mquinn@pro-gsplus.cts.com (Michael Quinn) (05/25/90)
In-Reply-To: message from johnmac@fawlty.towers.oz
This is in reply to the message about the SCSI trnsfer rate of about 960k per
second:
You asked if he made it up? No, This is a very popular demo that MANY people
have seen. It's been mentio{ed on this mail system quite a few times now from
many people. You were right about the program not using GS/OS calls. I read
that they "cheated" to accomplish this, but they DID accomplish it. There
have just been too many people that have seen it.
Oh! and YES it was {_using SHR pictures.
Somewhere you mentioned a 12 or 14 >NANO<second drive? Are you sure about
that? I think you probably mean MILIsecond. My RAMchips are about 10 times
SLOWER than 12 or 14 nanosxDeconds. (I know this last paragraph was nit-picky,
sorry).
--s\lK1eM$oW~{wMM"EBb}uY1ye~cJZ^eS$Xi]0~{E!S:'
Y=Vv: #n;8>; kNK~C^Yj-FS.
__
/ \
| |
/ / mquinn@pro-gsplus.cts.com
/ \______________
_/ _____________)
_(_)_____)
/(_)______)
- (_)___)
_____________/
johnmac@fawlty.towers.oz (John MacLean) (05/28/90)
>Believe me, it is. Matt Gulick put together a "Star Wars" demo on his hard >disk, which was 30 megabytes of frames stored consecutively at the block >level. The frame rate was approx 33 fps. Now we are getting even more than far-fetched. 33fps = 33*32K = 1056K per second. This is *MORE* than Apple quotes as the theoretical upper limit of the DMA SCSI card transfer rate! All I can say is, well done Matt, you have achieved the impossible. I like to prove people wrong about the GS myself, but keep the facts straight. From what you have described, I can only conclude that: - the demo was very fast. - the graphics were stored at the block level (ie: no GS/OS or other file system on the partition). - palettes or scan line control bytes were not updated. Even with all of this (which I must say is very impressive, and I would really like to see it) I do not believe your frame rates. I would easily believe 20 fps (if the code is written in this way). I could possibly believe 25 fps or so. Any higher, and Apple has misquoted the transfer rate of thier card. It is simply not possible for the card to be run at its maximum transfer rate over multiple calls to the driver. This is because of delays between head movements, latency waiting for start of track (or at least start of block with full track caching). Would someone like to quote the ACTUAL number of SHR screens displayed, and the ACTUAL time it took to display them all. Regards, John MacLean. -- This net: johnmac@fawlty.towers.oz.au Phone: +61 2 427 2999 That net: uunet!fawlty.towers.oz.au!johnmac Fax: +61 2 427 7072 Snail: Tower Technology, Unit D 31-33 Sirius Rd, Home: +61 2 960 1453 Lane Cove, NSW 2066, Australia.
SOMCYC@NUSVM.BITNET (Chiang Yao Chye) (06/14/90)
I have 2 queries about the DMA SCSI card and I hope that some kind souls out there will have the answer... 1. Is the Applied Engineering GS-RAM plus card *REALLY* DMA compatible? Is it only DMA compatible to 4 megabytes, or compatible to 6 megabytes? 2. The older formatting/partitioning programs, specifically SCSI Hacker and Vanilla, do not work with the new DMA SCSI card. This is a loss, because these 2 programs can set the interleave factor when the hard disk is being formatted, whereas the SCSI Utilities that are supplied with the card cannot. So, is there a way to set the interleave of the drive by any other means, or is an upgrade to SCSI Hacker or Vanilla available to work with the new card? Thanks in advance. Chiang
gwyn@smoke.BRL.MIL (Doug Gwyn) (06/15/90)
In article <9006141515.AA13853@apple.com> SOMCYC@NUSVM.BITNET (Chiang Yao Chye) writes: >1. Is the Applied Engineering GS-RAM plus card *REALLY* DMA compatible? > Is it only DMA compatible to 4 megabytes, or compatible to 6 megabytes? I don't own a GS-RAM so I'm partly guessing, but (a) I don't think a 6MB configuration should be expected to work at all, due to the way the IIGS addresses memory; (b) I doubt that any RAM expansion card would be DMA compatible beyond 4MB, but this is not an issue for the High-Speed SCSI Card since it (or at least the GS/OS driver) doesn't use DMA beyond the lower 4MB anyway. (c) Early GS-RAMs were not DMA compatible, but an upgrade is available from AE. Recent GS-RAMs are supposed to be okay. A way to find out whether yours is is to boot GS/OS with the DMA switch ON on the High-Speed SCSI Card, and if it crashes (probably near the end of the boot thermometer) try again with the DMA switch OFF. >2. The older formatting/partitioning programs, specifically SCSI Hacker > and Vanilla, do not work with the new DMA SCSI card. This is a loss, > because these 2 programs can set the interleave factor when the hard > disk is being formatted, whereas the SCSI Utilities that are supplied > with the card cannot. So, is there a way to set the interleave of > the drive by any other means, or is an upgrade to SCSI Hacker or > Vanilla available to work with the new card? The Chinook SCSI Utilities have the same problem. I don't know about upgrades; however, 1:1 interleave works well with many SCSI drives (at least on my TWGS-equipped system). Because of GS/OS caching, interleave should not be of much concern, except for rare special applications such as the Star Wars demo.