Cache-Integrated Network Interfaces: Flexible On-Chip Communication and Synchronization for Large-Scale CMPs

Per-core scratchpad memories (or local stores) allow direct inter-core communication, with latency and energy advantages over coherent cache-based communication, especially as CMP architectures become more distributed. We have designed cache-integrated network interfaces, appropriate for scalable multicores, that combine the best of two worlds – the flexibility of caches and the efficiency of scratchpad memories: on-chip SRAM is configurably shared among caching, scratchpad, and virtualized network interface (NI) functions. This paper presents our architecture, which provides local and remote scratchpad access, to either individual words or multiword blocks through RDMA copy. Furthermore, we introduce event responses, as a technique that enables software configurable communication and synchronization primitives. We present three event response mechanisms that expose NI functionality to software, for multiword transfer initiation, completion notifications for software selected sets of arbitrary size transfers, and multi-party synchronization queues. We implemented these mechanisms in a four-core FPGA prototype, and measure the logic overhead over a cache-only design for basic NI functionality to be less than 20%. We also evaluate the on-chip communication performance on the prototype, as well as the performance of synchronization functions with simulation of CMPs with up to 128 cores. We demonstrate efficient synchronization, low-overhead communication, and amortized-overhead bulk transfers, which allow parallelization gains for fine-grain tasks, and efficient exploitation of the hardware bandwidth.

Renal replacement treatment for diabetic nephropathy in Northern Ireland 1979-1987

Twenty-three patients with end-stage renal failure due to diabetic nephropathy received renal replacement treatment. All patients had insulin-dependent diabetes mellitus. Nineteen transplants were performed in seventeen patients. Two-year graft survival for all transplants was 74% with a two-year patient survival post-transplantation of 81%. Overall two-year patient survival was 73%, compared with 82% in non-diabetic patients receiving renal replacement treatment. In diabetic patients accepted for treatment there was a high incidence of non-renal complications, particularly vascular disease. An aggressive approach to the treatment of vascular disease in these patients may improve overall survival rates.

Renal replacement therapy in multiple myeloma and systemic amyloidosis

Renal failure frequently complicates both multiple myeloma and systemic amyloidosis. Renal replacement therapy (RRT) may be poorly tolerated and its role in such patients is not clearly defined. Of fifty patients (26 males and 24 females) referred to a single centre because of renal failure associated with multiple myeloma or systemic amyloidosis 37 progressed to end-stage renal failure and 30 of these patients received RRT. Nine patients have been treated by CAPD, 13 by haemodialysis, and 8 patients have required both forms of dialysis. Overall one year and two year survival rates were 66% and 57% respectively. The median duration on RRT was 7.5 months (range 1-96 months) with a 51% one year, and a 46% two year survival rate. Of 7 patients with amyloidosis who underwent renal transplantation, 3 died within 6 months of transplantation. Undiagnosed cardiac involvement contributed to this early mortality. We conclude that renal replacement therapy is appropriate for some patients with multiple myeloma and systemic amyloidosis who develop endstage renal failure. Careful assessment and selection of patients is necessary prior to renal transplantation.