Abstract:PCIe (Peripheral Component Interconnect express) protocol is the de facto protocol to bridge CPU and peripheral devices like GPU, NIC, and SSD drive. There is an increasing demand to install more peripheral devices on a single machine, but the PCIe interfaces offered by Intel CPUs are fixed. To resolve such contention, PCIe switch, PCH (Platform Controller Hub), or virtualization cards are installed on the machine to allow multiple devices to share a PCIe interface. Congestion happens when the collective PCIe traffic from the devices overwhelm the PCIe link capacity, and transmission delay is then introduced.In this work, we found the PCIe delay not only harms device performance but also leaks sensitive information about a user who uses the machine. In particular, as user’s activities might trigger data movement over PCIe (e.g., between CPU and GPU), by measuring PCIe congestion, an adversary accessing another device can infer the victim’s secret indirectly. Therefore, the delay resulted from I/O congestion can be exploited as a side-channel. We demonstrate the threat from PCIe congestion through 2 attack scenarios and 4 victim settings. Specifically, an attacker can learn the workload of a GPU in a remote server by probing a RDMA NIC that shares the same PCIe switch and measuring the delays. Based on the measurement, the attacker is able to know the keystroke timings of the victim, what webpage is rendered on the GPU, and what machine-learning model is running on the GPU. Besides, when the victim is using a low-speed device, e.g., an Ethernet NIC, an attacker controlling an NVMe SSD can launch a similar attack when they share a PCH or virtualization card. The evaluation result shows our attack can achieve high accuracy (e.g., 96.31% accuracy in inferring webpage visited by a victim).

GPU.zip: On the Side-Channel Implications of Hardware-Based Graphical Data Compression

CODAG: Characterizing and Optimizing Decompression Algorithms for GPUs

GPU Side-Channel Attacks are Everywhere: A Survey

MagView++: Data Exfiltration Via CPU Magnetic Signals under Video Decoding

A New Profiling-Based Side-Channel Attack on Graphics Processing Units

Whispering Pixels: Exploiting Uninitialized Register Accesses in Modern GPUs

Surface Compression Using Dynamic Color Palettes

Hot Pixels: Frequency, Power, and Temperature Attacks on GPUs and ARM SoCs

GPULZ: Optimizing LZSS Lossless Compression for Multi-byte Data on Modern GPUs

A Parallel Compression Pipeline for Improving GPU Virtualization Data Transfers

Hybrid-Smash: A Heterogeneous CPU-GPU Compression Library

IOTLB-SC: An Accelerator-Independent Leakage Source in Modern Cloud Systems

FPGA Acceleration of Zstd Compression Algorithm

Towards Practical Meshlet Compression

Vulnerable GPU Memory Management: Towards Recovering Raw Data from GPU

Accelerating Video Decoding Using GPU

FZ-GPU: A Fast and High-Ratio Lossy Compressor for Scientific Computing Applications on GPUs

Invisible Probe: Timing Attacks with PCIe Congestion Side-channel.

cuSZ-$i$: High-Ratio Scientific Lossy Compression on GPUs with Optimized Multi-Level Interpolation

Beyond the Bridge: Contention-Based Covert and Side Channel Attacks on Multi-GPU Interconnect

CacheZoom: How SGX Amplifies The Power of Cache Attacks