Earlier this week, we posted a bit of an alarmist post about AMD’s future, based on a report from Kerrisdale Capital Investments. AMD got wind of the report, and took umbrage with a great deal of it, letting us know of inaccuracies contained within.
For starters, though much of the information in the report is sourced the way any report of its ilk might be, one thing that isn’t made clear in the thing is that it’s a report compiled by students.
“The “analysis,” was actually done by a team of MBA students from Warwick Business School who entered competition run by Kerrisdale Capital Investments” contends AMD’s Director of Corporate Communications, Drew Prairie. “Students were asked to predict the future based on modeling and their understanding of the market and trends.”
So while the report may have contained a great deal of factual information, it was also heavily based on conjecture, and a few of the report’s claims were based on faulty reasoning, with Prairie adding that they made gross over-simplifications. He also says that the students’ lack of faith in AMD’s roadmap was misplaced – especially seeing as at the time the report was made, that AMD hadn’t disclosed that information yet.
Those plans include its ZEN core, high-performance server-grade ARM cores and high-performance server APUs. The next-generation 64-bit x86 processor core, codenamed “Zen” is what AMS except to help drive its re-entry in to the top-end desktop and server markets, making good on CEO Lisa Su’s promises not to always be the low cost solution. Its first custom 64-bit ARM core, “K12? core, is an embeddable chip that’s made for efficiency. AMD is also banking on its stacked HBM memory, coming soon to its enthusiast cards, to help shake up the GPU market. Its newly announced, budget, eSports-focused APU’s should also help with its return to consumer favour.
In the article, I also said that AMD’s got “close to no R&D budget,” which Prairie pointed out wasn’t quite right too. While it is true that AMD’s annual research budget has dropped sharply (now half a billion Dollars less than it was 5 years ago), Prairie asserts that it’s down to efficiency.
“Nearly $1.1B in 2014 isn’t “close to no research budget”…and our CTO spent a fair amount of time at our Financial Analyst Day earlier this month talking through how we have re-engineered our R&D functions to work more efficiently by adopting a modular, building block design approach for future products so that we are re-using key IP more effectively across products.
We also talked about future GPU cores planned for next year that will deliver 2x perf/watt gains over current products as well as a new x86 CPU core that will deliver 40% IPC uplift. Innovation is lifeblood of the company, and we are by all means investing to drive this.”
For reference, Nvidia’s annual R&D budgets sits at around $1.3 Billion USD – but it has its fingers in fewer pies – investing more R&D in to a narrow set of products. Intel, AMD’s largest semi-conductor rival spends $11.5 Billion annually on its R&D.
While AMD’s financial situation isn’t at all where the company would like it to be, It’s unlikely that AMD will declare bankruptcy by 2020 – though any sort of turnaround will be an uphill struggle. And it’s one I certainly hope that AMD overcomes.
Last Updated: May 29, 2015