2009-08-01

The Pentium contains a complicated circuit to multiply by three

A detailed analysis of the Intel Pentium's floating-point multiplier reveals a complex ×3 circuit containing around 9,000 transistors, utilizing base-8 multiplication for improved performance. The circuit combines advanced techniques like carry lookahead, Kogge-Stone addition, and carry-select addition to maximize speed and efficiency.

Original archive.is archive.ph web.archive.org

Log in to get one-click access to archived versions of this article.

read comments on news aggregators:

Related articles

Zen 5's AVX-512 Frequency Behavior

AMD's Zen 5 processor introduces full-width AVX-512 datapaths with impressive performance at high clock speeds, demonstrating significant improvements over Intel's Skylake-X implementation. The architecture employs sophisticated IPC throttling and clock management techniques to handle heavy AVX-512 workloads, maintaining optimal performance while avoiding fixed frequency offsets.

TSMC Faces Tough Choices Amid Rumors for Intel Collaboration

TSMC executives are set to negotiate with the U.S. government over three potential cooperation options, including setting up an advanced packaging factory and potentially partnering with Intel Foundry Services. The proposed arrangements could significantly impact TSMC's operations and relationships with existing partners, while raising concerns about balancing U.S. interests with shareholder value.

Game Bub: open-source FPGA retro emulation handheld

Game Bub is an open-source FPGA-based handheld device that can play Game Boy, Game Boy Color, and Game Boy Advance games from both physical cartridges and ROM files. The project showcases comprehensive hardware development including custom PCB design, FPGA implementation, and enclosure creation, featuring multiplayer support through Game Link Cable and various hardware improvements across two revisions.

Intel’s Battlemage Architecture

Intel's new Battlemage architecture powers the Arc B580 GPU, offering improved performance over its Alchemist predecessor despite fewer cores and a narrower memory bus, targeting the midrange market at $250 with 12GB VRAM. The architecture features significant improvements in compute utilization, cache latency, and memory handling, while maintaining Intel's unique approach to GPU design distinct from AMD and Nvidia.

LINUX is obsolete

A lively technical debate from 1992 between Andrew Tanenbaum and Linus Torvalds discusses the merits of microkernel vs monolithic kernel architectures, with Tanenbaum advocating for microkernels while Torvalds defends his monolithic design choices in Linux based on practical implementation considerations.

6502 is a Good Starting Point for Beginning Assembly Programming

The 6502 processor is recommended as an ideal starting point for learning assembly programming due to its simplicity with only 6 registers and 56 instructions, making core concepts easily digestible. Extensive learning resources and emulators are available through multiple platforms, though its practical applications are limited to retro programming.