Behnam Montazeri

Bio: Behnam Montazeri is an academic researcher from Stanford University. The author has contributed to research in topics: Latency (engineering) & Laser. The author has an hindex of 8, co-authored 11 publications receiving 1152 citations. Previous affiliations of Behnam Montazeri include Google.

Demonstration of electron acceleration in a laser-driven dielectric microstructure

Abstract: Acceleration of relativistic electrons in a dielectric laser accelerator at high electric field gradients is reported, setting the stage for the development of future multi-staged accelerators of this type. Conventional particle accelerators, based on radio-frequency technology, are large-scale installations that are expensive to run. Micro-fabricated dielectric laser accelerators (DLAs) offer an attractive alternative, as they are able to support much larger accelerating fields than current accelerators, while being compact, economical and simple to manufacture using lithographic techniques. This paper presents the first experimental demonstration of a DLA capable of sustained, high-gradient (beyond 250 MeV m−1) acceleration of relativistic electrons. The results set the stage for the development of future multi-staged DLA devices composed of integrated on-chip systems, which would enable compact table-top MeV–GeV-scale accelerators. Applications include security scanners and medical therapy, X-ray light sources for biological and materials research, and portable medical imaging devices. The enormous size and cost of current state-of-the-art accelerators based on conventional radio-frequency technology has spawned great interest in the development of new acceleration concepts that are more compact and economical. Micro-fabricated dielectric laser accelerators (DLAs) are an attractive approach, because such dielectric microstructures can support accelerating fields one to two orders of magnitude higher than can radio-frequency cavity-based accelerators. DLAs use commercial lasers as a power source, which are smaller and less expensive than the radio-frequency klystrons that power today’s accelerators. In addition, DLAs are fabricated via low-cost, lithographic techniques that can be used for mass production. However, despite several DLA structures having been proposed recently1,2,3,4, no successful demonstration of acceleration in these structures has so far been shown. Here we report high-gradient (beyond 250 MeV m−1) acceleration of electrons in a DLA. Relativistic (60-MeV) electrons are energy-modulated over 563 ± 104 optical periods of a fused silica grating structure, powered by a 800-nm-wavelength mode-locked Ti:sapphire laser. The observed results are in agreement with analytical models and electrodynamic simulations. By comparison, conventional modern linear accelerators operate at gradients of 10–30 MeV m−1, and the first linear radio-frequency cavity accelerator was ten radio-frequency periods (one metre) long with a gradient of approximately 1.6 MeV m−1 (ref. 5). Our results set the stage for the development of future multi-staged DLA devices composed of integrated on-chip systems. This would enable compact table-top accelerators on the MeV–GeV (106–109 eV) scale for security scanners and medical therapy, university-scale X-ray light sources for biological and materials research, and portable medical imaging devices, and would substantially reduce the size and cost of a future collider on the multi-TeV (1012 eV) scale.

Dielectric laser accelerators

Abstract: The use of infrared lasers to power optical-scale lithographically fabricated particle accelerators is a developing area of research that has garnered increasing interest in recent years The physics and technology of this approach is reviewed, which is referred to as dielectric laser acceleration (DLA) In the DLA scheme operating at typical laser pulse lengths of 01 to 1 ps, the laser damage fluences for robust dielectric materials correspond to peak surface electric fields in the $\mathrm{GV}/\mathrm{m}$ regime The corresponding accelerating field enhancement represents a potential reduction in active length of the accelerator between 1 and 2 orders of magnitude Power sources for DLA-based accelerators (lasers) are less costly than microwave sources (klystrons) for equivalent average power levels due to wider availability and private sector investment Because of the high laser-to-particle coupling efficiency, required pulse energies are consistent with tabletop microJoule class lasers Combined with the very high (MHz) repetition rates these lasers can provide, the DLA approach appears promising for a variety of applications, including future high-energy physics colliders, compact light sources, and portable medical scanners and radiative therapy machines

The RAMCloud Storage System

Abstract: RAMCloud is a storage system that provides low-latency access to large-scale datasets. To achieve low latency, RAMCloud stores all data in DRAM at all times. To support large capacities (1PB or more), it aggregates the memories of thousands of servers into a single coherent key-value store. RAMCloud ensures the durability of DRAM-based data by keeping backup copies on secondary storage. It uses a uniform log-structured mechanism to manage both DRAM and secondary storage, which results in high performance and efficient memory usage. RAMCloud uses a polling-based approach to communication, bypassing the kernel to communicate directly with NICs; with this approach, client applications can read small objects from any RAMCloud storage server in less than 5μs, durable writes of small objects take about 13.5μs. RAMCloud does not keep multiple copies of data online; instead, it provides high availability by recovering from crashes very quickly (1 to 2 seconds). RAMCloud’s crash recovery mechanism harnesses the resources of the entire cluster working concurrently so that recovery performance scales with cluster size.

Homa: a receiver-driven low-latency transport protocol using network priorities

Abstract: Homa is a new transport protocol for datacenter networks. It provides exceptionally low latency, especially for workloads with a high volume of very short messages, and it also supports large messages and high network utilization. Homa uses in-network priority queues to ensure low latency for short messages; priority allocation is managed dynamically by each receiver and integrated with a receiver-driven flow control mechanism. Homa also uses controlled overcommitment of receiver downlinks to ensure efficient bandwidth utilization at high load. Our implementation of Homa delivers 99th percentile round-trip times less than 15 μs for short messages on a 10 Gbps network running at 80% load. These latencies are almost 100x lower than the best published measurements of an implementation. In simulations, Homa's latency is roughly equal to pFabric and significantly better than pHost, PIAS, and NDP for almost all message sizes and workloads. Homa can also sustain higher network loads than pFabric, pHost, or PIAS.

Swift: Delay is Simple and Effective for Congestion Control in the Datacenter

Abstract: We report on experiences with Swift congestion control in Google datacenters. Swift targets an end-to-end delay by using AIMD control, with pacing under extreme congestion. With accurate RTT measurement and care in reasoning about delay targets, we find this design is a foundation for excellent performance when network distances are well-known. Importantly, its simplicity helps us to meet operational challenges. Delay is easy to decompose into fabric and host components to separate concerns, and effortless to deploy and maintain as a congestion signal while the datacenter evolves. In large-scale testbed experiments, Swift delivers a tail latency of

Terahertz-driven linear electron acceleration

Abstract: The cost, size and availability of electron accelerators are dominated by the achievable accelerating gradient. Conventional high-brightness radio-frequency accelerating structures operate with 30-50 MeV m(-1) gradients. Electron accelerators driven with optical or infrared sources have demonstrated accelerating gradients orders of magnitude above that achievable with conventional radio-frequency structures. However, laser-driven wakefield accelerators require intense femtosecond sources and direct laser-driven accelerators suffer from low bunch charge, sub-micron tolerances and sub-femtosecond timing requirements due to the short wavelength of operation. Here we demonstrate linear acceleration of electrons with keV energy gain using optically generated terahertz pulses. Terahertz-driven accelerating structures enable high-gradient electron/proton accelerators with simple accelerating structures, high repetition rates and significant charge per bunch. These ultra-compact terahertz accelerators with extremely short electron bunches hold great potential to have a transformative impact for free electron lasers, linear colliders, ultrafast electron diffraction, X-ray science and medical therapy with X-rays and electron beams.

Quantum coherent optical phase modulation in an ultrafast transmission electron microscope

Abstract: Coherent manipulation of quantum systems with light is expected to be a cornerstone of future information and communication technology, including quantum computation and cryptography. The transfer of an optical phase onto a quantum wavefunction is a defining aspect of coherent interactions and forms the basis of quantum state preparation, synchronization and metrology. Light-phase-modulated electron states near atoms and molecules are essential for the techniques of attosecond science, including the generation of extreme-ultraviolet pulses and orbital tomography. In contrast, the quantum-coherent phase-modulation of energetic free-electron beams has not been demonstrated, although it promises direct access to ultrafast imaging and spectroscopy with tailored electron pulses on the attosecond scale. Here we demonstrate the coherent quantum state manipulation of free-electron populations in an electron microscope beam. We employ the interaction of ultrashort electron pulses with optical near-fields to induce Rabi oscillations in the populations of electron momentum states, observed as a function of the optical driving field. Excellent agreement with the scaling of an equal-Rabi multilevel quantum ladder is obtained, representing the observation of a light-driven 'quantum walk' coherently reshaping electron density in momentum space. We note that, after the interaction, the optically generated superposition of momentum states evolves into a train of attosecond electron pulses. Our results reveal the potential of quantum control for the precision structuring of electron densities, with possible applications ranging from ultrafast electron spectroscopy and microscopy to accelerator science and free-electron lasers.

NOVA: a log-structured file system for hybrid volatile/non-volatile main memories

Abstract: Fast non-volatile memories (NVMs) will soon appear on the processor memory bus alongside DRAM. The resulting hybrid memory systems will provide software with sub-microsecond, high-bandwidth access to persistent data, but managing, accessing, and maintaining consistency for data stored in NVM raises a host of challenges. Existing file systems built for spinning or solid-state disks introduce software overheads that would obscure the performance that NVMs should provide, but proposed file systems for NVMs either incur similar overheads or fail to provide the strong consistency guarantees that applications require. We present NOVA, a file system designed to maximize performance on hybrid memory systems while providing strong consistency guarantees. NOVA adapts conventional log-structured file system techniques to exploit the fast random access that NVMs provide. In particular, it maintains separate logs for each inode to improve concurrency, and stores file data outside the log to minimize log size and reduce garbage collection costs. NOVA's logs provide metadata, data, and mmap atomicity and focus on simplicity and reliability, keeping complex metadata structures in DRAM to accelerate lookup operations. Experimental results show that in write-intensive workloads, NOVA provides 22% to 216× throughput improvement compared to state-of-the-art file systems, and 3.1× to 13.5× improvement compared to file systems that provide equally strong data consistency guarantees.

NetCache: Balancing Key-Value Stores with Fast In-Network Caching

Abstract: We present NetCache, a new key-value store architecture that leverages the power and flexibility of new-generation programmable switches to handle queries on hot items and balance the load across storage nodes. NetCache provides high aggregate throughput and low latency even under highly-skewed and rapidly-changing workloads. The core of NetCache is a packet-processing pipeline that exploits the capabilities of modern programmable switch ASICs to efficiently detect, index, cache and serve hot key-value items in the switch data plane. Additionally, our solution guarantees cache coherence with minimal overhead. We implement a NetCache prototype on Barefoot Tofino switches and commodity servers and demonstrate that a single switch can process 2+ billion queries per second for 64K items with 16-byte keys and 128-byte values, while only consuming a small portion of its hardware resources. To the best of our knowledge, this is the first time that a sophisticated application-level functionality, such as in-network caching, has been shown to run at line rate on programmable switches. Furthermore, we show that NetCache improves the throughput by 3-10x and reduces the latency of up to 40% of queries by 50%, for high-performance, in-memory key-value stores.

Demonstration of electron acceleration in a laser-driven dielectric microstructure

Abstract: Acceleration of relativistic electrons in a dielectric laser accelerator at high electric field gradients is reported, setting the stage for the development of future multi-staged accelerators of this type. Conventional particle accelerators, based on radio-frequency technology, are large-scale installations that are expensive to run. Micro-fabricated dielectric laser accelerators (DLAs) offer an attractive alternative, as they are able to support much larger accelerating fields than current accelerators, while being compact, economical and simple to manufacture using lithographic techniques. This paper presents the first experimental demonstration of a DLA capable of sustained, high-gradient (beyond 250 MeV m−1) acceleration of relativistic electrons. The results set the stage for the development of future multi-staged DLA devices composed of integrated on-chip systems, which would enable compact table-top MeV–GeV-scale accelerators. Applications include security scanners and medical therapy, X-ray light sources for biological and materials research, and portable medical imaging devices. The enormous size and cost of current state-of-the-art accelerators based on conventional radio-frequency technology has spawned great interest in the development of new acceleration concepts that are more compact and economical. Micro-fabricated dielectric laser accelerators (DLAs) are an attractive approach, because such dielectric microstructures can support accelerating fields one to two orders of magnitude higher than can radio-frequency cavity-based accelerators. DLAs use commercial lasers as a power source, which are smaller and less expensive than the radio-frequency klystrons that power today’s accelerators. In addition, DLAs are fabricated via low-cost, lithographic techniques that can be used for mass production. However, despite several DLA structures having been proposed recently1,2,3,4, no successful demonstration of acceleration in these structures has so far been shown. Here we report high-gradient (beyond 250 MeV m−1) acceleration of electrons in a DLA. Relativistic (60-MeV) electrons are energy-modulated over 563 ± 104 optical periods of a fused silica grating structure, powered by a 800-nm-wavelength mode-locked Ti:sapphire laser. The observed results are in agreement with analytical models and electrodynamic simulations. By comparison, conventional modern linear accelerators operate at gradients of 10–30 MeV m−1, and the first linear radio-frequency cavity accelerator was ten radio-frequency periods (one metre) long with a gradient of approximately 1.6 MeV m−1 (ref. 5). Our results set the stage for the development of future multi-staged DLA devices composed of integrated on-chip systems. This would enable compact table-top accelerators on the MeV–GeV (106–109 eV) scale for security scanners and medical therapy, university-scale X-ray light sources for biological and materials research, and portable medical imaging devices, and would substantially reduce the size and cost of a future collider on the multi-TeV (1012 eV) scale.