From: marc@Autodesk.COM (Marc Ramsey) Newsgroups: alt.sys.sun Subject: Re: wanted: any info on Sun RoadRunner Message-ID: <7827@autodesk.COM> Date: 19 Sep 91 00:50:56 GMT References: <18353@life.ai.mit.edu> Organization: Autodesk Inc., Sausalito CA, USA Lines: 42 >The sunview windowing system runs PAINFULLY slow on this box, I'm >suprised sun allowed this to ship with such poor UI performance. I >used to have a 250 model with 16MB ram and color, it ran X11R4 very >well. If you're thinking of getting one of these, try to get the docs >with it. There are some tricky differences from "classic" sunOS in >the 386i OS. And don't even think about using the standard sunview >windows on this thing. As one of the 386i developers, and as the one who should most directly take the blame for the "PAINFULLY" slow sunview implementation, I'd like to make a few comments. Every benchmark we ran indicated that the overall performance of sunview on the 386i exceeded that on the Sun3, and was comparable to the 4/110 (the SPARCstation came later). One of the main problems was that the Sun3 MMU provided far superior performance to the internal 386 mechanism. This meant that things ran great as long as you didn't have to page, but would bog down somewhat if you did. The other big problem was the "clever" engineering trick of putting the cache on the memory boards, which meant that all frame buffer accesses were totally uncached. The tricky differences from the Moto/Sparc versions of SunOS were mostly confined to the system administration area. The 386i was running a version of 4.X at a time when most Sun3 developers were still using 3.X, and a lot of problems cropped up during the transistion process. It was also, of course, necessary to be a little more aware of byte ordering and alignment issues. The 486i (of which only a few upgrades made it out the door) solved many of the performance issues. It actually benchmarked and felt faster than a SPARC 1+. The main reason for the demise of the x86i product line was political, i.e., "throwing all of the wood behind one arrow". Solaris 2.0 for the x86 reopens the argument that we used to have during the early stages of 386i development. Us software folks were perfectly happy running on Compaq 386/16s, and we argued loudly that Sun should simply sell SunOS for generic 386s, rather than trying to build the box. Had they done so at the time, they would now probably own the desktop UNIX market, and made some inroads into what has become the Windows 3.0 market. I think now it is too little, too late... Marc Ramsey marc@autodesk.com