Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uunet!convex!eugene!swarren From: swarren@eugene.uucp (Steve Warren) Newsgroups: comp.sys.amiga Subject: Re: Kronos vs. Hardframe Keywords: Kronos, Hardframe, controller Message-ID: <3481@convex.UUCP> Date: 28 Nov 89 21:25:16 GMT References: <3302@convex.UUCP> <8687@cbmvax.UUCP> Sender: usenet@convex.UUCP Reply-To: swarren@convex.COM (Steve Warren) Organization: Convex Computer Corporation, Richardson, Tx. Lines: 152 In article <8687@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >in article <3302@convex.UUCP>, swarren@eugene.uucp (Steve Warren) says: >> Keywords: Kronos, Hardframe, controller > >> The reason the Kronos can beat a DMA controller is because it has a 16-bit >> path to memory. Other controllers only have an 8-bit path to memory. > >Pure bunk. All DMA controllers have 16 bit data paths. You'd have to go >to a great deal of extra trouble to build a DMA device that supports 8 bit You are right, Dave, I didn't have the data sheet with me, so I came away with this impression even though it didn't really say this in the sheet. So I picked up a copy during lunch today. I will treat you to some of the statements in it. Here is what they said (which I misremembered, until I picked up a copy today - emphasis in original): "C Ltd's KRONOS controllers are the *only* non-DMA controllers utilizing full bus wide 16-bit data transfer. Using a unique dual-buffered psuedo- DMA design, KRONOS systems transfer data into the Amiga at full Amiga bus speed, in fact, KRONOS systems are so fast that the hard drive itself is now the limiting factor." >> Their literature is misleading... > >If they've managed to make you think that DMA controllers are doing 8 bit >transfers, they're far more misleading than you imagine. Well, this was my misunderstanding, C.Ltd did not make this claim. >> claiming that DMA controllers are fundamentally inferior when there is >> chip-ram contention. This is a false claim, and it is also unnecessary >> if the disk-perf results they published are true. (Their disk-perf results >> were significantly better than all the others) > >What were their DiskPerf results? Our disk folks are seeing over 1 meg/sec >with 2091s through the filesystem, with a good hard disk. The main trouble "...For example, when tested with the latest optimized Micropolis high- performance hard drive, the KRONOS controller read in and wrote back out 10 one-megabyte data segments in just under 20 seconds for an astounding 1009K/sec data transfer rate." >I've seen with the magazine reviews of HD controllers is that the reviewers >don't quite understand or bother to deal with a real, sound, scientifically >accurate comparison. For such a test, you've got to benchmark each controller >with exactly the same hard disk, and preferably several different ones, under >a variety of system load conditions. I haven't designed a hard disk >controller myself, but unless there's something really stupid in the design >of the 2091 and the Hardframe that I've somehow missed up to this point, >there's no way in the world any CPU driven controller should benchmark >faster than either of there in a fair test on a stock A2000 or even a >stock A2500/30. >-- >Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" > {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy > Too much of everything is just enough No, I agree 100%, that was the point I was making, that DMA *is* better. But here are some excerpts from the "marketing release" from C.Ltd re the KRONOS controller: "You can't buy a faster SCSI controller, hardcard or hard drive system for your Amiga at any price. (See the performance test report inside for details.)" Here is an excerpt from the DPerf 2.0 test results, in which C.Ltd says all tests used Quantun 40 Meg harddrives except Supradrive & the Flash!Card (apparently those cards don't support Quantum, so ST138N drives were used instead). These results were for 256K, read is first, then write. CLtd KRONOS/2000 648K 371K non-DMA MicroBotics HardFrame 590K 352K DMA Commodore A-2090A 584K 337K DMA CLtd ECONO/2000 453K 170K non-DMA Xetec FastCard 403K 276K non-DMA Supra Corp SupraDrive 391K 267K DMA IVS Trump Card 292K 230K non-DMA GVP Impact A2000 275K 154K non-DMA Expansion Tech Flash!Card 55K 50K DMA You asked for the disk-perf results; this is part of what they published. But here is some of the misleading stuff: "DMA OR NON-DMA. There are currently two types of SCSI controllers available for the Amiga computers, DMA (Direct Memory Access) and non- DMA. While DMA type systems in the past have been considerably faster than the non-DMA systems, the DMA systems have had problems competing with the Amiga's custom chips which are also DMA and have priority over other DMA devices..." ^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^ (misleading - the custom chips also have priority over the CPU, so this is not an advantage of non-DMA systems over DMA) "...The result of these conflicts has been a substantial reduction in DMA drive performance and possible disruption of video and sound..." ^^^^^^^^^^^^^^^^^^^^^ The problem I think they are referring to did not result from DMA, it was a design flaw that could have been made on any controller. The controller it affected happened to be a DMA drive. Now manufacturers are going around trying to scare people away from DMA. This is silly. "...These problems are seldom catastrophic, but are only destined to get worse as Commodore upgrades the custom chips allowing access to more chip memory and bigger screens thereby requiring more DMA for the custom chips and leaving less for other DMA devices..." ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This affects the CPU the same way it does DMA. If less time is available for other DMA devices, it is also not available for the CPU, at least with the current bus arbitration scheme. It would be a very complex system where loss of memory bandwidth would significantly reduce the time available to DMA without impacting CPU usage. "...The non-DMA design of the KRONOS controller provides faster data transfer than any available DMA controller while avoiding any conflicts (present or future) with the Amiga's custom chip set." C.Ltd may indeed have developed a nice way of avoiding custom chip conflicts, but whatever it is, it is not because it is non-DMA. The same techniques could be used for a DMA controller. This is what I meant when I said their literature is claiming that non-DMA is fundamentally superior to DMA on the Amiga. And, as I said, there is no reason to misrepresent or mislead this way, if their disk-perf results are accurate. If they really have such a fast controller, why don't just say, "This here wiz-bang controller is so well-designed, it beats the DMA controllers at their own game," or some such nonsense. There *are* some really excellent features claimed in their literature, like support for all kinds of different SCSI devices such as tape drives, removeable media, WORM drives, SCSI laser printers, page scanners, etc. They also claim to be the only controller that is supplied with "...a fully functional AmigaDOS type library that allows users to write and create custom software that can issue SCSI commands directly to SCSI devices." So, with all the features, I wish they didn't feel the need to obfuscate things. With all the dealers and developers on the net, I am suprised no one else has seen this four-page promotion for Kronos. I picked up a photocopy at the local dealer. Cheers, --Steve ------------------------------------------------------------------------- {uunet,sun}!convex!swarren; swarren@convex.COM