This paper presents the architecture and motivation for a clusterbased, many-core computing architecture for energy-efficient, dataintensive computing. FAWN, a Fast Array of Wimpy Nodes, consists of a large number of slower but efficient nodes coupled with low-power storage. We present the computing trends that motivate a FAWN-like approach, for CPU, memory, and storage. We follow with a set of microbenchmarks to explore under what workloads these “wimpy nodes ” perform well (or perform poorly). We conclude with an outline of the longer-term implications of FAWN that lead us to select a tightly integrated stacked chip-and-memory architecture for future FAWN development.

This paper presents a new, memory efficient and cacheoptimized algorithm for simultaneously searching for a large number of patterns in a very large corpus. This algorithm builds upon the Rabin-Karp string search algorithm and incorporates a new type of Bloom filter that we call a feed-forward Bloom filter. While it retains the asymptotic time complexity of previous multiple pattern matching algorithms, we show that this technique, along with a CPU architecture aware design of the Bloom filter, can provide speedups between 2 × and 30×, and memory consumption reductions as large as 50 × when compared with grep. 1

Abstract. In the ongoing arms race against malware, antivirus software is at the forefront, as one of the most important defense tools in our arsenal. Antivirus software is flexible enough to be deployed from regular users desktops, to corporate e-mail proxies and file servers. Unfortunately, the signatures necessary to detect incoming malware number in the tens of thousands. To make matters worse, antivirus signatures are a lot longer than signatures in network intrusion detection systems. This leads to extremely high computation costs necessary to perform matching of suspicious data against those signatures. In this paper, we present GrAVity, a massively parallel antivirus engine. Our engine utilized the compute power of modern graphics processors, that contain hundreds of hardware microprocessors. We have modified ClamAV, the most popular open source antivirus software, to utilize our engine. Our prototype implementation has achieved end-to-end throughput in the order of 20 Gbits/s, 100 times the performance of the CPUonly ClamAV, while almost completely offloading the CPU, leaving it free to complete other tasks. Our micro-benchmarks have measured our engine to be able to sustain throughput in the order of 40 Gbits/s. The results suggest that modern graphics cards can be used effectively to perform heavy-duty anti-malware operations at speeds that cannot be matched by traditional CPU based techniques. 1

This paper evaluates the prospects for protecting critical social functions from “cyber ” attacks carried out over electronic information networks. In particular, it focuses on the feasibility of devising international laws, conventions or agreements to deter and/or punish perpetrators of such attacks. First, it briefly summarizes existing conventions and laws, and explains to which technological issues they can apply. The paper then turns to a technical discussion of the threats faced by critical infrastructure. By distinguishing between the different types of attacks (theft of information, destructive penetration, denial of service, etc.) that can be conducted, and examining the role of collateral damages in information security, the paper identifies the major challenges in devising and implementing international conventions for critical infrastructure protection. It then turns to a practical examination of how these findings apply to specific instances of critical networks (power grids and water systems, financial infrastructure, air traffic control and hospital networks), and draws conclusions about potential remedies. A notable finding is that critical functions should be isolated from non-critical functions in the network to have a chance to implement viable international agreements; and that, given the difficulty in performing attack attribution, other relevant laws should be designed with the objective of reducing negative externalities that facilitate such attacks. 1

This paper presents the architecture and motivation for a clusterbased, many-core computing architecture for energy-efficient, dataintensive computing. FAWN, a Fast Array of Wimpy Nodes, consists of a large number of slower but efficient nodes coupled with low-power storage. We present the computing trends that motivate a FAWN-like approach, for CPU, memory, and storage. We follow with a set of microbenchmarks to explore under what workloads these FAWN nodes perform well (or perform poorly), and briefly examine scenarios in which both code and algorithms may need to be re-designed or optimized to perform well on an efficient platform. We conclude with an outline of the longer-term implications of FAWN that lead us to select a tightly integrated stacked chipand-memory architecture for future FAWN development.

0716287, and CCF-0964474, Intel, by gi�s from Network Appliance and Google, and through fellowships

This article presents a new, memory efficient and cache-optimized algorithm for simultaneously searching for a large number of patterns in a very large corpus. This algorithm builds upon the Rabin-Karp string search algorithm and incorporates a new type of Bloom filter that we call a feed-forward Bloom filter. While it retains the asymptotic time complexity of previous multiple pattern matching algorithms, we show that this technique, along with a CPU architecture aware design of the Bloom filter, can provide speedups between 2× and 30×, and memory consumption reductions as large as 50 × when compared with grep. Our algorithm is also well suited for implementations on GPUs: a modern GPU can search for 3 million patterns at a rate of 580 MB/s, and for 100 million patterns (a prohibitive number for traditional algorithms) at a rate of 170 MB/s.