Samsung Electronics has begun mass producing NVM Express (NVMe) PCIe solid state drive (SSD) in a single ball grid array (BGA) package for next-generation PCs and ultra-slim notebook PCs. The new BGA NVMe SSD, named PM971-NVMe, features a compact package that contains all essential SSD components including NAND flash memory, DRAM and controller. Configuring the PM971-NVMe SSD in a single BGA package was enabled by combining 16 of Samsung's 48-layer 256Gb V-NAND flash chips, one 20nm 4Gb LPDDR4 mobile DRAM chip and a high-performance Samsung controller. The new SSD is 20x16x1.5mm and weighs only about one gram. The single-package SSD's volume is approximately a 100th of a 2.5-inch SSD or HDD, and its surface area is about a fifth of an M.2 SSD. The PM971-NVMe SSD enables sequential read and write speeds of up to 1,500MB/s and 900MB/s respectively, when TurboWrite technology is used. The performance figures can be directly compared to transferring a 5GB-equivalent, Full-HD movie in about three seconds or downloading it in about six seconds. It also boasts random read and write IOPS of up to 190K and 150K respectively. The PM971-NVMe SSD line-up will be available in 512GB, 256GB and 128GB storage options. Samsung will start providing the new SSDs to its customers in June worldwide.

Samsung Electronics has started mass production of NVMe PCIe solid state drive (SSD) which has an M.2 form factor for use in PCs and workstations. Samsung has also begun shipping NVMe SSDs to OEMs for the PC market. Samsung has been offering an AHCI-based PCIe 3.0 version of its SM951 SSD since early January. Now, it has added an NVMe version to the SSD portfolio. The new NVMe-based SM951 SSD achieves sequential data read and write performance up to 2,260 megabytes per second (MB/s) and 1,600MB/s, respectively. The drive attains its high-speeds by using four 8Gbps lanes (PCIe Gen 3.0 x4) of simultaneous data flow. This allows for a data transfer rate of 32Gbps and a maximum throughput of 4GB/s. When it comes to random read operations, the SM951-NVMe can process 300,000 IOPS (input output per second). Meeting all M.2 form factor requirements, the drive's thickness does not exceed 4mm (3.73mm maximum, when equipping chips on both sides of the board). Capacities are 512 gigabyte (GB), 256GB and 128GB.

Nvidia has announced that it plans to integrate a high-speed interconnect, called NVLink into its future GPUs, enabling GPUs and CPUs to share data five to 12 times faster than they can today. This will eliminate a longstanding bottleneck and help pave the way for a new generation of exascale supercomputers that are 50-100 times faster than today's most powerful systems, the vendor said. Nvidia will add NVLink technology into its Pascal GPU architecture - expected to be introduced in 2016 - following Nvidia's new Maxwell compute architecture for 2013. The new interconnect was co-developed with IBM, which is incorporating it in future versions of its Power CPUs. With NVLink technology tightly coupling IBM Power CPUs with Nvidia Tesla GPUs, the Power data center ecosystem will be able to fully leverage GPU acceleration for a diverse set of applications, such as high performance computing, data analytics and machine learning. Today's GPUs are connected to x86-based CPUs through the PCI Express (PCIe) interface, which limits the GPU's ability to access the CPU memory system and is four- to five-times slower than typical CPU memory systems. PCIe is an even greater bottleneck between the GPU and IBM Power CPUs, which have more bandwidth than x86 CPUs. As the NVLink interface will match the bandwidth of typical CPU memory systems, it will enable GPUs to access CPU memory at its full bandwidth.