New Intel CPU Vulnerability Bodes Well For AMD
Intel processors are vulnerable to an attack, nicknamed Spoiler, to which AMD processors are immune according to researchers at the Worcester Polytechnic Institute and the University of Lübeck. Intel will not be able to spin this as an industry-wide problem as they did last January when two other vulnerabilities, Spectre and Meltdown, were revealed. This bodes well for AMD shareholders.
In January 2018, when Spectre and Meltdown were revealed, Intel said it was an industry-wide problem implying that Intel processors were not at a disadvantage to AMD. This time around the researchers tested AMD processors and found them to be immune. Consequently, Spoiler will give AMD an advantage over Intel.
The question now is whether AMD's advantage will be big enough and last long enough for them to gain significant market share.
Intel initially addressed Spectre and Meltdown by releasing performance sapping software patches to the microcode in their processors.
Last January, early estimates of the performance penalty for the Spectre and Meltdown patches ranged from 5% to 25%. Since then datacenter system admins have told me that the patches have gotten more efficient and the performance penalty has decreased.
However, recently Intel changed the licensing agreement for these software patches to prevent developers from publishing benchmark results.
In the near-term, I expect Intel will come out with a software patch for Spoiler. However, researchers say Spoiler, “is not something you can patch easily with microcode without losing tremendous performance”. The degree of the performance penalty exacted by these patches is a good measure of the size of AMD's advantage. Intel can prevent developers from publishing their benchmark results, but they can't stop them from talking to each other. This information will get out. Investors will have to look for it on more technical websites frequented by developers.
The researchers are of the opinion that Spoiler cannot be fully fixed with a software patch. They believe changes to Intel's chip architecture will be required.
Intel's is already years behind schedule in moving from 14 nm production lines to 10 nm. In contrast, AMD will soon be making its processors on a 7 nm production line.
If Intel now also needs to redesign their processors to address Spoiler I cannot see how this can be accomplished in less than 5 years. That's enough time for AMD to take significant market share.
My Take: Intel has some tough decisions to make for their processor product line. Moving from 14 nm to 10 nm production lines requires multi-billion dollar investments with long lead-times. Changing their processor architecture at the same time compounds the problem.
Bob Swan, Intel's CEO, does not have expertise in these technical areas. He will, no doubt, have access to the best consultants in the industry, but in the end, investors have to trust the CEO to make the right decision. The fact that Intel restricts developers from publishing their benchmark results reduces my trust in them. How much performance is sapped by Intel's software patches is a key piece of information that should not be kept from investors. I cannot recommend Intel stock until Swan is more forthcoming.
In contrast, AMD was the best performing stock in the S&P 500 for 2018 with an almost 69% gain. Tony Mitchell, one of my managers, first bought AMD in October 2014 at $3.45. At today’s price of $23.50, he has already made a lot of money on his original investment and he sees a lot more upside in 2019.
After publication, an Intel spokesperson provided this statement:
In plain English, I think Intel is laying the blame for Spoiler on software developers who don't follow "side channel safe software development practices", and manufacturers of memory modules that have not "mitigated" against this kind of attack. In my view, if Spoiler was entirely due to sub-par software and memory modules, it would affect AMD processors as well. Intel's statement does not change my take on this news.