jms@tardis.Tymnet.COM (Joe Smith) (09/20/90)
>>I believe that the method employed here is using a 16 bit wide bus (dubbed >>SCSI II), instead of the former 8 bits. In article <14418@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: >SCSI II isn't 16 bits wide, it's still 8 bits. It specifies the SCSI command >set in a much more standardized fashion than original SCSI, and has a fast >synchronous transfer mode capable of 10 megabytes/second. I've heard that from a software point of view, SCSI-II is mainly the enhanced set of commands that Dave mentions. But from a hardware point of view, SCSI-II has two options, which can be combined. 1) Just the new commands (including better connect/reconnect) = 5M bytes/sec (this is copatible with SCSI-1 devices) 2) Fast SCSI = 10 megabytes per second 3) Wide SCSI = 16 bit transfers at 5 million per second = 10 megabytes/sec (SCSI-I and SCSI-II devices can co-exist on the 16-bit bus) 4) Fast and wide = 16 bit transfers at 10 M = 20 megabytes per second. At Sun-Expo, several companies announced plans for 16-bit SCSI-II. No single disk produces 20 megabytes per second, but some companies are working on disk clusters that act like a single, large, fast disk. -- Joe Smith (408)922-6220 | SMTP: jms@tardis.tymnet.com or jms@gemini.tymnet.com BT Tymnet Tech Services | UUCP: ...!{ames,pyramid}!oliveb!tymix!tardis!jms PO Box 49019, MS-C41 | BIX: smithjoe | 12 PDP-10s still running! "POPJ P," San Jose, CA 95161-9019 | humorous dislaimer: "My Amiga speaks for me."
guineau@wjg.enet.dec.com (W. John Guineau) (09/21/90)
In article <1232@tardis.Tymnet.COM>, jms@tardis.Tymnet.COM (Joe Smith) writes: [stuff deleted] |> |> I've heard that from a software point of view, SCSI-II is mainly the enhanced |> set of commands that Dave mentions. But from a hardware point of view, |> SCSI-II has two options, which can be combined. |> |> 1) Just the new commands (including better connect/reconnect) = 5M bytes/sec |> (this is copatible with SCSI-1 devices) |> 2) Fast SCSI = 10 megabytes per second |> 3) Wide SCSI = 16 bit transfers at 5 million per second = 10 megabytes/sec |> (SCSI-I and SCSI-II devices can co-exist on the 16-bit bus) |> 4) Fast and wide = 16 bit transfers at 10 M = 20 megabytes per second. |> SCSI II also specifies a 32 bit data path option. Couple that with FAST and you get 40MB/s. Both 16 and 32 bit require a second cable (termed the B cable). I think there are actually 2 or 3 cabling schemes defined in the latest spec revision. |> At Sun-Expo, several companies announced plans for 16-bit SCSI-II. |> No single disk produces 20 megabytes per second, but some companies are |> working on disk clusters that act like a single, large, fast disk. Yup, they're called arrays. There is a paper from Berkeley describing various array architectures (RAID levels | RAID = Redundant Array using Inexpensive Disks). RAID 3 and 5 are the most common. RAID 3 is optimum for getting lots of KB/s while RAID 5 gives lots of IO/s. |> -- |> Joe Smith (408)922-6220 | SMTP: jms@tardis.tymnet.com or jms@gemini.tymnet.com |> BT Tymnet Tech Services | UUCP: ...!{ames,pyramid}!oliveb!tymix!tardis!jms |> PO Box 49019, MS-C41 | BIX: smithjoe | 12 PDP-10s still running! "POPJ P," |> San Jose, CA 95161-9019 | humorous dislaimer: "My Amiga speaks for me." |> -- W. John Guineau guineau@wjg.enet.dec.com Digital Equipment Corporation Marlboro MA. 01752
wayned@wddami.spoami.com (Wayne Diener) (09/21/90)
>In article <1990Sep20.125810@wjg.enet.dec.com> guineau@wjg.enet.dec.com (W. John Guineau) writes: >In article <1232@tardis.Tymnet.COM>, jms@tardis.Tymnet.COM (Joe Smith) writes: > >[stuff deleted] > >|> >|> 3) Wide SCSI = 16 bit transfers at 5 million per second = 10 megabytes/sec >|> (SCSI-I and SCSI-II devices can co-exist on the 16-bit bus) >|> 4) Fast and wide = 16 bit transfers at 10 M = 20 megabytes per second. >|> > >SCSI II also specifies a 32 bit data path option. Couple that with FAST >and you get 40MB/s. Both 16 and 32 bit require a second cable (termed the Before everyone gets too excited about the _possible_ data transfer rates for high speed disk drives, just remember that you _cannot_ take the data off the disk & put it somewhere useful (i.e. memory) at any higher rate than the bandwidth of the bus. I don't know what the bandwidths of the Zorro I/II/III busses are, but I have serious doubts that a 40 MB (or even 20 MB) per second transfer rate is possible. -- |---------------------------------------------------------------| | // Wayne D. Diener | | // Spokane, WA | | \\ // E-mail reply to: | | \X/ To: isc-br!hawk!wddami!wayned@uunet.uu.net |
daveh@cbmvax.commodore.com (Dave Haynie) (09/25/90)
In article <wayned.9568@wddami.spoami.com> wayned@wddami.spoami.com (Wayne Diener) writes: >>In article <1990Sep20.125810@wjg.enet.dec.com> guineau@wjg.enet.dec.com (W. John Guineau) writes: >>SCSI II also specifies a 32 bit data path option. Couple that with FAST >>and you get 40MB/s. Both 16 and 32 bit require a second cable (termed the >Before everyone gets too excited about the _possible_ data transfer rates >for high speed disk drives, just remember that you _cannot_ take the data >off the disk & put it somewhere useful (i.e. memory) at any higher rate >than the bandwidth of the bus. I don't know what the bandwidths of the >Zorro I/II/III busses are, but I have serious doubts that a 40 MB (or even >20 MB) per second transfer rate is possible. Zorro III can handle 20 MB/s on-bus with no problems, though not going to the A3000's 68030 bus fast memory. And between a Zorro III master and slave with burst capability, 40 MB/s is certainly possible. Which doesn't matter in this case, because the A3000's DMA controller is not on the Zorro III bus, it's on the 68030 bus. That bus would have no trouble supporting 40 MB/s transfers, though the standard A3000 memory isn't quite that fast. Which doesn't matter in this case, because the A3000's SCSI chip is only an 8 bit SCSI, which, even with the fastest SCSI-2 transfer, only yields 10 MB/s. So you're not likely to have much trouble with the A3000's internal bus speeds. Anyone planning a 32 bit SCSI-2 bus mastering Zorro III board, feel free to give me a call. I'll help you all you need if I get free sample :-)... >| // Wayne D. Diener | -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Standing on the shoulders of giants leaves me cold -REM