| 1. |
- Anderson, Ian, et al.
(författare)
-
Indigenous and tribal peoples' health (The Lancet-Lowitja Institute Global Collaboration) : a population study
- 2016
-
Ingår i: The Lancet. - : Elsevier. - 0140-6736 .- 1474-547X. ; 388:10040, s. 131-157
-
Tidskriftsartikel (refereegranskat)abstract
- Background: International studies of the health of Indigenous and tribal peoples provide important public health insights. Reliable data are required for the development of policy and health services. Previous studies document poorer outcomes for Indigenous peoples compared with benchmark populations, but have been restricted in their coverage of countries or the range of health indicators. Our objective is to describe the health and social status of Indigenous and tribal peoples relative to benchmark populations from a sample of countries.Methods: Collaborators with expertise in Indigenous health data systems were identified for each country. Data were obtained for population, life expectancy at birth, infant mortality, low and high birthweight, maternal mortality, nutritional status, educational attainment, and economic status. Data sources consisted of governmental data, data from non-governmental organisations such as UNICEF, and other research. Absolute and relative differences were calculated.Findings: Our data (23 countries, 28 populations) provide evidence of poorer health and social outcomes for Indigenous peoples than for non-Indigenous populations. However, this is not uniformly the case, and the size of the rate difference varies. We document poorer outcomes for Indigenous populations for: life expectancy at birth for 16 of 18 populations with a difference greater than 1 year in 15 populations; infant mortality rate for 18 of 19 populations with a rate difference greater than one per 1000 livebirths in 16 populations; maternal mortality in ten populations; low birthweight with the rate difference greater than 2% in three populations; high birthweight with the rate difference greater than 2% in one population; child malnutrition for ten of 16 populations with a difference greater than 10% in five populations; child obesity for eight of 12 populations with a difference greater than 5% in four populations; adult obesity for seven of 13 populations with a difference greater than 10% in four populations; educational attainment for 26 of 27 populations with a difference greater than 1% in 24 populations; and economic status for 15 of 18 populations with a difference greater than 1% in 14 populations.Interpretation: We systematically collated data across a broader sample of countries and indicators than done in previous studies. Taking into account the UN Sustainable Development Goals, we recommend that national governments develop targeted policy responses to Indigenous health, improving access to health services, and Indigenous data within national surveillance systems.
|
|
| 2. |
- Umuroglu, Yaman, et al.
(författare)
-
Optimizing Bit-Serial Matrix Multiplication for Reconfigurable Computing
- 2019
-
Ingår i: ACM Transactions on Reconfigurable Technology and Systems. - : ASSOC COMPUTING MACHINERY. - 1936-7406 .- 1936-7414. ; 12:3
-
Tidskriftsartikel (refereegranskat)abstract
- Matrix-matrix multiplication is a key computational kernel for numerous applications in science and engineering, with ample parallelism and data locality that lends itself well to high-performance implementations. Many matrix multiplication-dependent applications can use reduced-precision integer or fixed-point representations to increase their performance and energy efficiency while still offering adequate quality of results. However, precision requirements may vary between different application phases or depend on input data, rendering constant-precision solutions ineffective. BISMO, a vectorized bit-serial matrix multiplication overlay for reconfigurable computing, previously utilized the excellent binary-operation performance of FPGAs to offer a matrix multiplication performance that scales with required precision and parallelism. We show how BISMO can be scaled up on Xilinx FPGAs using an arithmetic architecture that better utilizes six-input LUTs. The improved BISMO achieves a peak performance of 15.4 binary TOPS on the Ultra96 board with a Xilinx UltraScale+ MPSoC.
|
|
