Google looks to differentiate with premium cloud tier: The public cloud market is dominated by Amazon, Microsoft, Google and IBM. While all offer general-purpose and specialized compute power for increasingly lower cost, Google is betting that its proprietary fiber network can give it a competitive edge.
The company has been building out its private fiber network for nearly 20 years, and it now has more than 100 POPs (points of presence) around the world. That translates to low latency and high availability. Current customers already benefit from this private network, but going forward, Google will also offer a lower-cost standard tier, which runs on ISP networks and has comparable performance to competing clouds, it said in a blog post. In this sense, Google isn't offering a higher-powered option than previously but can be more cost-competitive with other clouds.
Once generally available, Standard Tier will be priced up to 33 percent lower than Premium Tier in North America and Europe. While Premium Tier will remain the "default tier for your workloads," Standard will be preferable for some scenarios, Google says. It has certain limitations as well, such as regional versus global load balancing capabilities.
POV: "It's good to offer customers choices, and Google has realized that the best solution doesn't have to be the right solution for all," says Constellation VP and principal analyst Holger Mueller. "But the ramifications go deeper, as with all network changes. Minutes, hours, or days of downtime can quickly become magnitudes more expensive than a higher-quality network that costs enterprises more money."
Walmart fires another e-commerce arrow at Amazon: The world's largest company has been in an e-commerce dogfight against Amazon, and in its latest move to fend off the online giant is teaming up with Google. Walmart customers will be able to shop for products via voice Google Assistant or through the Google Express website and application:
If you’re an existing Walmart customer, you can choose to link your Walmart account to Google and receive personalized shopping results based on your online and in-store Walmart purchases. For example, if you order Tide PODS or Gatorade, your Google Assistant will let you know which size and type you previously ordered from Walmart, making it easy for you to buy the right product again.
POV: Google had already signed up for similar partnerships with Costco and other retailers, but the Walmart deal will certainly move the needle. The partnership has benefits for both sides; Google Express is a small fish in the e-commerce pond right now, but access to Walmart's massive customer pool—and related historical data that will help improve Google Assistant's accuracy—will be a boon. In turn, Walmart will benefit from the large ecosystem that exists for Google Assistant across Android devices. While it's far from a killing blow against Amazon, this is one partnership that should make a mark on the matrix commerce landscape.
Apple opens up (a little) about Siri's R&D: Speaking of digital assistants and speech recognition technology, this week is the occasion of the preeminent academic conference covering those topics. Many industry mega-vendors, such as IBM and Microsoft, will have researchers on hand presenting papers, but one surprising participant is notoriously secretive Apple.
During the Interspeech conference this week in Stockholm, representatives from Cupertino will present a paper titled "Siri On-Device Deep Learning-Guided Unit Selection Text-to-Speech System. It describes "Apple’s hybrid unit selection speech synthesis system, which provides the voices for Siri with the requirement of naturalness, personality and expressivity."
While the paper is as technically dense as its kind tend to be, Apple has also released a blog post that provides a more accessible summary of the methodology behing Siri's development.
POV: A H/T goes to the Register for spotting Apple's paper on the Interspeech conference's website. Its release follows a pledge made last year by Apple's head of AI research that the company would be more open about AI going forward. It made good on that promise in December, releasing an initial academic paper focused on image recognition, which is an important AI capability for technologies such as driverless cars and security.
As we wrote at the time, Apple's motivations for opening up about AI are likely driven in part by a desire to attract top talent. It's good to see Apple continue to live up to its AI openness pledge, but expect it to understandably hold back on sharing its juiciest research.