admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) (05/05/90)
I got to play with an A3000 the other day. So .....
Are you listening CBM???
GOOD:
- The box. Nice job, looks much much better than the panzer A2000.
Smaller footprint. SCSI port on back. Nice height.
- OS 2.0 (or whatever). Very well done, even our Sun GUI could take
a few lessons from this one. Very professional.
- Price. CBM hit this one right on the head. Under cutting the MAC II's
(which isn't too hard) will help sell a lot of these. But that may
change (see below).
BAD:
- The architecture really hasn't changed. From what I've seen the
A3000 still has the same old chip/fast ram set up. My guess is
that the machine is going to real I/O bound with chip ram and
retrace on the bus. It's really time to change this. A Full 32-bit
chipset, maybe some type of dual-port ram or frame buffer for the
video. Anything to speed up the bus. The A4000 maybe??
- ZIP memories. I've always had problems with these things in sockets.
I know they save a lot of acreage on the board but I don't know. I
sure hope this doesn't cause problems.
- Calling it a workstation. When you do this you put this machine up
against some real heavy hitters (Sun, DEC, MIPS and to lesser extent,
Apple). This box has a nice price-point and that will help but not that
much. Check out Friday's S.F. Chronicle. There's an article on Sun
releasing a new workstation in the $4-5K range (list). And it will be
I figure 4X's faster. This trend of high speed/low cost workstations
isn't going away. As a personal computer the A3000 is fast but as a
workstation it's a slug.
Since CBM is pushing the multimedia angle, change the name to something
like "Media-Station". This avoids the workstaion label and opens a whole
new way to promote the A3000.
- Need 8-bit color hi-res, if your going to call it a workstaion.
OVERALL GRADE = B (not counting bugs at this point)
I've read quite a few reviews over the last couple of days and from the
prespective of past Amiga's it's a great machine. But it's evolutionary,
not revolutionary by any means. This machine can give CBM the time
necessary to build something really nasty for the workstaion market. I
just hope they do.
-------------------------------------------------------------------
"I think there's a world market for about 5 computers."
- Thomas J. Watson, Chairman of the Board, IBM (around 1948)
-------------------------------------------------------------------
uucp: {anywhere}!sun!admiral
-------------------------------------------------------------------eachus@linus.mitre.org (Robert I. Eachus) (05/08/90)
In article <135251@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes:
- Calling it a workstation. When you do this you put this machine up
against some real heavy hitters (Sun, DEC, MIPS and to lesser extent,
Apple). This box has a nice price-point and that will help but not that
much.
Come on!!! The A3000 out benches and outperforms ALL Sun-3's and
before, and is in the same performance ballpark as the SparcStation.
I would really have to do a lot of side-by-side benchmarking to
determine which is "faster," but the user percieved reality seems to
be close to no difference.
Check out Friday's S.F. Chronicle. There's an article on Sun
releasing a new workstation in the $4-5K range (list). And it will be
I figure 4X's faster. This trend of high speed/low cost workstations
isn't going away. As a personal computer the A3000 is fast but as a
workstation it's a slug.
Its nice that your company (Sun) is coming out with new products,
but what has that got to do with the definition of a workstation? And
if the Amiga 3000 at 6-7x an Amiga 2000 is a slug what is a Sun-3? I
prefer the response of the Amiga to that of any Sun, and I will often
flip screens (on my 2000) to do something else while waiting for a Sun
to finish some complex command like a directory listing...
--
Robert I. Eachus
with STANDARD_DISCLAIMER;
use STANDARD_DISCLAIMER;
function MESSAGE (TEXT: in CLEVER_IDEAS) return BETTER_IDEAS is...lphillips@lpami.wimsey.bc.ca (Larry Phillips) (05/08/90)
In <8812@sbcs.sunysb.edu>, rosej@sbsynchem.cs.sunysb.edu (John Rose) writes: >In article <135378@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes: >>In article <EACHUS.90May7145111@aries.linus.mitre.org>, eachus@linus.mitre.org (Robert I. Eachus) writes: >>> In article <135251@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes: >>> >>> Its nice that your company (Sun) is coming out with new products, >>> but what has that got to do with the definition of a workstation? And >>> if the Amiga 3000 at 6-7x an Amiga 2000 is a slug what is a Sun-3? >>> >>Again, Sun-3 is old stuff. Nobody in the market compares older, no longer >>sold machines to new ones. > >This just ain't so. In the mid and late '80s most computer companies in >the scientific computing market advertized their machines as being >X times faster than the VAX780 which was decade old technology. >Surely the Sun-3 in my office isn't that old ;-) (too be sure, the response >time sometimes seems that long ;-) Ahh. You are confusing a comparison with a base. When Sun (or anyone else) quotes MIPS, they are _basing_ the measurement on a machine that has well known performance. Folks can look at a figure and take a guess that their application will run n times faster. Though this can be called a comparison, nobody runs around saying "Oh boy! My machine is faster than a Vax 11/780." Suns tend to be 'old' after a very few years. A year ago, anyone in our office would have gladly accepted a 3/60 as a machine for their desk, now they would mutter things about SPARCstations and how hobbled they'd be by only 2 MIPs. >>What a meant to say is that when >>you call your machine a workstation you invite comaparisons to others who >>also call their machines a workstation. What I would like CBM to do is to >>AVOID this intanglement and really show their strength which is multimedia >>and price (not speed). That's what brought on the idea of Media-Station" >>or maybe "Multi-Station" to mind. >> > >Excellent point! Why put out "yet another workstation now" (YAWN)? >Why not draw attention to the A3000's unique strong points by calling >it a "Multi-Station"? Hear hear! -larry -- NeXT. The hardware makes it a PC. The software makes it a workstation. The units shipped makes it a mainframe. -=stolen from Hazy=- +-----------------------------------------------------------------------+ | // Larry Phillips | | \X/ lphillips@lpami.wimsey.bc.ca -or- uunet!van-bc!lpami!lphillips | | COMPUSERVE: 76703,4322 -or- 76703.4322@compuserve.com | +-----------------------------------------------------------------------+
barrett@jhunix.HCF.JHU.EDU (Dan Barrett) (05/08/90)
In article <EACHUS.90May7145111@aries.linus.mitre.org> eachus@linus.mitre.org (Robert I. Eachus) writes: > Come on!!! The A3000 out benches and outperforms ALL Sun-3's and >before, and is in the same performance ballpark as the SparcStation. Reality check -- no way! The A3000 is a cool machine, but it is not "in the same performance ballpark as the SparcStation". Reports say that the A2500/30 did about 7000 Dhrystones/sec. So maybe the A3000 does 8-9000. The Sparc does over 22,000. Dan //////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ | Dan Barrett, Department of Computer Science Johns Hopkins University | | INTERNET: barrett@cs.jhu.edu | | | COMPUSERVE: >internet:barrett@cs.jhu.edu | UUCP: barrett@jhunix.UUCP | \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////////
eric@oakhill.UUCP (Eric Quintana) (05/09/90)
In article <5203@jhunix.HCF.JHU.EDU> barrett@jhunix.UUCP (Dan Barrett) writes: > > Reality check -- no way! The A3000 is a cool machine, but it is >not "in the same performance ballpark as the SparcStation". > > Reports say that the A2500/30 did about 7000 Dhrystones/sec. So >maybe the A3000 does 8-9000. The Sparc does over 22,000. > > Dan However, an A3000 with an 040 SHOULD be "in the same performance ballpark as the SparcStation." Furthermore, the Dhrystones quoted above are for an 030 at 25MHz. No reason why someone couldn't build a card with a 030 running at 50MHz. The point is: the A3000 (or A2000 for that matter) COULD keep pace with a workstation if you plug in the right card. But alas, these cards do not exist today. Perhaps Commodore or GVP will unvale one of these gems within a year. One can hope. Eric Quintana ...!cs.utexas.edu!oakhill!eric
admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) (05/09/90)
In article <EACHUS.90May7145111@aries.linus.mitre.org>, eachus@linus.mitre.org (Robert I. Eachus) writes: > In article <135251@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes: > > - Calling it a workstation. When you do this you put this machine up > against some real heavy hitters (Sun, DEC, MIPS and to lesser extent, > Apple). This box has a nice price-point and that will help but not that > much. > > Come on!!! The A3000 out benches and outperforms ALL Sun-3's and > before, and is in the same performance ballpark as the SparcStation. > I would really have to do a lot of side-by-side benchmarking to > determine which is "faster," but the user percieved reality seems to > be close to no difference. Benchmarks running in that small cache or Fast Ram maybe but Chip ram is another thing. Also, comparing a brand new A3000 to 3/50-80's which are outdated today isn't a very strong comparison. If your trying to tell me that a 25 Mhz A3000 is comparable to a 25 Mhz SS-1+ or it's new cousin you need a strong dose of reality. And this isn't just because I work at Sun. The numbers just don't add up. ( This is where I ask CBM to publish Specmarks for the A3000 again...) > > Check out Friday's S.F. Chronicle. There's an article on Sun > releasing a new workstation in the $4-5K range (list). And it will be > I figure 4X's faster. This trend of high speed/low cost workstations > isn't going away. As a personal computer the A3000 is fast but as a > workstation it's a slug. > > Its nice that your company (Sun) is coming out with new products, > but what has that got to do with the definition of a workstation? And > if the Amiga 3000 at 6-7x an Amiga 2000 is a slug what is a Sun-3? I > prefer the response of the Amiga to that of any Sun, and I will often > flip screens (on my 2000) to do something else while waiting for a Sun > to finish some complex command like a directory listing... > Again, Sun-3 is old stuff. Nobody in the market compares older, no longer sold machines to new ones. As to the definition of a workstation. What a meant to say is that when you call your machine a workstation you invite comaparisons to others who also call their machines a workstation. What I would like CBM to do is to AVOID this intanglement and really show their strength which is multimedia and price (not speed). That's what brought on the idea of Media-Station" or maybe "Multi-Station" to mind. Anyway to each his own. Mick ------------------------------------------------------------------- "I think there's a world market for about 5 computers." - Thomas J. Watson, Chairman of the Board, IBM (around 1948) ------------------------------------------------------------------- uucp: {anywhere}!sun!admiral -------------------------------------------------------------------
rosej@sbsynchem.cs.sunysb.edu (John Rose) (05/09/90)
In article <135378@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes: >In article <EACHUS.90May7145111@aries.linus.mitre.org>, eachus@linus.mitre.org (Robert I. Eachus) writes: >> In article <135251@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes: >> >> Its nice that your company (Sun) is coming out with new products, >> but what has that got to do with the definition of a workstation? And >> if the Amiga 3000 at 6-7x an Amiga 2000 is a slug what is a Sun-3? >> >Again, Sun-3 is old stuff. Nobody in the market compares older, no longer >sold machines to new ones. This just ain't so. In the mid and late '80s most computer companies in the scientific computing market advertized their machines as being X times faster than the VAX780 which was decade old technology. Surely the Sun-3 in my office isn't that old ;-) (too be sure, the response time sometimes seems that long ;-) >What a meant to say is that when >you call your machine a workstation you invite comaparisons to others who >also call their machines a workstation. What I would like CBM to do is to >AVOID this intanglement and really show their strength which is multimedia >and price (not speed). That's what brought on the idea of Media-Station" >or maybe "Multi-Station" to mind. > Excellent point! Why put out "yet another workstation now" (YAWN)? Why not draw attention to the A3000's unique strong points by calling it a "Multi-Station"? -john (formerly of Sunshine Enterprises, where state of the art is a state of mind)
barrett@jhunix.HCF.JHU.EDU (Dan Barrett) (05/09/90)
I wrote: >> A3000 is a cool machine, but it is not "in the same performance >> ballpark as the SparcStation". In article <3268@rorschach.oakhill.UUCP> eric@rorschach.UUCP (Eric Quintana) writes: >However, an A3000 with an 040 SHOULD be "in the same performance ballpark as >the SparcStation." Oh sure. So would an IBM/XT, if you replaced it's processor with a 50 MHz 80486. That's irrelevant. We're comparing an EXISTING A3000 with an EXISTING Sparc. Of COURSE a machine goes faster if you add a faster processor. Sheesh! >But alas, these cards do not exist today. And it is silly to compare an existing machine (SPARC) to one that *might* exist in a year's time. Dan //////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ | Dan Barrett, Department of Computer Science Johns Hopkins University | | INTERNET: barrett@cs.jhu.edu | | | COMPUSERVE: >internet:barrett@cs.jhu.edu | UUCP: barrett@jhunix.UUCP | \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////////
eachus@linus.mitre.org (Robert I. Eachus) (05/09/90)
In article <135378@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes:
Again, Sun-3 is old stuff. Nobody in the market compares older, no
longer sold machines to new ones.
As to the definition of a workstation. What a meant to say is that
when you call your machine a workstation you invite comaparisons
to others who also call their machines a workstation....
I wasn't aware that Sun had stopped selling Sun-3's, and I am
certainly surprised to hear someone from Sun saying that Sun-3's
are no longer considered workstations. Amazing how fast technology
moves. :-)
Seriously, by the definitions we use here, the Sun-3/50, or an
original Mac II just make it across the line into workstations. A
Sun-3/80 or an Amiga 2500/30 is a mid-range workstation. The 3000
makes it into the high-end (mostly because of the greater than 16M
memory capacity*, and the Super Hi-Res video modes).
I am sorry that you feel that Commodore is poaching in Sun's
territory, but there are going to sell a LOT of machines into the
workstation market, with or without Unix. Whether you call the
machines workstations or not will not change that. If Sun can sell
the SPARCstation against the A3000, go right ahead. If Sun thinks
that the Sun-3 models can compete at the current prices, think again.
Robert I. Eachus
Amiga 3000 - The hardware makes it great, the software makes it
awesome, and the price will make it ubiquitous.
*p. s. no flames please about the availability of memory boards, I
know that TODAY if you want to go beyond 18 Meg you would have to
use 16 bit boards, but I am sure that 32-bit 16 Meg Zorro III
boards will be available soon after the 3000 starts shipping.
From what I hear of AmigaVision it is a typical memory hungry
Amiga graphics program. I used to say that the Amiga reminded me
of the plant in the two-pit room. When I first got a 1000, it
wispered "memory...memory," when I got it more it said "Memory...
Memory," now it is still asking for more, and my machine at work is
deciding that that 5 MEG is not enough and shouting "MEMORY,
MEMORY."
--
Robert I. Eachus
with STANDARD_DISCLAIMER;
use STANDARD_DISCLAIMER;
function MESSAGE (TEXT: in CLEVER_IDEAS) return BETTER_IDEAS is...wayneck@tekig5.PEN.TEK.COM (Wayne C Knapp) (05/10/90)
In article <EACHUS.90May7145111@aries.linus.mitre.org>, eachus@linus.mitre.org (Robert I. Eachus) writes: > In article <135251@sun.Eng.Sun.COM> admiral@m-5.Sun.COM (Michael Limprecht SUN Microsystems Mt. View Ca.) writes: > > - Calling it a workstation. When you do this you put this machine up > against some real heavy hitters (Sun, DEC, MIPS and to lesser extent, > Apple). This box has a nice price-point and that will help but not that > much. > > Come on!!! The A3000 out benches and outperforms ALL Sun-3's and > before, and is in the same performance ballpark as the SparcStation. > I would really have to do a lot of side-by-side benchmarking to > determine which is "faster," but the user percieved reality seems to > be close to no difference. > From what I can figure, an A3000 is pretty close to a Sun 3/80 which is about the same as a Next machine and a Mac IIxc (or something like that.). Well, when it comes to speed I heard many bitches in this group about how slow the Mac II is. It is well know that the Next machine gets pretty slow at times. The Sun 3/80 can't hold a candle to the SPARCstation 1. So I would have to agree that if you put UNIX on the A3000 you are going to have a real slug on your hands. I use the a SPARCstation 1 almost everyday. To say that any 68030 machine even the new Mac IIfx is as fast as a SPARC is clearly insane, or perhaps just very uninformed. IMHO, the SPARCstation 1 is the first computer I worked on that can run UNIX reasonably. We have two SPARCstations here that do more work than most mainframes I've seen. One of them is used to for cross-compiling and linking a lot. (Upto a couple hunderd compiles and links by five people a day.) The SPARC I use also is used by about 10 analog engineers to run spice. Some days I seen over 30 large spice jobs run on my SPARC. There is no way that the Sun 3/80 could do it. (We have one of those too.) Even when a SPICE job is running on my SPARCstation I have plenty of left over power to do anything I need. The A3000 is a very fast Amiga, but clearly by todays standards, it is a slug if you call it a workstation. I personally feel that it would simply be a waste of time to even bother to run UNIX on such a slow machine. Note the SPARCs we are using are over a year old and not even that fast by today's standards. Two years ago 68030 machines were okay in the speed department, but not today. However, this is not to degrade the A3000, as clearly Commodore isn't going head to head with Sun. There is a market for the A3000, and while it isn't a fantasic price for that level of power, it is a good price. Some people are likely to choose the A3000 over a Mac II or 386 PC. Surely that is the market that Commodore is after and not the workstation market. If they are after the workstation market they blew it, SPARCs are getting cheap! > Check out Friday's S.F. Chronicle. There's an article on Sun > releasing a new workstation in the $4-5K range (list). And it will be > I figure 4X's faster. This trend of high speed/low cost workstations > isn't going away. As a personal computer the A3000 is fast but as a > workstation it's a slug. > > Its nice that your company (Sun) is coming out with new products, > but what has that got to do with the definition of a workstation? And > if the Amiga 3000 at 6-7x an Amiga 2000 is a slug what is a Sun-3? I > prefer the response of the Amiga to that of any Sun, and I will often > flip screens (on my 2000) to do something else while waiting for a Sun > to finish some complex command like a directory listing... > > If Sun comes out with a faster SPARC for $4000, it will hurt some PC sales, some Mac sales and most likely the A3000 sales. The A3000 may be fast, but the SPARC is blazing. The Sun 3 line is old, the Sun 4 line is really some- thing. If you don't believe me, just look at Sun sales and profits. Now, if the 68040 were out in production, you might be talking a different story. Wayne Knapp
wayneck@tekig5.PEN.TEK.COM (Wayne C Knapp) (05/10/90)
In article <EACHUS.90May9111048@aries.linus.mitre.org>, eachus@linus.mitre.org (Robert I. Eachus) writes: > Seriously, by the definitions we use here, the Sun-3/50, or an > original Mac II just make it across the line into workstations. A > Sun-3/80 or an Amiga 2500/30 is a mid-range workstation. The 3000 > makes it into the high-end (mostly because of the greater than 16M > memory capacity*, and the Super Hi-Res video modes). The SPARCstation 1 comes with 8 megs. Runs at least 3x faster on interger type code. I don't think you can even compare the floating point performace as the SPARC does well, not great like the new IBM 6000 line but pretty good. Don't even talk about real tools. Graphics, well when compared to the top of the line Sun graphics, the A3000 is pretty limited unless you dump a lot of money into it. Even then you end up with just a hack. If you look at the price/performace ratio the SPARC is the bargin. Now, if the SPARCstation 1 is quickly falling into the mid-range workstation range, where does that leave the A3000? All this talk of comparing the A3000 to workstations is silly. If the A3000 was out two years ago it would have been a different story, but it just doesn't cut it today! However, in the new and growing multi-media market the A3000 maybe a winner. I hope, in fact I have a lot riding on it. Workstation no, useful PC yes. Wayne Knapp
kurt@tc.fluke.COM (Kurt Guntheroth) (05/10/90)
Gosh, what's the configured price for a Sun4-whatever-is-cheapest WITH local disk and software? $5K? $10K? $15K? Maybe the sun is faster. Maybe. But what about graphics? Are we comparing b&w to color? (Are we comparing 12" to 19"?) How much memory does the OS eat? What about dollars per MIP? Of course the 3000 isn't a sun. That would be silly. Maybe its just a better deal. Who needs all those MIPs at home or in a small business anyway. And if I was doing solid modelling, I'd need more than a sun anyway.
bga@pnet51.orb.mn.org (Bruce Albrecht) (05/26/90)
I don't really want to get involved in the discussion of the A3000 vs.
workstations, because I don't care whether the A3000 is labelled a workstation
or not. However, I agree with Wayne Throop that a multi-user system would be
nice. It may not be absolutely necessary (as John Sparks claims), but I'd
rather have one, even if the default is that there is only one user, and it
can do anything. After all, it's a lot easier to ignore the multi-user stuff,
than to fake it. And I can think of several things where it would be very
handy to have multi-user, like for NFS access, BBS's, and other forms of
remote access.
UUCP: {amdahl!bungia, uunet!rosevax, crash}!orbit!pnet51!bga
ARPA: crash!orbit!pnet51!bga@nosc.mil
INET: bga@pnet51.orb.mn.org