Nowadays a placement problem often involves multi-million objects and excessive fixed blockages. We present a new global placement algorithm that scales well to the modern large-scale circuit placement problems. We simulate the natural diffusion process to spread cells smoothly over the placement region, and use both analytical and discrete techniques to improve the wire length. Although any analytical wire length technique can be used in our new framework, by using the quadratic wire length model, the hessian of our formulation is extremely sparse compared with conventional formulations, which brings 24x speed up on quadratic solver. We also propose a wire linearization technique that transform quadratic star model into HPWL exactly. The overall runtime of our tool is close to the fastest placement tool in existing literature and significantly better than others. And meanwhile, we obtain competitive wire length results to the best known ones. The average total wire length is 2.2% higher than mPL6, 0.2%, 3.1%, and 9.1% better than FastPlace3.0, APlace2.0, and Capo10.2 respectively.

Physical-synthesis tools are responsible for achieving timing closure. Starting with 130-nm designs, multiple cycles are required to cross the chip, making latch placement critical to success. We present a new physical-synthesis optimization for latch placement called Rip Up and Move Boxes with Linear Evaluation (RUMBLE) that uses a linear timing model to optimize timing by simultaneously replacing multiple gates. RUMBLE runs incrementally and in conjunction with static timing analysis to improve the timing for critical paths that have already been optimized by placement, gate sizing, and buffering. Experimental results validate the effectiveness of the approach: Our techniques improve slack by 41.3% of cycle time on average for a large commercial ASIC design.

The purpose of global placement is to find non-overlapping locations for cells, typically while minimizing a wirelength objective. Because of this objective, however, when more timing information about the design is known, some cells will inevitably be sub-optimally placed from a timing perspective. In this paper, we present two new techniques to incrementally improve placements by moving cells to their optimal timing locations. We call our approach Pyramids, since it uses pyramid-shaped delay surfaces to solve for the optimal location, rather than running a more expensive linear programming solver. We show how to apply these techniques to timing-driven detailed placement and also for more accurate late-stage incremental timing correction. Experimental results validate the effectiveness of Pyramids by showing significantly improved timing after an industrial placement algorithm. Furthermore, compared to the linear programming solvers, the speedup of Pyramids solver is 373times vs. CLP and 448times vs. GLPK.

Power dissipation is quickly becoming one of the most important limiters in nanometer IC design for leakage increases exponentially as the technology scaling down. However, power and timing are often conflicting objectives during optimization. In this paper, we propose a novel total power optimization flow under performance constraint. Instead of using placement, gate sizing, and multiple-Vt assignment techniques independently, we combine them together through the concept of slack distribution management to maximize the potential for power reduction. We propose to use the linear programming (LP) based placement and the geometric programming (GP) based gate sizing formulations to improve the slack distribution, which helps to maximize the total power reduction during the Vt-assignment stage. Our formulations include important practical design constraints, such as slew, noise and short circuit power, which were often ignored previously. We tested our algorithm on a set of industrial-strength manually optimized circuits from a multi-GHz 65 nm microprocessor, and obtained very promising results. To our best knowledge, this is the first work that combines placement, gate sizing and Vt swapping systematically for total power (and in particular leakage) management.

Although circuit placement has been studied for decades, it continuously attracts research attentions. The placement problems grow rapidly in both problem size and complexity. Some industry placement problems contain multimillion gates and excessive number of blockages [1,2]. In this chapter, we introduce DPlace, an anchor cell and diffusion spreading-based quadratic placement engine that can handle largescale placement problem.

In this paper, we propose a new linear programming based timing driven placement framework for high performance designs. Our LP framework is mainly net-based, but it takes advantage of the path-based delay sensitivity with limited-stage slew propagation, thus it enjoys certain hybrid feature of net and path-based timing driven placement. Our LP formulation considers not only cells on the critical paths, but also cells that are logically adjacent to the critical paths (i.e., the criticality adjacency network) in a unified manner. We further present a timing aware spreading method to preserve timing in legalization for high performance designs. Our algorithm has been tested on a set of 65nm industry circuits from a multi-GHz microprocessor, and shown to achieve much improved timing on hand-tuned circuits

Nowadays a circuit placement algorithm often faces a problem size of multi-million objects, with excessive fixed blockages. We present a new efficient quadratic placement algorithm that scales well to the large-scale circuit placement problems. Instead of adding spreading forces into the quadratic system to push cells away from congestions, we simulate the cell diffusion process and use dummy anchor cellsto guide the real cell movements. By using the concept of anchor cells, our placer has explicit control on how and where to move cells. Meanwhile, the usage of anchor cells significantly reduces the complexity of the Hessian of the quadratic system, which encapsulates the netlist conductivity. The Hessian matrix in our quadratic placer is extremely sparse, with 2-3 nonzero entries in most of rows/colums. Such a matrix is trivial to solve. From our experiments on the large-scale ISPD 2005 placement benchmarks, our placer achieved a 23x speedup on the CPU usage of the quadratic solver, and comparable wirelengths to existing state-of-art placers.

Placement migration is a critical step to address a variety of postplacement design closure issues, such as timing, routing congestion, signal integrity, and heat distribution. To fix a design problem, one would like to perturb the design as little as possible while preserving the integrity of the original placement. This work presents a novel computational geometry based placement migration method, and a new stability metric to more accurately measure the "similarity" between two placements. It has two stages, a bin-based spreading, at coarse scale and a Delaunay triangulation based spreading at finer grain. It has clear advantage over conventional legalization algorithms such that the neighborhood characteristics of the, original placement are preserved. Thus, the placement migration is much more stable, which is important to maintain. Applying this technique to placement legalization demonstrates significant improvements in wire length and stability compared to other popular legalization algorithms.

We describe an efficient framework for Web personalization based on sequential and non-sequential pattern discovery from usage data. Our experimental results performed on real usage data indicate that more restrictive patterns, such as contiguous sequential patterns (e.g., frequent navigational paths) are more suitable for predictive tasks, such as Web prefetching, (which involve predicting which item is accessed next by a user), while less constrained patterns, such as frequent item sets or general sequential patterns are more effective alternatives in the context of Web personalization and recommender systems.

Web usage mining, possibly used in conjunction with standard approaches to personalization such as collaborative filtering, can help address some of the shortcomings of these techniques, including reliance on subjective user ratings, lack of scalability, and poor performance in the face of high-dimensional and sparse data. However, the discovery of patterns from usage data by itself is not sufficient for performing the personalization tasks. The critical step is the effective derivation of good quality and useful (i.e., actionable) “aggregate usage profiles” from these patterns. In this paper we present and experimentally evaluate two techniques, based on clustering of user transactions and clustering of pageviews, in order to discover overlapping aggregate profiles that can be effectively used by recommender systems for real-time Web personalization. We evaluate these techniques both in terms of the quality of the individual profiles generated, as well as in the context of providing recommendations as an integrated part of a personalization engine. In particular, our results indicate that using the generated aggregate profiles, we can achieve effective personalization at early stages of users' visits to a site, based only on anonymous clickstream data and without the benefit of explicit input by these users or deeper knowledge about them.

To engage visitors to a Web site at a very early stage (i.e., before registration or authentication), personalization tools must rely primarily on clickstream data captured in Web server logs. The lack of explicit user ratings as well as the sparse nature and the large volume of data in such a setting poses serious challenges to standard collaborative filtering techniques in terms of scalability and performance. Web usage mining techniques such as clustering that rely on offline pattern discovery from user transactions can be used to improve the scalability of collaborative filtering, however, this is often at the cost of reduced recommendation accuracy. In this paper we propose effective and scalable techniques for Web personalization based on association rule discovery from usage data. Through detailed experimental evaluation on real usage data, we show that the proposed methodology can achieve better recommendation effectiveness, while maintaining a computational advantage over direct approaches to collaborative filtering such as the k -nearest-neighbor strategy.

Recommender systems based on collaborative filtering usually require real-time comparison of users’ ratings on objects. In the context of Web personalization, particularly at the early stages of a visitor’s interaction with the site (i.e., before registration or authentication), recommender systems must rely on anonymous clickstream data. The lack of explicit user ratings and the shear amount of data in such a setting poses serious challenges to standard collaborative filtering techniques in terms of scalability and performance. Offline clustering of users transactions can be used to improve the scalability of collaborative filtering, however, this is often at the cost of reduced recommendation accuracy. In this paper we study the impact of various preprocessing techniques applied to clickstream data, such as clustering, normalization, and significance filtering, on collaborative filtering. Our experimental results, performed on real usage data, indicate that with proper data preparation, the clustering-based approach to collaborative filtering can achieve dramatic improvements in terms of recommendation effectiveness, while maintaining the computational advantage over the direct approaches such as the k-NearestNeighbor technique.

Recent proposals have suggested Web usage mining as an enabling mechanism to overcome the problems associated with more traditional Web personalization techniques such as collaborative or contentbased filtering. These problems include lack of scalability, reliance on subjective user ratings or static profiles, and the inability to capture a richer set of semantic relationships among objects (in content-based systems). Yet, usage-based personalization can be problematic when little usage data is available pertaining to some objects or when the site content changes regularly. For more effective personalization, both usage and content attributes of a site must be integrated into a Web mining framework and used by the recommendation engine in a uniform manner. In this paper we present such a framework, distinguishing between the offline tasks of data preparation and mining, and the online process of customizing Web pages based on a user’s active session. We describe effective techniques based on clustering to obtain a uniform representation for both site usage and site content profiles, and we show how these profiles can be used to perform real-time personalization.

Web usage mining, possibly used in conjunction with standard approaches to personalization such as collaborative filtering, can help address some of the shortcomings of these techniques, including reliance on subjective user ratings, lack of scalability, and poor performance in the face high- dimensional and sparse data. However, the discovery of patterns from usage data by itself is not sufficient for performing the personalization tasks. The critical step is the effective derivation of good quality and useful (i.e., actionable) "aggregate usage profiles" from these patterns. In this paper we present and experimentally evaluate two techniques, based on clustering of user transactions and clustering of pageviews, in order to discover overlapping aggregate profiles that can be effectively used by recommender systems for real-time personalization. We evaluate these techniques both in terms of the quality of the individual profiles generated, as well as in the context of providing recommendations as an integrated part of a personalization engine.