The integration of any two companies is a painful process. I got further explanation about the exodus of former MIPS people -- sales and FAEs -- in Korea, Taiwan, Japan and China.
I received more messages and phone calls than expected responding to my previous two blog posts on Imagination/MIPS -- "Yoshida in China: Imagination dumps MIPS sales team," and "Yoshida in China: Imagination's MIPS roadmap."
Clearly, the integration of any two companies is a painful process. I got further explanation about the exodus of former MIPS people -- sales and FAEs -- in Korea, Taiwan, Japan and China.
Sources close to MIPS/Imagination cite a fundamental difference in pay scale between the U.S.-based MIPS and the U.K.-based Imagination. Former MIPS sales people were reportedly asked to accept pay cuts -- some as much as 60 percent; all will be receiving no commissions as of April. Since it was Imagination who acquired MIPS, you might say itís only natural that Imagination favors its own people (each company had a sales team in the region before the acquisition). Put it all together and it makes sense to wonder if Imagination ever intended to keep any MIPS sales people.
Obviously, the new MIPS/Imagination team needs to get past any hard feelings born out of the recent changes. At the same time, the team also needs to convince, if not convert, skeptics who arenít sure of the future of MIPS processor cores in Imaginationís hands.
I interviewed several industry sources including MIPS customers, industry analysts and former MIPS employees. Their concerns about MIPS/Imaginationís future can be distilled into three questions.
1. Itís no secret that 15 to 20 percent of Android apps do not run well on MIPS-based devices. As J. Scott Gardner, a senior analyst at the Linley Group, wrote in the Feb. Microprocessor Report, ďFew mobile customers want a device that runs Ďmostí apps when it is so easy to buy an ARM-based device that runs all apps.Ē How will Imagination respond?
2. Several sources point out that enabling binary translation of ARM code to MIPS code on the fly might be the way to mitigate this. What progress has MIPS/Imagination made in their efforts to offer that tool?
3. Can Imagination afford not to go after the mobile business with its newly acquired MIPS cores?
Armed with these questions, I went back to Imagination.