NetSpectre: CPU vulnerability enables remote data theft

The newly discovered CPU vulnerability NetSpectre allows data theft via remote access. We tell you what you should know now.

With NetSpectre, Graz University of Technology discovered a new variant of the CPU security gap Spectre, which enables data theft via remote access. Of the known attack vectors, which can be derived from Meltdown and Spectre, only NetSpectre has this advantage. This means that a hacker could read the system memory without having to execute code locally.

In order to minimize the vulnerability potential of the Meltdown and Spectre vulnerabilities, Microsoft, AMD, Intel and other companies released a number of updates and hardware protection measures. However, researchers have since the beginning of the year identified more versions of the CPU vulnerability Spectre. Until now, one of the prerequisites for a successful attack was that hackers must first install malware on the target system.

NetSpectre: Threat (still) low

The NetSpectre vulnerability discovered by the TU Graz research team circumvents this requirement, but does not represent an acute threat for the time being. On the one hand, NetSpectre is based on the Spectre attack variant 1, against which various companies already offer patches for download. On the other hand, the data transfer rate achieved in the experiment was only between 15 and 60 bits per hour.

For comparison: 97,600 bit memory (about 12 KB) is required for a Word 2010 document saved as an empty sheet. With additional text information, the Word 2010 document quickly takes up over 100,000 bits of disk space. It would take at least 70 days to download the Word document with the maximum transmission rate of 60 bits per hour achieved in the attempt. The transfer rates are currently too slow for a serious attack, but this could change in the future.