brion.com brion.com ................................................... THE INDUSTRY’S LEADING IMAGE-BASED HARDWARE ACCELERATED COMPUTATIONAL LITHOGRAPHY PLATFORM WITH UNCOMPROMISED PERFORMANCE ...................................................... Based on the revolutionary Tachyon 2.0 architecture, Tachyon 2.5 continues to lead the industry with simulation accuracy, computational performance and cost of ownership to meet the low k lithography challenges of the 32nm and 1 22nm technology nodes. Pioneering Technology Tachyon 2.5 builds upon the pioneering technology of Brion’s Tachyon, the industry’s first high-performance, image-based, dense-simulation computational lithography platform. Since first shipped in 2003, two generations of Tachyon systems have been widely adopted by the global semiconductor industry for full-chip, model-based optical proximity correction (OPC) and OPC verification. Optimized for fast and accurate lithography simulation, Tachyon represents a fundamental improvement over polygon-based sparse simulation approaches. Tachyon 2.5 inherits its predecessor Tachyon 2.0’s hybrid computing architecture, which combines tier-1 Linux servers with a flexible coprocessing configuration. Tachyon 2.5 represents another milestone in computational lithography, combining innovations in hardware design, software design, and system integration. Tachyon 2.5 continues to deliver outstanding performance that surpasses anything that can be achieved using generic computing platforms for computational lithography applications. Now, with the flexible Tachyon 2.0 architecture, all existing Tachyon 2.0 systems are field upgradable to Tachyon 2.5. Product Features Platform Highlights  Fully integrated  Hardware accelerated Uncompromised Performance  Dense image-based computations Tachyon 2.5 gives users the uncompromised performance of a fully  Full-chip lithography applications integrated and optimized system solution for computational lithography. It Supported Application Products offers up to a 2.5x performance improvement over the already powerful  Lithography modeling: Tachyon FEM, Tachyon 2.0 for resolution enhancement technology (RET), OPC and Tachyon M3D, Tachyon iCAL verification applications. It continues to support up to 256 complex optical  Optical Proximity Correction: Tachyon OPC+ simulation kernels, up to 8 micron diameter optical interaction range for the  Lithography verification: Tachyon RDI/LMC 45nm node, and up to 7 micron diameter for 32nm and 22nm nodes. With Tachyon 2.5’s speed, a single system rack provides the equivalent  Resolution Enhancement Techniques: Tachyon DDL, Tachyon DPT production capacity of up to two and a half Tachyon 2.0 racks, and ten first-generation Tachyon racks. Tachyon 2.5 features new quad-core Uncompromised Performance servers, fully re-engineered acceleration hardware and software, and  Up to 256 complex kernels for numerous system level enhancements. With Tachyon 2.5, more optical modeling computationally intensive operations are off-loaded to the acceleration  Up to 8µm diameter optical interaction hardware, making full process window optimization and verification even range for 45nm node, and 7µm for 32nm and 22nm nodes more accessible for production uses. Tachyon 2.5 also introduces built-in coprocessing monitoring and management to ensure optimum use of the  Fast 3D mask simulation servers and the acceleration hardware. Tachyon 2.5 provides a powerful, Fast, Predictable Turnaround Time fast and efficient low k computational lithography platform for advanced 1  Linearly scales with chip area semiconductor design and manufacturing.  Linearly scales with system configuration Tachyon 2.5 platform yields enhanced Quad-Core performance Tachyon 2.5 decreases process window OPC overhead  Insensitive to design hierarchy, pattern 12 complexity and file size 8.0 Tachyon OPC+ Conventional OPC OPC + on Tachyon 2.0 OPC + on Tachyon 2.5 * Conventional OPC on Generic computing platform Tachyon 2.5 7.0 10 6.0 Supports Lithography Process Variability-Aware 8 5.0 Correction and Verification Tachyon 2.0 6 4.0  Wafer focus-exposure variations 3.0 4 Tachyon 1200 2.0  Mask CD variations 2 1.0 0  0.0 Scanner optical characteristics 1 3 5 7 Single-Core Dual-Core Quad-Core Number of Focus Conditions  Overlay misalignment The Tachyon platform consistently delivers performance By porting additional computations to the FPGA, surpassing that of generic computing platform. Tachyon Tachyon 2.5 further improves the computational System Configuration 2.5 upgrades enable additional speed improvements efficiency of process window based applications. relative to Quad-Core upgrades only. Unlike conventional OPC, Tachyon 2.5 enables  Flexible coprocessors process window OPC with minimal overhead for * Intel Performance Benchmark Testing, May 2007. each additional focus condition.  Powerful tier-1 Linux servers  Fast quad-core dual processors CD Error vs. Pitch for Different Ambit Sizes Runtime vs. Ambit (X Dipole, Y Polarized, s = 0.87/0.97, NA = 1.35)  Large RAID, extendable to 34TB 0 100 200 300 400 500 0 1.4  Integrated diagnostics and monitoring 1.2 -1 1 1.98 µm -2 0.8  Field upgradable from Tachyon 2.0 3.0 µm 0.6 -3 4.98 µm 0.4 -4 7.0 µm 0.2 -5 0 0 1 2 3 4 5 6 Pitch (nm) Ambit (µm) Rigorous simulations show the need for an ambit greater than 3 micron to achieve the necessary accuracy for sub-32nm process generations. Tachyon 2.5 provides optical ambits of up to 7 microns with 256 simulation kernels delivering best-in-class model accuracy without any productivity trade-off. © 2008 Brion Technologies, Inc. All rights reserved. Brion Technologies, Inc. • Phone: +1 (408) 653-1500 • Email: info@brion.com • brion.com Error (nm) OPC Throughput Normalized OPC Runtime Normalized Runtime