Background: Biphasic immediate-release (IR)/extended-release (ER) hydrocodone bitartrate (HB)/ace... more Background: Biphasic immediate-release (IR)/extended-release (ER) hydrocodone bitartrate (HB)/acetaminophen (APAP) 7.5/325-mg tablets are formulated with gastroretentive ER drug delivery technology that has been associated with clinically meaningful food effects in other approved products. Two phase 1 studies evaluated potential effects of food on single-dose pharmacokinetics of IR/ER HB/APAP tablets.
More than four million Americans present to U.S. Emergency Departments (EDs) each year after moto... more More than four million Americans present to U.S. Emergency Departments (EDs) each year after motor vehicle collision (MVC). More than 90% of these individuals have acute musculoskeletal pain (MSP), and, while less than 5% have a fracture or require hospital admission, 20-30% transition to chronic MSP. Individuals with acute MSP after MVC are most often discharged with oral nonsteroidal anti-inflammatory drugs (NSAIDs) or opioid medications, however optimal medication treatments are poorly defined and the influence of acute medication choices on post-MVC pain trajectory are unknown. In this study we evaluated the effect of opioid versus NSAID medication treatment at the time of ED discharge on the presence of moderate or severe MSP (MSMSP, NRS > 3) six weeks after MVC. Data was obtained from a large prospective cohort of adult European Americans (n=948) who presented to the ED after MVC. Propensity score matched analysis was used to compare the odds of MSMSP at six week follow-up (obtained in 859/948 (91%) among patients discharged with opioid analgesics alone (198/859 (23%)) vs. those discharged with NSAIDs alone (338/859 (39%)). Participants were propensity-matched on demographic and clinical characteristics, baseline pain scores, and MVC characteristics. At 6weeks, 49% of those receiving NSAIDS and 56% of those receiving opioids reported MSMSP. After propensity matching, there was no significant difference in MSMSP at six weeks between those discharged with opioids vs. those discharged with NSAIDS (OR=0.92; 95% CI 0.33-1.51). These results suggest that initial ED provider choice of NSAIDs versus opioid medication at discharge following MVC does not influence the development of MSMSP six weeks after MVC. Supported by NIAMS R01AR056328.
Although not powered to detect differences between GEn doses, the greatest numerical difference v... more Although not powered to detect differences between GEn doses, the greatest numerical difference vs placebo using the LOCF and MMRM methods was observed with GEn 3600 mg, while this was the case for GEn 1200 mg using the BOCF method. Conclusions: Regardless of the analysis methodology, there were significant differences between all three doses of GEn and placebo with regard to the primary endpoint, confirming the validity of the primary analysis. Poster 525 Exercise Increases Cephalad CSF Flow at Lumbar Spine.
MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodon... more MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodone bitartrate (HB) 7.5 mg and acetaminophen (APAP) 325 mg (HB/APAP ER) being developed for the management of moderate to moderately severe acute pain. This study evaluated single-dose pharmacokinetics and bioavailability following administration of 3 tablets of MNK-155 (HB 7.5 mg/APAP 325 mg) under fed conditions (high-and low-fat meals) compared with fasted conditions. Healthy volunteers (n=21) received a single dose (3 tablets) of MNK-155 after a high-fat meal, after a low-fat meal, and under fasted conditions. Nausea, pruritus, vomiting, somnolence, and dizziness were the most frequently reported treatment-emergent adverse events (AEs) in this population. No serious AEs were reported. Plasma hydrocodone and APAP concentrations rose rapidly after MNK-155 administration, with plasma hydrocodone concentrations sustained above pre-dose levels over the proposed 12-hour dosing interval. APAP concentrations approximated baseline levels by 10-12 hours after dosing. The confidence intervals for the total exposure (area under the concentration-time curve [AUC]) after administration of 3 tablets of MNK-155 under fed conditions (low-fat and high-fat meal) were within the no-effect range for both hydrocodone and APAP. Peak exposure (maximum plasma concentration [C max ]) for hydrocodone following administration of MNK-155 after a high-fat meal did not differ from that under the fasted condition. The lower peak exposure for APAP under both fed conditions and the slight increase in mean peak exposure for hydrocodone after a low-fat meal are consistent with food effects on peak exposure that have been reported for other hydrocodone-and APAP-containing products. These findings support the appropriate administration of MNK-155 without regard to food.
MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodon... more MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodone bitartrate 7.5 mg and acetaminophen 325 mg (HB/ APAP ER). The single-and multiple-dose pharmacokinetics and bioavailability of MNK-155 were compared with IR HB/APAP 7.5 mg/325 mg. Treatments were administered to healthy volunteers under fasted conditions. In the single-dose portion of the study, subjects received 1 dose of MNK-155 (3 tablets) or 3 doses of IR HB/APAP (1 tablet q4h). In the multiple-dose portion, subjects received MNK-155 (2 tablets q12h with a loading dose of 3 tablets) or IR HB/ APAP (1 tablet q6h) for 4.5 days. The most commonly reported adverse events (AEs) were nausea and vomiting. No serious AEs were reported. Plasma hydrocodone and APAP concentrations rose rapidly after MNK-155 administration. Hydrocodone concentrations were sustained above pre-dose levels throughout the proposed 12-h dosing interval, and APAP levels were low by 10-12 hours after dosing. Steady-state conditions were observed in 3 days for hydrocodone and 2 days for APAP. Total exposure (dose-normalized area under the concentration-time curve from 0-12h [AUC 0-12 ]) to hydrocodone and APAP with MNK-155 were comparable after single-dose administration and at steady state (with and without loading dose). At steady state, total exposure (dose-normalized AUC) to hydrocodone and APAP and peak exposure (dose-normalized maximum plasma concentration [C max ]) to hydrocodone for MNK-155 with the loading dose were equivalent to IR HB/APAP. Although peak exposure (dose-normalized C max) at steady state for APAP was higher, average concentrations were equivalent between MNK-155 and IR HB/APAP. Consistent with earlier studies, trough APAP plasma concentrations after administration of MNK-155 with a loading dose were lower compared with IR HB/APAP. These findings support a 12-hour dosing interval for MNK-155. Supported by funding from Mallinckrodt Inc.
Safety data from two phase 3 clinical trials of MNK-795 (oxycodone and acetaminophen extended-rel... more Safety data from two phase 3 clinical trials of MNK-795 (oxycodone and acetaminophen extended-release tablets [OC/APAP ER]) were analyzed for subjects who received the proposed indicated dose of 15/650 mg OC/APAP ER q12h. One study included a 48-hour placebo-controlled, double-blind phase followed by an optional #14 day open-label phase, and the other consisted of up to 42 days of open-label treatment. Of the 701 patients in the safety population, 607 received $1 dose of OC/APAP ER 15/650 mg (mean age, 48.2 years; 68.5% female). In addition, 163 received placebo during double-blind treatment (mean age, 44.6 years; 82.8% female). Mean duration of OC/APAP ER exposure was 20.3 days (range, 1-42 days), and 435 patients (71.7%) had a mean exposure of $5 days. Mean time on placebo was 2.8 days (range, 1-3 days) during the double-blind treatment period. AEs occurred in 55.9% of patients (60.8% OC/APAP ER, 21.5% placebo). The most common AEs with OC/APAP ER were nausea (25.7%), dizziness (13.0%), and vomiting (12.9%), whereas the most common AEs with placebo were nausea (5.5%) and headache (4.9%). Most AEs were mild or moderate in severity. Five patients (0.8%) who received OC/ APAP ER (none on placebo) experienced an increase in alanine transaminase >3 times the upper limit of the normal range, and 6 patients (1%) discontinued the study due to LFT abnormalities. However, no subjects experienced LFT abnormalities that met the Hy's Law criteria for drug-induced liver injury. These results support the use of OC/APAP ER (15/650 mg administered q12h) for the management of acute pain when an opioid analgesic is appropriate. Supported by funding from Mallinckrodt Inc.
MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodon... more MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodone bitartrate (HB) 7.5 mg and acetaminophen (APAP) 325 mg (HB/APAP ER) being developed for the management of moderate to moderately severe acute pain. This study evaluated single-dose pharmacokinetics and bioavailability following administration of 3 tablets of MNK-155 (HB 7.5 mg/APAP 325 mg) under fed conditions (high-and low-fat meals) compared with fasted conditions. Healthy volunteers (n=21) received a single dose (3 tablets) of MNK-155 after a high-fat meal, after a low-fat meal, and under fasted conditions. Nausea, pruritus, vomiting, somnolence, and dizziness were the most frequently reported treatment-emergent adverse events (AEs) in this population. No serious AEs were reported. Plasma hydrocodone and APAP concentrations rose rapidly after MNK-155 administration, with plasma hydrocodone concentrations sustained above pre-dose levels over the proposed 12-hour dosing interval. APAP concentrations approximated baseline levels by 10-12 hours after dosing. The confidence intervals for the total exposure (area under the concentration-time curve [AUC]) after administration of 3 tablets of MNK-155 under fed conditions (low-fat and high-fat meal) were within the no-effect range for both hydrocodone and APAP. Peak exposure (maximum plasma concentration [C max ]) for hydrocodone following administration of MNK-155 after a high-fat meal did not differ from that under the fasted condition. The lower peak exposure for APAP under both fed conditions and the slight increase in mean peak exposure for hydrocodone after a low-fat meal are consistent with food effects on peak exposure that have been reported for other hydrocodone-and APAP-containing products. These findings support the appropriate administration of MNK-155 without regard to food.
Proceedings of the 1st ACM/EIGSCC Symposium on Smart Cities and Communities - SCC '18, 2018
From the time and money lost sitting in congestion and waiting for traffic signals to change, to ... more From the time and money lost sitting in congestion and waiting for traffic signals to change, to the many people injured and killed in traffic crashes each year, to the emissions and energy consumption from our vehicles, the effects of transportation on our daily lives are immense. A wealth of transportation data is available to help address these problems; from data from sensors installed to monitor and operate the roadways and traffic signals to data from cell phone apps and-just over the horizon-data from connected vehicles and infrastructure. However, this wealth of data has yet to be effectively leveraged, thus providing opportunities in areas such as improving traffic safety, reducing congestion, improving traffic signal timing, personalizing routing, coordinating across transportation agencies and more. This paper presents opportunities and challenges in applying data management technology to the transportation domain.
Abstract: Despite the promising performance improvement observed in emerging many-core architectu... more Abstract: Despite the promising performance improvement observed in emerging many-core architectures in high performance processors, high power consumption prohibitively affects their use and marketability in the low-energy sectors, such as embedded processors, network processors and application specific instruction processors (ASIPs). While most chip architects design power-efficient processors by finding an optimal power-performance balance in their design, some use sophisticated on-chip autonomous power management units, which dynamically reduce the voltage or frequencies of idle cores and hence extend battery life and reduce operating costs. For large scale designs of many-core processors, a holistic approach integrating both these techniques at different levels of abstraction can potentially achieve maximal power savings. In this paper we present CASPER, a robust instruction trace driven cycle-accurate many-core multi-threading micro-architecture simulation platform where we ha...
While medium- and large-sized computing centers have increasingly relied on clusters of commodity... more While medium- and large-sized computing centers have increasingly relied on clusters of commodity PC hardware to provide cost-effective capacity and capability, it is not clear that this technology will scale to the PetaFLOP range. It is expected that semiconductor technology will continue its exponential advancements over next fifteen years; how-ever, new issues are rapidly emerging and the relative im-portance of current performance metrics are shifting. Fu-ture PetaFLOP architectures will require system designers to solve computer architecture problems ranging from how to house, power, and cool the machine, all-the-while re-maining sensitive to cost. The Reconfigurable Computing Cluster (RCC) project is a multi-institution, multi-disciplinary project investigating the use of FPGAs to build cost-effective petascale comput-ers. This paper describes the nascent project’s objectives and a 64-node prototype cluster. Specifically, the aim is to provide an detailed motivation for the pr...
In this work, we quantize a trained Transformer machine language translation model leveraging INT... more In this work, we quantize a trained Transformer machine language translation model leveraging INT8/VNNI instructions in the latest Intel$^\\circledR$ Xeon$^\\circledR$ Cascade Lake processors to improve inference performance while maintaining less than 0.5$\\%$ drop in accuracy. To the best of our knowledge, this is the first attempt in the industry to quantize the Transformer model. This has high impact as it clearly demonstrates the various complexities of quantizing the language translation model. We present novel quantization techniques directly in TensorFlow to opportunistically replace 32-bit floating point (FP32) computations with 8-bit integers (INT8) and transform the FP32 computational graph. We also present a bin-packing parallel batching technique to maximize CPU utilization. Overall, our optimizations with INT8/VNNI deliver 1.5X improvement over the best FP32 performance. Furthermore, it reveals the opportunities and challenges to boost performance of quantized deep learni...
Existing approaches to train neural networks that use large images require to either crop or down... more Existing approaches to train neural networks that use large images require to either crop or down-sample data during pre-processing, use small batch sizes, or split the model across devices mainly due to the prohibitively limited memory capacity available on GPUs and emerging accelerators. These techniques often lead to longer time to convergence or time to train (TTT), and in some cases, lower model accuracy. CPUs, on the other hand, can leverage significant amounts of memory. While much work has been done on parallelizing neural network training on multiple CPUs, little attention has been given to tune neural network training with large images on CPUs. In this work, we train a multiscale convolutional neural network (M-CNN) to classify large biomedical images for high content screening in one hour. The ability to leverage large memory capacity on CPUs enables us to scale to larger batch sizes without having to crop or down-sample the input images. In conjunction with large batch sizes, we find a generalized methodology of linearly scaling of learning rate and train M-CNN to state-of-the-art (SOTA) accuracy of 99% within one hour. We achieve fast time to convergence using 128 two socket Intel Xeon 6148 processor nodes with 192GB DDR4 memory connected with 100Gbps Intel Omnipath architecture.
We present a novel storage manager for multi-dimensional arrays that arise in scientific applicat... more We present a novel storage manager for multi-dimensional arrays that arise in scientific applications, which is part of a larger scientific data management system called TileDB. In contrast to existing solutions, TileDB is optimized for both dense and sparse arrays. Its key idea is to organize array elements into ordered collections called fragments. Each fragment is dense or sparse, and groups contiguous array elements into data tiles of fixed capacity. The organization into fragments turns random writes into sequential writes, and, coupled with a novel read algorithm, leads to very efficient reads. TileDB enables parallelization via multi-threading and multiprocessing, offering thread-/process-safety and atomicity via lightweight locking. We show that TileDB delivers comparable performance to the HDF5 dense array storage manager, while providing much faster random writes. We also show that TileDB offers substantially faster reads and writes than the SciDB array database system with both dense and sparse arrays. Finally, we demonstrate that TileDB is considerably faster than adaptations of the Vertica relational column-store for dense array storage management, and at least as fast for the case of sparse arrays.
An efficient design space exploration framework to optimize power-efficient heterogeneous many-core multi-threading embedded processor architectures
Abstract: By the middle of this decade, uniprocessor architecture performance had hit a roadblock... more Abstract: By the middle of this decade, uniprocessor architecture performance had hit a roadblock due to a combination of factors, such as excessive power dissipation due to high operating frequencies, growing memory access latencies, diminishing returns on deeper ...
Background: Biphasic immediate-release (IR)/extended-release (ER) hydrocodone bitartrate (HB)/ace... more Background: Biphasic immediate-release (IR)/extended-release (ER) hydrocodone bitartrate (HB)/acetaminophen (APAP) 7.5/325-mg tablets are formulated with gastroretentive ER drug delivery technology that has been associated with clinically meaningful food effects in other approved products. Two phase 1 studies evaluated potential effects of food on single-dose pharmacokinetics of IR/ER HB/APAP tablets.
More than four million Americans present to U.S. Emergency Departments (EDs) each year after moto... more More than four million Americans present to U.S. Emergency Departments (EDs) each year after motor vehicle collision (MVC). More than 90% of these individuals have acute musculoskeletal pain (MSP), and, while less than 5% have a fracture or require hospital admission, 20-30% transition to chronic MSP. Individuals with acute MSP after MVC are most often discharged with oral nonsteroidal anti-inflammatory drugs (NSAIDs) or opioid medications, however optimal medication treatments are poorly defined and the influence of acute medication choices on post-MVC pain trajectory are unknown. In this study we evaluated the effect of opioid versus NSAID medication treatment at the time of ED discharge on the presence of moderate or severe MSP (MSMSP, NRS > 3) six weeks after MVC. Data was obtained from a large prospective cohort of adult European Americans (n=948) who presented to the ED after MVC. Propensity score matched analysis was used to compare the odds of MSMSP at six week follow-up (obtained in 859/948 (91%) among patients discharged with opioid analgesics alone (198/859 (23%)) vs. those discharged with NSAIDs alone (338/859 (39%)). Participants were propensity-matched on demographic and clinical characteristics, baseline pain scores, and MVC characteristics. At 6weeks, 49% of those receiving NSAIDS and 56% of those receiving opioids reported MSMSP. After propensity matching, there was no significant difference in MSMSP at six weeks between those discharged with opioids vs. those discharged with NSAIDS (OR=0.92; 95% CI 0.33-1.51). These results suggest that initial ED provider choice of NSAIDs versus opioid medication at discharge following MVC does not influence the development of MSMSP six weeks after MVC. Supported by NIAMS R01AR056328.
Although not powered to detect differences between GEn doses, the greatest numerical difference v... more Although not powered to detect differences between GEn doses, the greatest numerical difference vs placebo using the LOCF and MMRM methods was observed with GEn 3600 mg, while this was the case for GEn 1200 mg using the BOCF method. Conclusions: Regardless of the analysis methodology, there were significant differences between all three doses of GEn and placebo with regard to the primary endpoint, confirming the validity of the primary analysis. Poster 525 Exercise Increases Cephalad CSF Flow at Lumbar Spine.
MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodon... more MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodone bitartrate (HB) 7.5 mg and acetaminophen (APAP) 325 mg (HB/APAP ER) being developed for the management of moderate to moderately severe acute pain. This study evaluated single-dose pharmacokinetics and bioavailability following administration of 3 tablets of MNK-155 (HB 7.5 mg/APAP 325 mg) under fed conditions (high-and low-fat meals) compared with fasted conditions. Healthy volunteers (n=21) received a single dose (3 tablets) of MNK-155 after a high-fat meal, after a low-fat meal, and under fasted conditions. Nausea, pruritus, vomiting, somnolence, and dizziness were the most frequently reported treatment-emergent adverse events (AEs) in this population. No serious AEs were reported. Plasma hydrocodone and APAP concentrations rose rapidly after MNK-155 administration, with plasma hydrocodone concentrations sustained above pre-dose levels over the proposed 12-hour dosing interval. APAP concentrations approximated baseline levels by 10-12 hours after dosing. The confidence intervals for the total exposure (area under the concentration-time curve [AUC]) after administration of 3 tablets of MNK-155 under fed conditions (low-fat and high-fat meal) were within the no-effect range for both hydrocodone and APAP. Peak exposure (maximum plasma concentration [C max ]) for hydrocodone following administration of MNK-155 after a high-fat meal did not differ from that under the fasted condition. The lower peak exposure for APAP under both fed conditions and the slight increase in mean peak exposure for hydrocodone after a low-fat meal are consistent with food effects on peak exposure that have been reported for other hydrocodone-and APAP-containing products. These findings support the appropriate administration of MNK-155 without regard to food.
MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodon... more MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodone bitartrate 7.5 mg and acetaminophen 325 mg (HB/ APAP ER). The single-and multiple-dose pharmacokinetics and bioavailability of MNK-155 were compared with IR HB/APAP 7.5 mg/325 mg. Treatments were administered to healthy volunteers under fasted conditions. In the single-dose portion of the study, subjects received 1 dose of MNK-155 (3 tablets) or 3 doses of IR HB/APAP (1 tablet q4h). In the multiple-dose portion, subjects received MNK-155 (2 tablets q12h with a loading dose of 3 tablets) or IR HB/ APAP (1 tablet q6h) for 4.5 days. The most commonly reported adverse events (AEs) were nausea and vomiting. No serious AEs were reported. Plasma hydrocodone and APAP concentrations rose rapidly after MNK-155 administration. Hydrocodone concentrations were sustained above pre-dose levels throughout the proposed 12-h dosing interval, and APAP levels were low by 10-12 hours after dosing. Steady-state conditions were observed in 3 days for hydrocodone and 2 days for APAP. Total exposure (dose-normalized area under the concentration-time curve from 0-12h [AUC 0-12 ]) to hydrocodone and APAP with MNK-155 were comparable after single-dose administration and at steady state (with and without loading dose). At steady state, total exposure (dose-normalized AUC) to hydrocodone and APAP and peak exposure (dose-normalized maximum plasma concentration [C max ]) to hydrocodone for MNK-155 with the loading dose were equivalent to IR HB/APAP. Although peak exposure (dose-normalized C max) at steady state for APAP was higher, average concentrations were equivalent between MNK-155 and IR HB/APAP. Consistent with earlier studies, trough APAP plasma concentrations after administration of MNK-155 with a loading dose were lower compared with IR HB/APAP. These findings support a 12-hour dosing interval for MNK-155. Supported by funding from Mallinckrodt Inc.
Safety data from two phase 3 clinical trials of MNK-795 (oxycodone and acetaminophen extended-rel... more Safety data from two phase 3 clinical trials of MNK-795 (oxycodone and acetaminophen extended-release tablets [OC/APAP ER]) were analyzed for subjects who received the proposed indicated dose of 15/650 mg OC/APAP ER q12h. One study included a 48-hour placebo-controlled, double-blind phase followed by an optional #14 day open-label phase, and the other consisted of up to 42 days of open-label treatment. Of the 701 patients in the safety population, 607 received $1 dose of OC/APAP ER 15/650 mg (mean age, 48.2 years; 68.5% female). In addition, 163 received placebo during double-blind treatment (mean age, 44.6 years; 82.8% female). Mean duration of OC/APAP ER exposure was 20.3 days (range, 1-42 days), and 435 patients (71.7%) had a mean exposure of $5 days. Mean time on placebo was 2.8 days (range, 1-3 days) during the double-blind treatment period. AEs occurred in 55.9% of patients (60.8% OC/APAP ER, 21.5% placebo). The most common AEs with OC/APAP ER were nausea (25.7%), dizziness (13.0%), and vomiting (12.9%), whereas the most common AEs with placebo were nausea (5.5%) and headache (4.9%). Most AEs were mild or moderate in severity. Five patients (0.8%) who received OC/ APAP ER (none on placebo) experienced an increase in alanine transaminase >3 times the upper limit of the normal range, and 6 patients (1%) discontinued the study due to LFT abnormalities. However, no subjects experienced LFT abnormalities that met the Hy's Law criteria for drug-induced liver injury. These results support the use of OC/APAP ER (15/650 mg administered q12h) for the management of acute pain when an opioid analgesic is appropriate. Supported by funding from Mallinckrodt Inc.
MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodon... more MNK-155 is a bilayer tablet formulation of immediate-release (IR)/extendedrelease (ER) hydrocodone bitartrate (HB) 7.5 mg and acetaminophen (APAP) 325 mg (HB/APAP ER) being developed for the management of moderate to moderately severe acute pain. This study evaluated single-dose pharmacokinetics and bioavailability following administration of 3 tablets of MNK-155 (HB 7.5 mg/APAP 325 mg) under fed conditions (high-and low-fat meals) compared with fasted conditions. Healthy volunteers (n=21) received a single dose (3 tablets) of MNK-155 after a high-fat meal, after a low-fat meal, and under fasted conditions. Nausea, pruritus, vomiting, somnolence, and dizziness were the most frequently reported treatment-emergent adverse events (AEs) in this population. No serious AEs were reported. Plasma hydrocodone and APAP concentrations rose rapidly after MNK-155 administration, with plasma hydrocodone concentrations sustained above pre-dose levels over the proposed 12-hour dosing interval. APAP concentrations approximated baseline levels by 10-12 hours after dosing. The confidence intervals for the total exposure (area under the concentration-time curve [AUC]) after administration of 3 tablets of MNK-155 under fed conditions (low-fat and high-fat meal) were within the no-effect range for both hydrocodone and APAP. Peak exposure (maximum plasma concentration [C max ]) for hydrocodone following administration of MNK-155 after a high-fat meal did not differ from that under the fasted condition. The lower peak exposure for APAP under both fed conditions and the slight increase in mean peak exposure for hydrocodone after a low-fat meal are consistent with food effects on peak exposure that have been reported for other hydrocodone-and APAP-containing products. These findings support the appropriate administration of MNK-155 without regard to food.
Proceedings of the 1st ACM/EIGSCC Symposium on Smart Cities and Communities - SCC '18, 2018
From the time and money lost sitting in congestion and waiting for traffic signals to change, to ... more From the time and money lost sitting in congestion and waiting for traffic signals to change, to the many people injured and killed in traffic crashes each year, to the emissions and energy consumption from our vehicles, the effects of transportation on our daily lives are immense. A wealth of transportation data is available to help address these problems; from data from sensors installed to monitor and operate the roadways and traffic signals to data from cell phone apps and-just over the horizon-data from connected vehicles and infrastructure. However, this wealth of data has yet to be effectively leveraged, thus providing opportunities in areas such as improving traffic safety, reducing congestion, improving traffic signal timing, personalizing routing, coordinating across transportation agencies and more. This paper presents opportunities and challenges in applying data management technology to the transportation domain.
Abstract: Despite the promising performance improvement observed in emerging many-core architectu... more Abstract: Despite the promising performance improvement observed in emerging many-core architectures in high performance processors, high power consumption prohibitively affects their use and marketability in the low-energy sectors, such as embedded processors, network processors and application specific instruction processors (ASIPs). While most chip architects design power-efficient processors by finding an optimal power-performance balance in their design, some use sophisticated on-chip autonomous power management units, which dynamically reduce the voltage or frequencies of idle cores and hence extend battery life and reduce operating costs. For large scale designs of many-core processors, a holistic approach integrating both these techniques at different levels of abstraction can potentially achieve maximal power savings. In this paper we present CASPER, a robust instruction trace driven cycle-accurate many-core multi-threading micro-architecture simulation platform where we ha...
While medium- and large-sized computing centers have increasingly relied on clusters of commodity... more While medium- and large-sized computing centers have increasingly relied on clusters of commodity PC hardware to provide cost-effective capacity and capability, it is not clear that this technology will scale to the PetaFLOP range. It is expected that semiconductor technology will continue its exponential advancements over next fifteen years; how-ever, new issues are rapidly emerging and the relative im-portance of current performance metrics are shifting. Fu-ture PetaFLOP architectures will require system designers to solve computer architecture problems ranging from how to house, power, and cool the machine, all-the-while re-maining sensitive to cost. The Reconfigurable Computing Cluster (RCC) project is a multi-institution, multi-disciplinary project investigating the use of FPGAs to build cost-effective petascale comput-ers. This paper describes the nascent project’s objectives and a 64-node prototype cluster. Specifically, the aim is to provide an detailed motivation for the pr...
In this work, we quantize a trained Transformer machine language translation model leveraging INT... more In this work, we quantize a trained Transformer machine language translation model leveraging INT8/VNNI instructions in the latest Intel$^\\circledR$ Xeon$^\\circledR$ Cascade Lake processors to improve inference performance while maintaining less than 0.5$\\%$ drop in accuracy. To the best of our knowledge, this is the first attempt in the industry to quantize the Transformer model. This has high impact as it clearly demonstrates the various complexities of quantizing the language translation model. We present novel quantization techniques directly in TensorFlow to opportunistically replace 32-bit floating point (FP32) computations with 8-bit integers (INT8) and transform the FP32 computational graph. We also present a bin-packing parallel batching technique to maximize CPU utilization. Overall, our optimizations with INT8/VNNI deliver 1.5X improvement over the best FP32 performance. Furthermore, it reveals the opportunities and challenges to boost performance of quantized deep learni...
Existing approaches to train neural networks that use large images require to either crop or down... more Existing approaches to train neural networks that use large images require to either crop or down-sample data during pre-processing, use small batch sizes, or split the model across devices mainly due to the prohibitively limited memory capacity available on GPUs and emerging accelerators. These techniques often lead to longer time to convergence or time to train (TTT), and in some cases, lower model accuracy. CPUs, on the other hand, can leverage significant amounts of memory. While much work has been done on parallelizing neural network training on multiple CPUs, little attention has been given to tune neural network training with large images on CPUs. In this work, we train a multiscale convolutional neural network (M-CNN) to classify large biomedical images for high content screening in one hour. The ability to leverage large memory capacity on CPUs enables us to scale to larger batch sizes without having to crop or down-sample the input images. In conjunction with large batch sizes, we find a generalized methodology of linearly scaling of learning rate and train M-CNN to state-of-the-art (SOTA) accuracy of 99% within one hour. We achieve fast time to convergence using 128 two socket Intel Xeon 6148 processor nodes with 192GB DDR4 memory connected with 100Gbps Intel Omnipath architecture.
We present a novel storage manager for multi-dimensional arrays that arise in scientific applicat... more We present a novel storage manager for multi-dimensional arrays that arise in scientific applications, which is part of a larger scientific data management system called TileDB. In contrast to existing solutions, TileDB is optimized for both dense and sparse arrays. Its key idea is to organize array elements into ordered collections called fragments. Each fragment is dense or sparse, and groups contiguous array elements into data tiles of fixed capacity. The organization into fragments turns random writes into sequential writes, and, coupled with a novel read algorithm, leads to very efficient reads. TileDB enables parallelization via multi-threading and multiprocessing, offering thread-/process-safety and atomicity via lightweight locking. We show that TileDB delivers comparable performance to the HDF5 dense array storage manager, while providing much faster random writes. We also show that TileDB offers substantially faster reads and writes than the SciDB array database system with both dense and sparse arrays. Finally, we demonstrate that TileDB is considerably faster than adaptations of the Vertica relational column-store for dense array storage management, and at least as fast for the case of sparse arrays.
An efficient design space exploration framework to optimize power-efficient heterogeneous many-core multi-threading embedded processor architectures
Abstract: By the middle of this decade, uniprocessor architecture performance had hit a roadblock... more Abstract: By the middle of this decade, uniprocessor architecture performance had hit a roadblock due to a combination of factors, such as excessive power dissipation due to high operating frequencies, growing memory access latencies, diminishing returns on deeper ...
Uploads
Papers by Kushal Datta