Conventional heterogeneous computing systems built on PCIe interconnects suffer from inefficient fine-grained host-device interactions and complex programming models. In recent years, many proprietary and open cache-coherent interconnect standards have emerged, among which compute express link (CXL) prevails in the open-standard domain after acquiring several competing solutions. Although CXL-based coherent heterogeneous computing holds the potential to fundamentally transform the collaborative computing mode of CPUs and XPUs, research in this direction remains hampered by the scarcity of available CXL-supported platforms, immature software/hardware ecosystems, and unclear application prospects. This paper presents Cohet, the first CXL-driven coherent heterogeneous computing framework. Cohet decouples the compute and memory resources to form unbiased CPU and XPU pools which share a single unified and coherent memory pool. It exposes a standard malloc/mmap interface to both CPU and XPU compute threads, which share a single per-process page table for user applications, leaving the OS dealing with smart memory allocation, page auto-migration, and management of heterogeneous resources. This design significantly simplifies heterogeneous parallel programming to a level comparable to homogeneous programming. To facilitate Cohet research, we also present a fullsystem cycle-level simulator named SimCXL, which is capable of modeling all CXL sub-protocols and device types. SimCXL has been rigorously calibrated against a real CXL testbed with various CXL memory and accelerators, showing an average simulation error of 3 %. Our evaluation reveals that CXL.cache reduces latency by 68 % and increases bandwidth by$14.4 \times$compared to DMA transfers at cacheline granularity. Building upon these insights, we demonstrate the benefits of Cohet with two killer apps, which are remote atomic operation (RAO) and remote procedure call (RPC). Compared to PCIe-NIC design, CXL-NIC achieves a 5.5 to$40.2 \times$speedup for RAO offloading and an average speedup of$\mathbf{1. 8 6} \times$for$\mathbf{R P C}$(de)serialization offloading.

With the rapid progress of medical image technology, medical image retrieval has attracted wide attention in medical data processing fields. Deep hashing methods have been proven effective for massive medical image retrieval. However, existing medical image retrieval methods ignore lesion context and category-level semantics, so it is difficult to correctly correspond to the context information and category of the lesion, resulting in poor performance. This paper addresses this dilemma with a novel Deep Semantic Ranking Hashing Based on Self-Attention (DSHA) approach. We first divide the medical triplet into smaller patches and send them to the multi-head self-attention module, which can more effectively encode the context information of the lesion and the interaction among patches. Meanwhile, the weight-sharing triplet networks are used to learn hash codes, which are forced to be semantically aligned. The proposed semantic enhancement loss effectively enhances the category-level semantics of hash codes. In addition, we added a semantic ranking penalty loss to optimize the retrieval accuracy. Extensive experiments on diverse medical image datasets prove that our DSHA method achieves remarkable results compared with the state-of-the-art medical image retrieval methods.