» Archived Articles | Login | Activate My Subscription | Subscribe | Free Sample Articles | Purchase an Article

Current MPR Articles

July 26, 2021

• Die Photos Show Cortex-A78 Shortfall
  Arm’s new CPUs were supposed to bring new efficiencies to smartphone processors, but initial implementations of Cortex-X1 and Cortex-A78 are larger than expected, reducing area efficiency.
  

• Editorial: Huawei Builds New Supply Chain
  To combat US sanctions, Huawei is attempting to build an entirely Chinese supply chain by investing in local companies and constructing its own fab. Fully escaping the sanctions, however, will be difficult.
  

July 19, 2021

• Training Competitors Battle on MLPerf
  The latest benchmark scores for AI training show that three challengers—Graphcore’s GC200, Habana’s Gaudi, and Huawei’s Ascend—are making progress but still lag Nvidia’s A100 and Google’s TPUv4.
  

• Cadence Floats K-Series DSPs
  The new Tensilica KP1, KP6, KQ7, and KQ8 DSPs are optimized for floating-point workloads, offering up to a 40% area reduction and 30% energy savings relative to previous Tensilica DSPs.
  

• Qualcomm Targets 5G Embedded
  The four new embedded products target industrial tablets, drones, point-of-sale terminals, kiosks, and similar applications, particularly those that can take advantage of its integrated 5G modem.
  

July 12, 2021

• Marvell Adds Vectors to Octeon 10
  The company’s newest data-processing unit (DPU) adds vector packet processing and an AI engine. Marvell expects to sample the first Octeon 10 processors later this year, targeting edge and cloud networking.
  

• NXP S32R Handles Automotive Radar
  The new chips address the entire range of automotive-radar applications from ADAS to self-driving cars. They include the S32R294 microcontroller, S32R45 imaging-radar processor, and TEF82xx transceivers.
  

• ProteanTecs Aids in Manufacturing
  Startup ProteanTecs licenses its Universal Chip Telemetry platform, which combines AI models and a collection of sensors, to reduce chip defect rates by 10x and to cut manufacturing costs.
  

July 5, 2021

• P550 Doubles Floating-Point Capability
  After picking up new customers Renesas and Tenstorrent, SiFive released two new CPU cores: the P270 and P550. These offerings sit atop the product stack, targeting Linux-based application processors.
  

• Data Centers Drive to 800Gbps Optics
  At the recent Optical Fiber Communication Conference (OFC), Broadcom, Marvell, and MaxLinear announced 112Gbps PAM4 physical-layer chips, including the first built in 5nm technology.
  

• Quadric’s Flexible Edge Accelerator
  Startup Quadric is now sampling its first chip, a data-flow processor that can handle a range of tasks from signal processing to AI. The 256-core design consumes less than 4W and targets edge applications.
  

June 28, 2021

• FinFET Roadmap Ends at 3nm
  The FinFET has enjoyed a good 10-year run, but the next process node will likely be its last. At 3nm, TSMC will stick with FinFETs, but Samsung plans to switch to next-generation nanosheet transistors.
  

• New AI Engines Give Versal an Edge
  Xilinx introduced the AI Edge line for automotive and industrial applications. An updated AI Engine doubles compute capacity relative to the previous generation.
  

• CertusPro-NX Delivers 10G Serdes
  Continuing to extend its Nexus family, Lattice is again taking advantage of fully depleted silicon-on-insulator (FD-SOI) technology to challenge Intel and Xilinx in low-density FPGAs.
  

June 21, 2021

• Intel Fills Out 11th-Gen Laptop Lineup
  Intel’s 10nm process is displacing 14nm across the Core laptop lineup. Tiger Lake-H replaces Comet Lake-H for mobile gamers, whereas new Tiger Lake-UP3 versions update the thin-and-light segment.
  

• Radeon 6800M Tackles GeForce 3080
  AMD updated its laptop GPUs to the RDNA2 architecture, boosting gaming performance by 50% over the previous generation and matching up against Nvidia’s RTX 3080 for Laptops.
  

• Upmem Nails RowHammer
  French startup Upmem has patented a new defense against RowHammer attacks that corrupt memory. Dubbed Silver Bullet, it’s available to DRAM manufacturers as licensable intellectual property.
  

June 14, 2021

• Editorial: US Fabs Need Government Aid
  The United States Congress is moving toward approving subsidies for building fabs in America. This funding is necessary to offset government fab investment in China, South Korea, and Taiwan.
  

• Marvell PHYs Drive Multigigabit Cars
  Marvell’s new 88Q4364 Ethernet PHY complies with the 802.3ch automotive Ethernet standard, which the IEEE ratified last year. It delivers 2.5Gbps, 5Gbps, or 10Gbps on shielded single-pair wiring.
  

• Efinix Fills Out 16nm FPGA Roadmap
  Three years after sampling its first chip, Efinix is sampling second-generation Trion Titanium FPGAs manufactured in TSMC 16nm. The latest family members offer greater density and new features.
  

June 7, 2021

• Arm Extends Valhall to Low-End GPUs
  Arm’s third-generation Valhall GPUs boosts performance across all Mali product tiers. The Mali-G310 is the company’s first new entry-level design in three years, succeeding the Bifrost-based Mali-G31.
  

• MLPerf 1.0 Adds Power Measurements
  Nvidia dominated the latest round of MLPerf Inference submissions, but Qualcomm’s Cloud AI delivered leading performance per watt on two of the six neural networks. Many other vendors were absent.
  

• Agilex Moves to SuperFin
  Intel’s initial 10nm FPGAs are now slated to reach general availability in Q3 after a delay owing to a redesign to the 10nm SuperFin process, which offers a dramatic reduction in power.
  

May 31, 2021

• Arm v9 Yields Three New CPUs
  In a company first, Arm released three CPUs at once, prompted by its transition to Arm v9. They include Cortex-X2, Cortex-A710, and Cortex-A510, the company’s first new “little” core in four years.
  

• Arm NI-700 Puts a NoC in CoreLink
  To support its new Arm v9 CPUs, the company has developed a new dynamic shared unit and two new CoreLink interconnect products, the CI-700 cache-coherent interconnect and the NI-700 network-on-a-chip.
  

• Dimensity 900 Jumps to Cortex-A78
  MediaTek has refreshed its Dimensity 800 family with the latest CPU and GPU cores as well as new standards such as Wi-Fi 6 and LPDDR5 to enhance performance in midrange 5G smartphones.
  

May 24, 2021

• IBM Test Chip Is First at 2nm
  IBM has manufactured the industry’s first test chip in 2nm technology. Compared with 7nm FinFETs, the gate-all-around (GAA) transistors allow 3.4x the density with 45% greater speed or 75% less power.
  

• Unisoc T760 Launches Tanggula
  Under its mountainous moniker Tanggula, Chinese vendor Unisoc plans to deliver four new 5G –chip families to revitalize its smartphone lineup. The first new product is a midrange chip called the T760.
  

• Lattice Sets Midrange FPGA Collision
  Known for tiny low-power FPGAs, Lattice is developing a new family called Avant that will target the midrange segment. It expects to introduce the first Avant devices late next year.
  

May 17, 2021

• Nvidia Orin Turbocharges ADAS
  Nvidia’s Orin chip delivers a big performance and power-efficiency boost compared with previous Drive Xavier systems. But the next-generation Atlan will raise the bar even higher—to 1,000 TOPS.
  

• Allwinner Creates First RISC-V AP
  Using a CPU licensed from Alibaba, Allwinner’s D1 is the industry’s first RISC-V application processor. Featuring a GPU, DSP, and audio/video codec, it is available in a $13 development board.
  

• Editorial: Nvidia Endangers Arm Model
  Leading Arm customers have criticized Nvidia’s planned acquisition, and some are even turning away from Arm cores. Nvidia’s new plans to enter the server market raise further conflicts of interest.
  

May 10, 2021

• Cadence Q8 Expands Vision DSPs
  Cadence’s new Tensilica Vision DSPs extend its intellectual-property (IP) lineup at the high end to target 3D computer vision and at the low end to reduce area and power in always-on smart sensors.
  

• SiFive 21G1 Update Boosts Hash Rates
  The intellectual-property (IP) company announced a new AI platform that uses the Intelligence extensions to reduce convolution-processing time by 4x compared with a standard RISC-V vector implementation.
  

May 3, 2021

• Nvidia Arms Its Data-Center Roadmap
  Nvidia aims to accelerate Arm-CPU adoption for cloud computing with the upcoming Grace processor as well as next-generation GPU and networking (DPU) products. The company also rolled out A10 and A30 GPUs for AI.
  

• Neoverse V1 Updates Cortex-X1
  Arm’s new Neoverse V1 and Neoverse N2 cores boost performance and add new instructions relative to the older Neoverse N1, but they still fall short of the leading x86 CPUs in single-thread performance.
  

• Movellus Improves Clock Meshes
  The startup’s Maestro intellectual property (IP) enhances clocking efficiency by up to 30%, boosting chip performance and saving power. The “virtual mesh” also accelerates timing closure by 5x.
  

More Articles

Purchase an Article