fbpx
Connect with us

Hi, what are you looking for?

Evolution Magazine

Tech

New CPU Design Baffles Hundreds of Expert Hackers in DARPA Tests

Hacking Tests Run On “Morpheus” Microprocessor Design Proves Near Impossible to Hack Into

A microprocessor codenamed “Morpheus” has baffled almost 600 professional hackers as they attempted to break into its system during a series of tests for the DARPA-funded project. The hackers attempted to hack a medical database utilizing this CPU through the injection of code into the machine. This resulted in 13000 collective hours of attempts and no success.

This new CPU design is designed differently than typical x86 processors. Morpheus’s secret is that it’s constantly rewriting its interior architecture, faster than modern hacking techniques can keep up. This prevents hackers from using Spectre and Meltdown-styled side-channel methods to infiltrate the system.

The excitement of Morpheus’s creation needs to be paused and assessed, however. Morpheus was implemented and tested on a considerably outdated system, one which isn’t even close to the modern computers and laptops you can purchase in the store. Not being tested on a higher functioning chip makes the analysis of the Morpheus slightly inaccurate to real-life applications and convenience.

Nevertheless, the results are quite astonishing as they’re still not comparable to any other microprocessor. The test was done on a gem5 simulator on a Xilinx FPGA. The simulator tested a MinorCPU 4-stage, in order core which would run at 2.5GHz. it would run with 32KB L1i and 32KB L1d and the L2 cache was 256KB.

The general idea behind this processor was rather than tackling only specific and identified methods of hacking, the processor would make itself a difficult target to grab hold of for any hacker. The best way to do this while keeping the actual coding intact is by using undefined semantics.

Undefined semantics are those pieces of information that users and programmers don’t need to know to keep the system operational and running. These pieces of information are encrypted by Morpheus every 100 milliseconds over and over again. This isn’t even enough time for the hacker to find an entrance window or weak point to infiltrate from.

The downfall to this process is that there’s a performance penalty of the overall processor by about 10%. The constant pointer encryption scheme that Morpheus creates is labeled as a ‘churn’. There is a positive correlation between the churn and the performance penalty, meaning the greater the churn, the greater the penalty.

This microprocessor does not stop every kind of hack either. SQL injection, man-in-the-middle web servers, Spearphisher techniques, they’re all still a vulnerability of the CPU. Todd Austin himself believes that the microprocessor is not impossible to hack, just extremely hard to.

One thing is for sure. If the Morpheus microprocessor can be made more efficient it’ll be a hot take on the previous outdated microprocessor models which get hacked as this one is self-sufficient. People will have a new level of personal safety with a microprocessor that takes its measures to make itself hard to hack.

Reported By: Imaaz

Written By: Leah Latif

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

You May Also Like

Science

The Neuralink brain chip is expected to be developed within a few years

Tech

Fastest internet speed set by Japanese Engineers.

Data Science

As part of Pakistan’s strategy to use drone technology to plant millions of trees and strengthen the ‘Clean Green Pakistan Movement,’ the Ministry of...

Science

Immortality might not be such an inconceivable concept anymore! University of California San Francisco provided a study where the cognitive ability of mice was...

Copyright © 2021 Powered by Emporia Digital Solutions (Pvt) Ltd.