SNP genotyping device is an essential tool in the upcoming era of personal genome, personalized medicine. Human genome has more than 10 million SNPs while conventional SNP genotyping device can only hold 1 million SNPs. Thus intelligent SNP contents selection is required to maximize the value of SNP genotyping device. Here, we developed a new selection algorithm and applied this method to design genotyping contents for cancer and pharmacogenomic association study. This approach significantly increased the product value when compared to contents of competitive SNP genotyping product.