Password Cracking of PDF 2.0 documents on GPU

PDF is one of the international documents standards and it is widely used in the world. Password cracking of the PDF documents has attracted the attention of forensic community. Some forensic tools can crack the password of PDF documents with GPU-acceleration only for PDF l.x documents. This paper studies the password cracking of PDF 2.0 documents, which is the newest PDF standard, on GPU. We analyze the complexity of the password cracking algorithm, and implement it on GPU. We adopt many optimization techniques and our experiments shows that our GPU implementation has about 47 times speedup, compared with the CPU implementation.