Efficient, out-of-memory sparse MTTKRP on massively parallel architectures
Efficient, out-of-memory sparse MTTKRP on massively parallel architectures
复制标题
DOI：
10.1145/3524059.3532363
复制
发表时间：
2022-01
期刊：
影响因子：
--
通讯作者：
A. Nguyen;Ahmed E. Helal;Fabio Checconi;Jan Laukemann;Jesmin Jahan Tithi;Yongseok Soh;Teresa M. Ranadive;F. Petrini;Jee W. Choi
中科院分区：
文献类型：
--
作者：
A. Nguyen;Ahmed E. Helal;Fabio Checconi;Jan Laukemann;Jesmin Jahan Tithi;Yongseok Soh;Teresa M. Ranadive;F. Petrini;Jee W. Choi
张量分解是从高维(多峰)稀疏数据中提取潜在信息的一种重要方法。该研究提出了一种在大规模并行GPU架构上加速基本TD运算的新框架。与以前的工作不同，所提出的分块线性化坐标(BLCO)格式能够使用在单个张量副本上工作的统一实现来实现张量算法的高效内存溢出计算。我们的自适应阻塞和线性化策略不仅满足了GPU设备的资源限制，而且加快了数据索引速度，消除了控制流和内存访问的不规则性，减少了内核启动开销。为了解决GPU上巨大的同步开销，我们引入了一种机会主义冲突解决算法，在该算法中，线程协作而不是竞争存储器访问来动态地发现和解决它们冲突的更新，而不保留任何辅助信息或在特定模式方向存储非零元素。因此，与以前最先进的框架相比，我们的框架提供了卓越的内存中性能，并且是唯一能够处理内存外张量的框架。在最新的Intel和NVIDIA图形处理器上，BLCO在一系列实际稀疏张量上，与最先进的混合模式压缩稀疏光纤(MM-CSF)相比，实现了2.12-2.6倍的几何平均加速比(最高可达33.35倍)。
Tensor decomposition (TD) is an important method for extracting latent information from high-dimensional (multi-modal) sparse data. This study presents a novel framework for accelerating fundamental TD operations on massively parallel GPU architectures. In contrast to prior work, the proposed Blocked Linearized Coordinate (BLCO) format enables efficient out-of-memory computation of tensor algorithms using a unified implementation that works on a single tensor copy. Our adaptive blocking and linearization strategies not only meet the resource constraints of GPU devices, but also accelerate data indexing, eliminate control-flow and memory-access irregularities, and reduce kernel launching overhead. To address the substantial synchronization cost on GPUs, we introduce an opportunistic conflict resolution algorithm, in which threads collaborate instead of contending on memory access to discover and resolve their conflicting updates on-the-fly, without keeping any auxiliary information or storing non-zero elements in specific mode orientations. As a result, our framework delivers superior in-memory performance compared to prior state-of-the-art, and is the only framework capable of processing out-of-memory tensors. On the latest Intel and NVIDIA GPUs, BLCO achieves 2.12 -- 2.6X geometric-mean speedup (with up to 33.35X speedup) over the state-of-the-art mixed-mode compressed sparse fiber (MM-CSF) on a range of real-world sparse tensors.