Skip to main content

Scientists Discover Way to Send Information into Black Holes Without Using Energy

3 comments

This New Memory Technology Sorts Data Without Comparisons, 160× More Energy Efficient & Could Power Next-Gen Computers

Sorting may sound like a basic operation, but it sits at the heart of almost every modern computing task. From organizing search results and managing databases to routing network traffic and running artificial intelligence models, fast and efficient sorting is essential. However, as data volumes explode, traditional sorting methods are increasingly struggling to keep up. A new breakthrough from researchers at Peking University offers a powerful alternative—one that could fundamentally change how computers sort data.

Why Sorting Is Becoming a Bottleneck

Most of today’s computing systems are built on the von Neumann architecture, where data is stored in memory and processed separately in a central processor. Every time data needs to be sorted, it must travel back and forth between memory and the processor. This constant movement creates two major problems.

First, it consumes a lot of time. Data transfer bandwidth becomes a limiting factor, especially when dealing with large datasets. Second, it wastes energy, since moving data costs far more power than simply computing on it. On top of that, traditional hardware sorters rely heavily on comparison operations—checking whether one value is larger or smaller than another. As datasets grow, the number of comparisons increases rapidly, slowing everything down.

To overcome these challenges, researchers have been exploring a concept called sort-in-memory, where sorting is performed directly inside memory itself. Among the most promising technologies for this approach are memristors.

What Are Memristors and Why Do They Matter?

A memristor is a type of memory device that can store data as resistance levels. Unlike conventional CMOS memory, memristors are non-volatile, meaning they retain information even when power is turned off. They are also compact, energy-efficient, and capable of performing computation directly where data is stored.

Many previous sort-in-memory designs using memristors still depended on traditional comparison-based methods. As a result, while they reduced data movement, they did not fully eliminate the performance limitations of comparisons. This is where the new research from Peking University stands out.

A Comparison-Free Approach to Sorting

The research team, led by Lianfeng Yu, introduced a fast and reconfigurable comparison-free sort-in-memory system built on one-transistor–one-resistor (1T1R) memristor arrays. Instead of comparing values one by one, their system uses a clever technique called digit reads.

In simple terms, digit reads allow the system to examine specific bits or digits of stored data across many memory cells at the same time. By organizing these reads in a structured way, the system can determine the correct order of data without performing explicit comparisons.

This is achieved through a method known as digit-read tree node skipping.

Understanding Digit-Read Tree Node Skipping

Imagine sorting numbers by checking their digits from the most significant to the least significant, much like how humans might sort words alphabetically. The researchers designed a tree-like structure where each node represents a digit read.

The key innovation is node skipping. If the system can determine that certain branches of the tree do not need to be explored—for example, because no data falls into that category—it simply skips them. This dramatically reduces the number of operations required.

What makes this approach even more powerful is its flexibility. It supports:

  • Different data sizes

  • Multiple data types

  • Varying sorting requirements

This makes the system highly adaptable for real-world applications.

Scaling Up with Advanced Strategies

To further enhance performance, the researchers extended their design using several advanced strategies.

One is the multi-bank strategy, which divides memory into multiple banks that can operate in parallel. This increases throughput by allowing many digit reads to happen simultaneously.

Another is the bit-slice strategy, where data is sliced into bits and processed across different arrays. This improves efficiency and scalability, especially for large datasets.

Finally, the multi-level strategy enables cross-array tree node skipping. In simple terms, the system can skip unnecessary operations not just within a single memory array, but across multiple arrays working together. This level of coordination is crucial for handling complex sorting tasks at high speed.

Impressive Experimental Results

The researchers didn’t just propose a theoretical model—they built and tested their system experimentally. The results are striking.

Compared with conventional sorting systems, their comparison-free sort-in-memory design achieved:

  • 7.70× higher throughput

  • 160.4× better energy efficiency

  • 32.46× higher area efficiency

These gains are significant. Higher throughput means faster sorting. Better energy efficiency means lower power consumption, which is critical for data centers and edge devices alike. Improved area efficiency means more functionality can be packed into smaller chips.

Beyond Sorting: Real-World Applications

To demonstrate that their approach is not limited to simple benchmarks, the researchers applied it to real computing tasks.

One example is Dijkstra’s shortest path algorithm, a widely used method in navigation systems and network routing. Sorting plays a key role in selecting the next closest node. The memristor-based sort-in-memory system significantly ускорated this process by reducing overhead.

Another application is neural network inference with in situ pruning. Modern AI models often need to remove unimportant connections to improve efficiency. Sorting helps identify which weights to prune. By performing sorting directly in memory, the system enables faster and more energy-efficient AI inference.

Importantly, the researchers showed that their sorting method is compatible with other compute-in-memory schemes, meaning it can be integrated into broader memristor-based computing systems.

Why This Matters for the Future of Computing

As data continues to grow at an unprecedented rate, traditional computing architectures are reaching their limits. Moving data back and forth between memory and processors is no longer sustainable in terms of speed, energy, or cost.

This new comparison-free sort-in-memory system points toward a future where memory and computation are tightly integrated. By eliminating unnecessary comparisons and data transfers, memristor-based designs can deliver massive gains in performance and efficiency.

While challenges remain—such as large-scale manufacturing and long-term reliability—this research marks an important step forward. It shows that sorting, one of the most basic operations in computing, can be reimagined in a way that better fits the needs of the data-driven world.

Conclusion

The work by Lianfeng Yu and his team at Peking University demonstrates how innovative hardware design can break long-standing bottlenecks in computing. By combining memristor technology with a clever, comparison-free sorting approach, they have opened the door to faster, greener, and more compact computing systems.

As applications like big data analytics, artificial intelligence, and graph processing continue to grow, such advances in sort-in-memory technology could play a crucial role in shaping the next generation of computers.

Reference: Yu, L., Zhang, T., Wang, Z. et al. A fast and reconfigurable sort-in-memory system based on memristors. Nat Electron 8, 597–609 (2025). https://doi.org/10.1038/s41928-025-01405-2

Comments

Post a Comment

Popular

Scientists Discover Way to Send Information into Black Holes Without Using Energy

For years, scientists believed that adding even one qubit (a unit of quantum information) to a black hole needed energy. This was based on the idea that a black hole’s entropy must increase with more information, which means it must gain energy. But a new study by Jonah Kudler-Flam and Geoff Penington changes that thinking. They found that quantum information can be teleported into a black hole without adding energy or increasing entropy . This works through a process called black hole decoherence , where “soft” radiation — very low-energy signals — carry information into the black hole. In their method, the qubit enters the black hole while a new pair of entangled particles (like Hawking radiation) is created. This keeps the total information balanced, so there's no violation of the laws of physics. The energy cost only shows up when information is erased from the outside — these are called zerobits . According to Landauer’s principle, erasing information always needs energy. But ...
3 comments
Read more

A New Type of Wormhole Could Slowly Become Stable Over Time, Scientists Suggest

For many years, wormholes have captured the imagination of scientists and science fiction fans. They are often shown as magical tunnels through space that can connect two faraway places in the universe. If wormholes really exist, they could one day make it possible to travel huge distances in a very short time. But there is one big problem—no one has ever found a real wormhole. They remain only theoretical objects predicted by the mathematics of Einstein's theory of general relativity. Even though they have never been observed, physicists continue studying them because they help us understand the limits of gravity and spacetime. Now, researchers Ditta and Channuie have proposed a new model of a time-dependent traversable wormhole . Unlike many earlier models, their wormhole is not completely still. Instead, it changes with time because energy flows through it. As this flow slowly fades away, the wormhole naturally becomes stable. Their study offers a new and simple way to understan...
Post a Comment
Read more

Black Holes That Never Dies

Black holes are powerful objects in space with gravity so strong that nothing can escape them. In the 1970s, Stephen Hawking showed that black holes can slowly lose energy by giving off tiny particles. This process is called Hawking radiation . Over time, the black hole gets smaller and hotter, and in the end, it disappears completely. But new research by Menezes and his team shows something different. Using a theory called Loop Quantum Gravity (LQG) , they studied black holes with quantum corrections. In their model, the black hole does not vanish completely. Instead, it stops shrinking when it reaches a very small size. This leftover is called a black hole remnant . They also studied something called grey-body factors , which affect how much energy escapes from a black hole. Their findings show that the black hole cools down and stops losing mass once it reaches a minimum mass . This new model removes the idea of a “singularity” at the center of the black hole and gives us a better ...
Post a Comment
Read more