Meteor Lake will introduce a multidie arrangement using Intel’s silicon-substrate and advanced packaging technologies. This chiplet approach will appear across all PC price tiers beginning late next year.
Renesas has introduced the first embedded processor with a user-programmable RISC-V CPU. Chinese processors are already adopting RISC-V; more will follow in a direct challenge to Arm.
The startup’s unique at-memory architecture targets an eye-popping 30 teraflop/s per watt and 12,000 teraflops in a single card. The second-generation chip is due to sample in 1H23.
The Israeli startup has built a proof-of-concept system for a simple optical neural network, but it will require additional effort to handle commercial AI models.
Broadcom is the first vendor to sample a 51Tbps data-center switch chip. The 5nm device features 100Gbps serdes and port speeds of up to 800G Ethernet.
Functional safety affects an increasing number of chip and IP designs. Vendors are precertifying chips, IP, and tools to make it easier for their customers to achieve certification.
TI has updated its low-cost SoC family with 64-bit CPUs, lower power, and improved security and functional-safety support, enabling it to target automotive as well as industrial applications.
Intel and Micron have developed a revolutionary new memory technology called 3D XPoint, but under the Optane brand it never met expectations as either an SSD or DIMM replacement.
Among flagship-smartphone processors, Google’s first custom design has the largest AI engine but the worst AI performance. In contrast, the Dimensity 9000’s AI engine is the most efficient.
Intel’s eight-laser silicon-photonics research chip could be a breakthrough for copackaged optics. It places the laser array on a silicon chip that can eventually be integrated in the package.
Qualcomm has updated its smartwatch platform with new chip designs. Reduced power enables a visible watch face even when idle; reduced chipset area enables smartwatches that operate globally.
Our die-photo analysis reveals that Samsung devoted too little area to its AMD-powered GPU, contributing to its performance shortfall. MediaTek and Intel both outperform the latest Apple chip.
Both leading foundries allowed customers to claim they were using a 4nm process when, in fact, they were using 5nm technology. This situation renders node names meaningless. Featured in our July 22 Microprocessor Report we are making this available, in full, for a limited time.
Overly cautious about rising development costs and risk, vendors aren’t refreshing their lower-cost processors, hurting themselves and customers. They must consider business cases other than merely chasing the biggest customers.
A research team led by Enlightra’s founders demonstrated a new method of photonic AI inference that is smaller than earlier photonics implementations. Using photonics also provides opportunities for resource sharing that aren’t available with electronics.