Young-Joon Lee

Bio: Young-Joon Lee is an academic researcher from Google. The author has contributed to research in topics: Three-dimensional integrated circuit & Integrated circuit design. The author has an hindex of 22, co-authored 42 publications receiving 1644 citations. Previous affiliations of Young-Joon Lee include Seoul National University & Daegu Haany University.

Titanium dioxide nanoparticles trigger p53-mediated damage response in peripheral blood lymphocytes.

Abstract: Titanium dioxide nanoparticles (nano-TiO2) are widely used as a photocatalyst in air and water remediation. These nanoparticles are known to induce toxicity; however, their cytotoxic mechanism is not fully understood. In this study, we investigated the underlying mechanism of nano-TiO2-induced cytotoxicity in peripheral blood lymphocytes. We examined the genotoxic effects of nano-TiO2 in lymphocytes using alkaline single-cell gel electrophoresis (Comet) and cytokinesis-block micronucleus (CBMN) assays. Lymphocytes treated with nano-TiO2 showed significantly increased micronucleus formation and DNA breakage. Western-blot analysis to identify proteins involved in the p53-mediated response to DNA damage revealed the accumulation of p53 and activation of DNA damage checkpoint kinases in nano-TiO2-treated lymphocytes. However, p21 and bax, downstream targets of p53, were not affected, indicating that nano-TiO2 does not stimulate transactivational activity of p53. The generation of reactive oxygen species (ROS) in nano-TiO2-treated cells was also observed, andN-acetylcysteine (NAC) supplementation inhibited the level of nano-TiO2-induced DNA damage. Given that ROS-induced DNA damage leads to p53 activation in the DNA damage response, our results suggest that nano-TiO2 induces ROS generation in lymphocytes, thereby activating p53-mediated DNA damage checkpoint signals.

3D-MAPS: 3D Massively parallel processor with stacked memory

Abstract: Several recent works have demonstrated the benefits of through-silicon-via (TSV) based 3D integration [1–4], but none of them involves a fully functioning multicore processor and memory stacking. 3D-MAPS (3D Massively Parallel Processor with Stacked Memory) is a two-tier 3D IC, where the logic die consists of 64 general-purpose processor cores running at 277MHz, and the memory die contains 256KB SRAM (see Fig. 10.6.1). Fabrication is done using 130nm GlobalFoundries device technology and Tezzaron TSV and bonding technology. Packaging is done by Amkor. This processor contains 33M transistors, 50K TSVs, and 50K face-to-face connections in 5×5mm2 footprint. The chip runs at 1.5V and consumes up to 4W, resulting in 16W/cm2 power density. The core architecture is developed from scratch to benefit from single-cycle access to SRAM.

Chip Placement with Deep Reinforcement Learning

Abstract: In this work, we present a learning-based approach to chip placement, one of the most complex and time-consuming stages of the chip design process. Unlike prior methods, our approach has the ability to learn from past experience and improve over time. In particular, as we train over a greater number of chip blocks, our method becomes better at rapidly generating optimized placements for previously unseen chip blocks. To achieve these results, we pose placement as a Reinforcement Learning (RL) problem and train an agent to place the nodes of a chip netlist onto a chip canvas. To enable our RL policy to generalize to unseen blocks, we ground representation learning in the supervised task of predicting placement quality. By designing a neural architecture that can accurately predict reward across a wide variety of netlists and their placements, we are able to generate rich feature embeddings of the input netlists. We then use this architecture as the encoder of our policy and value networks to enable transfer learning. Our objective is to minimize PPA (power, performance, and area), and we show that, in under 6 hours, our method can generate placements that are superhuman or comparable on modern accelerator netlists, whereas existing baselines require human experts in the loop and take several weeks.

A graph placement methodology for fast chip design

Abstract: Chip floorplanning is the engineering task of designing the physical layout of a computer chip. Despite five decades of research1, chip floorplanning has defied automation, requiring months of intense effort by physical design engineers to produce manufacturable layouts. Here we present a deep reinforcement learning approach to chip floorplanning. In under six hours, our method automatically generates chip floorplans that are superior or comparable to those produced by humans in all key metrics, including power consumption, performance and chip area. To achieve this, we pose chip floorplanning as a reinforcement learning problem, and develop an edge-based graph convolutional neural network architecture capable of learning rich and transferable representations of the chip. As a result, our method utilizes past experience to become better and faster at solving new instances of the problem, allowing chip design to be performed by artificial agents with more experience than any human designer. Our method was used to design the next generation of Google’s artificial intelligence (AI) accelerators, and has the potential to save thousands of hours of human effort for each new generation. Finally, we believe that more powerful AI-designed hardware will fuel advances in AI, creating a symbiotic relationship between the two fields. Machine learning tools are used to greatly accelerate chip layout design, by posing chip floorplanning as a reinforcement learning problem and using neural networks to generate high-performance chip layouts.

TSV stress aware timing analysis with applications to 3D-IC layout optimization

Abstract: As the geometry shrinking faces severe limitations, 3D wafer stacking with through silicon via (TSV) has gained interest for future SOC integration. Since TSV fill material and silicon have different coefficients of thermal expansion (CTE), TSV causes silicon deformation due to different temperatures at chip manufacturing and operating. The widely used TSV fill material is copper which causes tensile stress on silicon near TSV. In this paper, we propose systematic TSV stress aware timing analysis and show how to optimize layout for better performance. First, we generate a stress contour map with an analytical radial stress model. Then, the tensile stress is converted to hole and electron mobility variations depending on geometric relation between TSVs and transistors. Mobility variation aware cell library and netlist are generated and incorporated in an industrial timing engine for 3D-IC timing analysis. It is interesting to observe that rise and fall time react differently to stress and relative locations with respect to TSVs. Overall, TSV stress induced timing variations can be as much as ± 10% for an individual cell. Thus as an application for layout optimization, we can exploit the stress-induced mobility enhancement to improve timing on critical cells. We show that stress-aware perturbation could reduce cell delay by up to 14.0% and critical path delay by 6.5% in our test case.

Titanium dioxide nanoparticles: a review of current toxicological data

Abstract: Titanium dioxide (TiO2) nanoparticles (NPs) are manufactured worldwide in large quantities for use in a wide range of applications. TiO2 NPs possess different physicochemical properties compared to their fine particle (FP) analogs, which might alter their bioactivity. Most of the literature cited here has focused on the respiratory system, showing the importance of inhalation as the primary route for TiO2 NP exposure in the workplace. TiO2 NPs may translocate to systemic organs from the lung and gastrointestinal tract (GIT) although the rate of translocation appears low. There have also been studies focusing on other potential routes of human exposure. Oral exposure mainly occurs through food products containing TiO2 NP-additives. Most dermal exposure studies, whether in vivo or in vitro, report that TiO2 NPs do not penetrate the stratum corneum (SC). In the field of nanomedicine, intravenous injection can deliver TiO2 nanoparticulate carriers directly into the human body. Upon intravenous exposure, TiO2 NPs can induce pathological lesions of the liver, spleen, kidneys, and brain. We have also shown here that most of these effects may be due to the use of very high doses of TiO2 NPs. There is also an enormous lack of epidemiological data regarding TiO2 NPs in spite of its increased production and use. However, long-term inhalation studies in rats have reported lung tumors. This review summarizes the current knowledge on the toxicology of TiO2 NPs and points out areas where further information is needed.

Titanium dioxide nanoparticles induce DNA damage and genetic instability in vivo in mice.

Abstract: Titanium dioxide (TiO2) nanoparticles are manufactured worldwide in large quantities for use in a wide range of applications including pigment and cosmetic manufacturing. Although TiO2 is chemically inert, TiO2 nanoparticles can cause negative health effects, such as respiratory tract cancer in rats. However, the mechanisms involved in TiO2-induced genotoxicity and carcinogenicity have not been clearly defined and are poorly studied in vivo. The present study investigates TiO2 nanoparticles–induced genotoxicity, oxidative DNA damage, and inflammation in a mice model. We treated wild-type mice with TiO2 nanoparticles in drinking water and determined the extent of DNA damage using the comet assay, the micronuclei assay, and the γ-H2AX immunostaining assay and by measuring 8-hydroxy-2′-deoxyguanosine levels and, as a genetic instability endpoint, DNA deletions. We also determined mRNA levels of inflammatory cytokines in the peripheral blood. Our results show that TiO2 nanoparticles induced 8-hydroxy-2′-deoxyguanosine, γ-H2AX foci, micronuclei, and DNA deletions. The formation of γ-H2AX foci, indicative of DNA double-strand breaks, was the most sensitive parameter. Inflammation was also present as characterized by a moderate inflammatory response. Together, these results describe the first comprehensive study of TiO2 nanoparticles–induced genotoxicity in vivo in mice possibly caused by a secondary genotoxic mechanism associated with inflammation and/or oxidative stress. Given the growing use of TiO2 nanoparticles, these findings raise concern about potential health hazards associated with TiO2 nanoparticles exposure. [Cancer Res

Nanomaterial cytotoxicity is composition, size, and cell type dependent

Abstract: Despite intensive research efforts, reports of cellular responses to nanomaterials are often inconsistent and even contradictory. Additionally, relationships between the responding cell type and nanomaterial properties are not well understood. Using three model cell lines representing different physiological compartments and nanomaterials of different compositions and sizes, we have systematically investigated the influence of nanomaterial properties on the degrees and pathways of cytotoxicity. In this study, we selected nanomaterials of different compositions (TiO2 and SiO2 nanoparticles, and multi-wall carbon nanotubes [MWCNTs]) with differing size (MWCNTs of different diameters 50 nm; but same length 0.5-2 μm) to analyze the effects of composition and size on toxicity to 3T3 fibroblasts, RAW 264.7 macrophages, and telomerase-immortalized (hT) bronchiolar epithelial cells. Following characterization of nanomaterial properties in PBS and serum containing solutions, cells were exposed to nanomaterials of differing compositions and sizes, with cytotoxicity monitored through reduction in mitochondrial activity. In addition to cytotoxicity, the cellular response to nanomaterials was characterized by quantifying generation of reactive oxygen species, lysosomal membrane destabilization and mitochondrial permeability. The effect of these responses on cellular fate - apoptosis or necrosis - was then analyzed. Nanomaterial toxicity was variable based on exposed cell type and dependent on nanomaterial composition and size. In addition, nanomaterial exposure led to cell type dependent intracellular responses resulting in unique breakdown of cellular functions for each nanomaterial: cell combination. Nanomaterials induce cell specific responses resulting in variable toxicity and subsequent cell fate based on the type of exposed cell. Our results indicate that the composition and size of nanomaterials as well as the target cell type are critical determinants of intracellular responses, degree of cytotoxicity and potential mechanisms of toxicity.