Month: November 2019

Enlarge / The 49th Beidou navigation satellite was successfully launched by a Long March 3b carrier rocket from the Xichang Satellite Launch Center in southwest China on November 5, 2019. (credit: Costfoto / Barcroft Media via Getty Images)

In recent weeks, China's space program has made news by revealing some of its long-term ambitions for spaceflight. These include establishing an Earth-Moon space economic zone by 2050, which, if successful, could allow the country to begin to dictate the rules of behavior for future space exploration.

Some have questioned whether China, which has flown six human spaceflights in the last 16 years, can really build a large low-Earth space station, send taikonauts to the Moon, return samples from Mars, and more in the coming decade or two. But what seems clear is that the country's authoritarian government has long-term plans and is taking steps toward becoming a global leader in space exploration.

By one important metric—orbital launches—China has already reached this goal.

Enlarge / Intel's "Mega Trends in HPC" boil down to AI workloads, running on many kinds of hardware, largely in cloud—not on-premise—environments. (credit: Intel Corporation)

Saturday afternoon (Nov. 16) at Supercomputing 2019, Intel launched a new programming model called oneAPI. Intel describes the necessity of tightly coupling middleware and frameworks directly to specific hardware as one of the largest pain points of AI/Machine Learning development. The oneAPI model is intended to abstract that tight coupling away, allowing developers to focus on their actual project and re-use the same code when the underlying hardware changes.

This sort of "write once, run anywhere" mantra is reminiscent of Sun's early pitches for the Java language. However, Bill Savage, general manager of compute performance for Intel, told Ars that's not an accurate characterization. Although each approach addresses the same basic problem—tight coupling to machine hardware making developers' lives more difficult and getting in the way of code re-use—the approaches are very different.

• 

  In this simplified block diagram of AI/ML development, Intel wants to isolate "Languages & Libraries" as the biggest pain point. [credit: Intel Corporation ]

When a developer writes Java code, the source is compiled to bytecode, and a Java Virtual Machine tailored to the local hardware executes that bytecode. Although many optimizations have improved Java's performance in the 20+ years since it was introduced, it's still significantly slower than C++ code in most applications—typically, anywhere from half to one-tenth as fast. By contrast, oneAPI is intended to produce direct object code with no or negligible performance penalties.