Microprocessor Report (MPR) Subscribe

Current MPR Articles

May 16, 2022

  • D-Matrix Transforms SRAMs for AI
    The AI-processor startup has emerged from stealth mode with a chiplet-based inference engine that employs digital in-memory computing (IMC) to run Transformer models like those in Bert and GPT.
  • Versal Adds AI to Premium Line
    AMD has added two new FPGA products to its Versal Premium line. They combine AI engines to accelerate neural networks along with plenty of programmable logic and DSP blocks for signal processing.
  • Intel Blockscale Boosts Mining Profit
    Intel has introduced a new bitcoin-mining ASIC that competes respectably where the original test chip didn’t. The company will sell the chip directly rather than building its own system for sale.

May 9, 2022

  • AMD Ships 3D V-Cache Processors
    The Ryzen 7 5800X3D PC processor and Milan X server processor use TSMC’s chip-on-wafer (CoW) technology to stack more cache on the compute die, improving game and HPC performance.
  • Intel GNA Engine Adds Vision
    The Gaussian and Neural Accelerator supports always-on audio and vision workloads, reducing power and adding security. It appears on Intel PC processors, other chips, and the Clover Falls add-on chip.
  • Chip Vendors Help Power Up 6GHz
    Broadcom and Qualcomm will support a new automated-frequency-coordination (AFC) system to allow Wi-Fi and unlicensed 5G devices to use higher power in the 6GHz band, increasing performance and range.

May 2, 2022

  • Fewer Find 3nm Worth the Wait
    TSMC and Samsung have suffered lengthy delays in their 3nm processes, and the gain in density and other characteristics is smaller than in previous nodes. Intel says its future nodes are on schedule.
  • GF Fotonix Unites CMOS, Photonics
    The new GlobalFoundries process combines photonic and digital components on a single 45nm chip. The primary application is data-center communications, but lidar and computing will benefit as well.
  • AMD Adds DPUs With Pensando Buy
    By acquiring Pensando, AMD will instantly offer DPUs alongside its CPUs and GPUs. Although it principally sells smart NICs, the startup recently announced a design win for its 7nm second-generation chip.

April 25, 2022

  • Synopsys NPX6 Expands AI Options
    The new deep-learning accelerator (DLA) can scale to more than 2,000 TOPS, providing a licensable core for applications such as autonomous driving and natural-language processing.
  • Sakura Debuts for Edge AI
    EdgeCortix launched its first edge-AI inference chip by hardening its DNA IP, delivering low latency and high power efficiency for applications that fit into 5W to 20W.
  • TetraMem Touts In-Memory DLA
    The Silicon Valley startup uses a new type of memristor to perform analog math for deep-learning acceleration. It provides nonvolatile storage and performs low-power AI computations.

April 18, 2022

  • Alchemist Turns Intel Into GPU Vendor
    Intel’s new Arc products are its first significant discrete GPUs, using competitive performance and hardware ray tracing to bring new competition to laptop-PC graphics cards.
  • Editorial: Where’s the AI Competition?
    The dominant AI-chip vendor, Nvidia has raised the bar with its new Hopper and Orin accelerators. Startups, hyperscalers, and large chip vendors try to compete with the company but keep falling short.
  • Nvidia Orin Appears in MLPerf
    The newest set of MLPerf Inference results showcase the same old vendors; almost all the data-center and edge accelerators came from Nvidia and Qualcomm. Orin was the notable newcomer.

April 11, 2022

  • Nvidia Hopper Leaps Ahead
    The next-generation AI architecture powers the H100 card and DGX-H100 system. The 700W flagship card triples peak performance over Ampere while adding FP8 support for more-efficient training.
  • Merging Memory and Compute
    Near-memory and in-memory compute are techniques for reducing computing power—especially for AI. But they mean different things to different companies.
  • Nvidia Reveals 51Tbps Ethernet Switch
    Due to sample late this year, Spectrum-4 will quadruple the bandwidth of Spectrum-3 by both doubling per-lane speed to 100Gbps and doubling the number of serdes lanes to 512, yielding 51.2Tbps.

April 4, 2022

  • Renesas Refocuses on Enhanced ADAS
    The R-Car V4H is a new automotive system-on-chip for vehicles with extended ADAS capabilities. It addresses functional safety and the greater AI performance for Level 2+ driving.
  • Agilex Dials M for Memory
    Intel announced its first Agilex to copackage DRAM, sporting a pair of HBM2e stacks that deliver up to 32GB. It’s also using the Intel 7 process to further cut power dissipation.

March 28, 2022

  • FuriosaAI Seeks Data-Center Efficiency
    The startup’s first product, Warboy, offers a flexible architecture that excels on single-stream workloads and depthwise-convolution models such as SSD. It’s more power efficient than Nvidia’s low-end A2.
  • UCIe Addresses Chiplet Interconnect
    The new Universal Chiplet Interconnect Express (UCIe) specification provides a way to connect chiplets in a package at the physical and protocol level with very high data rates.
  • Ceva’s PentaG2 Reduces 5G Power
    The new 5G baseband design adds hardware accelerators that can completely offload data-plane processing from the DSP, reducing energy per bit. Ceva is licensing it for mobile and IoT applications.
  • M1 Ultra Sports 20 CPU Cores
    For Apple’s new desktop processor, two die are better than one. The company has finished its move from x86 by connecting two M1 Max die in one package to create a huge product for its highest-end desktops.

March 21, 2022

  • Graphcore First With WoW
    The lead customer for TSMC’s wafer-on-wafer technology achieved a 40% clock-speed boost by applying deep-trench capacitors, creating the new Bow accelerator for data-center AI training.
  • Dimensity 8100 Raises AI Bar
    MediaTek is doubling AI performance for mid-premium smartphones with its Dimensity 8100 and 8000. The new processors include features from the premium Dimensity 9000 but at a lower price.
  • Mythic Enables Smaller PCIe Card
    The startup’s new M1076 replaces the previous M1108 while adopting a smaller package and lower performance. The company is now on track to ship its first production products in September.

March 14, 2022

  • Xeon D Gets Major Network Upgrade
    Intel’s Xeon D-2700 family, based on the 10nm Ice Lake D, delivers fresh networking features that first appeared in Snow Ridge, such as an integrated Ethernet switch and in-line IPSec processing.
  • Wi-Fi 7 Quadruples Peak Data Rates
    Wi-Fi 7 will bring higher data rates and lower latency to home networks in the 6GHz band. Qualcomm, MediaTek, and others are already developing chips, aiming for products on shelves early next year.
  • Mike Demler Retires, Bryon Moyer Joins
    We thank Mike for his extensive contributions to Microprocessor Report over a nine-year run with the publication. Bryon brings a strong combination of technical and writing experience to his new role.

March 7, 2022

  • Intel’s Bitcoin-Mining Bonanza
    Gold miners helped California boom in the nineteenth century, but in the twenty-first century, cryptocurrency mining is the new mania. Intel is tapping this vein with its new Bonanza Mine product.
  • Editorial: Xeon Roadmap Shows Delays
    General availability of Sapphire Rapids has slipped, creating a ripple effect on future generations. But Intel plans a new line of power-efficient server processors beginning with Sierra Forest in 2024.
  • AMD Powers Steam Deck Console
    Valve, the company that operates the Steam platform for online gaming, has introduced a PC-compatible handheld console using a semicustom AMD processor with Zen 2 CPUs and an RDNA2 GPU.

February 28, 2022

  • Exynos 2200 First With AMD GPU
    Samsung’s flagship processor features a PC GPU architecture with hardware ray tracing, a first for smartphones. It also upgrades to Arm v9 CPUs for a big performance gain while doubling AI TOPS.
  • Intel Pushes Foundry Service
    To aid its IDM 2.0 strategy, Intel acquired Tower Semiconductor, launched a $1 billion fund to invest in companies that support its foundry (IFS), announced two foundry customers, and joined RISC-V International.
  • Picocom Thinks Big for Small Cells
    Picocom’s new PC802 is a Layer 1 baseband SoC for the emerging 5G small-cell market. One chip can handle up to 2,000 users, and it typically consumes only 4W.
  • Seeing Machines Eyes Unsafe Drivers
    For more than 20 years, the Australian company has developed computer-vision algorithms that can track a person’s eyes, and it introduced one of the first driver-monitoring systems (DMSs).

February 21, 2022

  • Kompanio 1380 Improves CPU, AI
    Targeting midrange Chromebooks, MediaTek’s new processor raises multithread-CPU performance by 50% and doubles AI performance compared with the previous Kompanio 820.
  • LeapMind Scales Binary DLA
    LeapMind is licensing its second-generation Efficiera deep-learning accelerator (DLA). The binary-optimized design scales to 24 trillion binary operations per second (TOPS).

More Articles


 

Free Newsletter

Linley Newsletter
Analysis of new developments in microprocessors and other semiconductor products

Events

Linley Spring Processor Conference 2022
Registration now open: April 20-21, 2022
Hyatt Regency Hotel, Santa Clara, CA
Linley Fall Processor Conference 2021
Held October 20-21, 2021
Proceedings available
Linley Spring Processor Conference 2021
April 19 - 23, 2021
Proceedings Available
More Events »