Intel hopes new chip-level security features move it past the Meltdown/Spectre fallout

Intel CEO Brian Krzanich. (GeekWire Photo / Taylor Soper)

As one of the biggest security events of the year kicks off Monday, Intel promised that its future chip designs would improve the state of information security, after flaws in older designs kicked off one of the biggest security messes in years.

Intel announced three new measures Monday night at the RSA Conference in San Francisco that will be built into future hardware and assume some of the load previously put on security software, which should improve performance and stability. CEO Brian Krzanich promised shortly after the disclosure of the Meltdown and Spectre vulnerabilities that the chip company would be overhauling its road maps to improve the security of its products, after a six-month scramble last year to patch operating systems and cloud services with workloads to deal with the design flaws, which lay undetected for almost two decades.

As part of Monday’s announcement, Intel is committing to a security training program run in conjunction with Purdue University to train a new generation of security professionals. There are also two new product-oriented initiatives being unveiled at RSA: Intel Threat Detection Technology and Intel Security Essentials.

Intel said its Threat Detection Technology will improve system security by freeing up graphics hardware resources that will be capable of scanning computer memory for malware, taking on in hardware what is often done by software running on general-purpose processors. Not only should this improve overall system performance, but using specialized hardware will increase the number of times memory can be scanned for issues, said Rick Echevarria, vice president, Software & Services Group and general manager, Platform Security Division at Intel.

Rick Echevarria, vice president, Software & Services Group; general manager, Platform Security Division, Intel (LinkedIn Photo)

“Malware is one of the fastest evolving workloads that we’re dealing with,” Echevarria said, and that’s even harder for hardware companies that update their designs yearly, if that often.

As malicious hacking operations increase in size and sophistication, detecting and removing them from a computer’s memory requires more and more horsepower, he said. In response, Intel is also unveiling new technology called Advanced Platform Telemetry, which uses machine-learning algorithms to help detect security threats on a corporate network. Cisco will be first partner to use this technology in a product.

Intel didn’t provide a ton of detail around Intel Security Essentials, describing them as a set of hardware-level features that will give systems builders security features such as a secure boot zone and secure storage for cryptographic keys.

The new features should help reassure the consumer and business markets that Intel is doing something about security, but monopoly power means that they don’t really have much of a choice when it comes to the processors that power our customers. Intel’s stranglehold on the datacenter market means that it could actually profit over time from its security gaffes, which will require big cloud computing providers to upgrade their systems over time to processors designed to prevent attackers from using the side-channel exploits identified in the Meltdown and Spectre research.

Companies are working on servers based around ARM processors, and AMD has a new server chip out that’s compatible with Intel’s designs, but virtually all the servers that power the cloud use Intel’s processors, and that’s not going to change any time soon. It’s a little different on the consumer side, where AMD is stronger and PC buyers can also entertain ARM-based tablet devices like the iPad Pro or certain Windows 10 devices.

Powered by WPeMatico