic development tools.
Firefly Aerospace's Blue Ghost lander, part of NASA's Commercial Lunar Payload Services (CLPS) program, has accomplished a lot on the Moon in the last two weeks. It has drilled into the Moon's surface, tested an extraterrestrial vacuum cleaner, demonstrated that GPS can be used to navigate on the lunar surface, and sent daily imagery back to Earth. All of the mission objectives for the Blue Ghost mission have been met. This article details how Firefly managed to secure the first fully successful commercial lunar landing, the challenges Firefly's engineers have had to overcome, some of the discoveries that Blue Ghost has made so far, and the progress of other missions in the CLPS program.
On-call, in its current state in Big Tech, is flawed. AI is already changing the scope of on-call and the responsibility of the engineer in the loop. Some on-call will always be unavoidable, but it should be the exception rather than the norm. This will give engineers time to concentrate on meaningful tasks, which will make customers happier.
Services should minimize the number of synchronous requests they make to other services as much as possible. Synchronous calls are costly and impact the availability of services. The more services that an app relies on synchronously, the lower the availability of that app will be. Making a synchrony budget is about being mindful of how communication flows are implemented between services.
Keith O'Brien, a global payroll-compliance manager at Rippling, a business software startup, has been accused of stealing trade secrets for a rival, Deel, as part of a spying operation. O'Brien allegedly searched for 'deel' an average of 23 times a day across Rippling's internal systems, accessing message boards that had nothing to do with his job in a mode that hid his presence from others. Rippling claims that it set a honeypot trap that proved O'Brien was being directed by Deel's top executives. O'Brien fleed Rippling's offices with his smartphone after being accused of being a corporate spy despite being told that failing to turn over the device violated a court order and could result in imprisonment.
Core Devices, a company by the original creator of the Pebble smartwatch, is getting ready to ship two new watches. They will have all of the features of the old Pebble devices and run the newly open-sourced Pebble operating system. The Core 2 Duo will ship in July for $149. It has the exact same black-and-white e-paper display as the old Pebble 2 and even comes in the exact same frame. It has a battery that lasts more than 30 days and an in-device speaker for chatting with AI assistants. The Core Time 2 will ship in December and cost $225 - it will have a touchscreen.
NVIDIA Dynamo is an inference engine agnostic, high-throughput, low-latency inference framework designed for serving generative AI and reasoning models in multi-node distributed environments.
The construction for the site, which will have a capacity of 1.2 gigawatts of power, will be completed by mid-2026.
EBS failure is constant, variable, and all by design.
CoreWeave has filed paperwork for a public listing - if the initial public offering goes ahead, it will be the first prominent AI startup to reach the stock market.
This post shows how to increase jq's performance by 90% by just compiling the package again using the same source package that Ubuntu uses.
Llama has achieved widespread success since launching in 2023, with companies including Spotify, AT&T, and DoorDash using Llama models in production today.