Date: Fri, 3 Nov 2000 10:33:21 -0800 From: "Heredity Choice" <stork@QNET.COM> To: "Allen Landsidel" <all@biosys.net>, "Nick Slager" <nicks@albury.net.au>, "jadream" <jadream@chat.ru> Cc: <freebsd-questions@FreeBSD.ORG> Subject: RE: FBSD & Itanium? Message-ID: <005001c045c4$8b05bed0$15c6ddd1@STORK> In-Reply-To: <4.3.2.7.2.20001028110103.00c8c978@mail.megapathdsl.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Since its introduction in 1992, the DEC Alpha has always been the fastest processor, with only Hewlett Packard coming close. Since the demise of DEC, development continues, but for it to maintain its lead throughout this coming decade will depend on the continuing support of IBM, whose long-term commitment is unknown. For the prices on new Alphas, I quote the following from www.microway.com : "From the 533-LX Performance Value to the 667MHZ SuperCache-8, Microway is your best source for high-performance Alpha LX Workstations and Servers. LX provides the most cost-effective platform for Tru64 UNIX and Linux. Prices start at $1,995." Yes, an Athlon 600 is cheaper than an Alpha 21164 533 MHz, and about half as fast. I am afraid there is a little more to porting an OS than having gcc for the processor, especially as FreeBSD is not optimized for portability. FreeBSD will be ported to the Itanium if people want to make the considerable effort. So far the Itanium has been all promises and no substance. Paul Smith > -----Original Message----- > From: owner-freebsd-questions@FreeBSD.ORG > [mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of Allen Landsidel > Sent: Saturday, 28 October, 2000 08:55 AM > To: Heredity Choice; Nick Slager; jadream > Cc: freebsd-questions@FreeBSD.ORG > Subject: RE: FBSD & Itanium? > > > At 07:48 10/28/2000 -0700, Heredity Choice wrote: > >Predictions I have seen suggest that the Itanium will be nowhere near as > >fast as the DEC/Samsung/Compaq/IBM Alpha, but will be much more > expensive to > >manufacture. > > > >IBM has an Alpha prototype running at 1200 MH. To do the same amount of > >work, an Athlon would have to run at about 2400 MH. > > Just as a little FYI, here are come current numbers. > > 21164 Alpha @ 600MHz : 18.8 SpecInt, 29.0 SpecFP. > 21264 Alpha @ 575MHz : 25.9 SpecInt, 40.9 SpecFP. > Athlon @ 600MHz : 28.0 SpecInt, 22 SpecFP. > PIII @ 600MHz : 24.0 SpecInt, 15.9 SpecFP. > > Those numbers are for BASE processors.. The original K7 Athlon, and the > Katmai core PIII. > For the Thunderbird I wasn't able to find any scores, and the other P III > scores for the coppermine core aren't that much different from > the Katmai core. > > The 21364 Alpha (not in production yet) is running at 1.2GHz, > SpecInt 75.0, > SpecFP 160.0 (estimated speeds.) > That is probably the Alpha you're talking about, and it's a ways off in > production. For timeframe purposes, it's pointless to compare it to any > x86 machine available now, as by the time that is out, the > Sledgehammer/Clawhammer (K8-64bit) processors, as well as the > Willamette/Foster/Etc (IA-64 - P7) processors should be out. > > You have to always keep in mind floating-point vs. integer math as well.. > other considerations such as price/performance etc are kind of > moot I guess > here, considering the absurd direction this email is taking.. > > > >The lineup of Alphas currently produced range from a Microway workstation > >for about $2000 to a Compaq cluster being developed for the U.S. > Government > >which will be the most powerful computer in the world. > > A Microway 21264 workstation (single processor) with a 667MHz > processor and > 128MB of memory is $8,500, not $2,000. A dual 500 21264 is $13,000. On > that note, a 1GHz Athlon system from the same company is $3,300, with > 512MB. So, maybe yes an Athlon system will have to be twice as fast MHz > wise as an Alpha to be equally as fast in benchmarks, but who cares if it > still costs only 1/3 as much? To equip the Alpha with memory is > not cheap > either, considering that it's nearly 3x as expensive as PC133 memory for > the Athlon. > > Currently #1 system is the ASCI White at LLNL, 12 Trillion > ops/sec, and is > a cluster of 8,192 RS/6000 processors. Plans to extend this to the 30 > Teraflop range are on track for next year, with plans for 100 Teraflop > range for ~ 2003. > > www.llnl.gov/asci for more information. > > The machine you're talking about is ASCI Q, and as usual with your > statements in this email.. does not exist yet. It is not going to exist > until 2004 (scheduled), and is going to be roughly the same speed as ASCI > White in 2003. > > http://www5.compaq.com/hpc/tsn/iss017/hptc_iss017_fa.html#1 > > > > >The best news is that most of the single-processor Alphas already are > >equipped to run FreeBSD. Apple's Darwin, with its microkernel, > should be an > >easy port to the multiprocessor Alpha and the Alpha clusters. > > > >I have just acquired an Alpha EB164 for 5% of its cost new 5 > years ago. The > >only hardware incompatible with FreeBSD was the video card, which has no > >driver in XFree86 4.01 and has swapped places with the card in my Windows > >box. From the same period, a Deskstation Raptor, a DEC XL or a > DEC XLT will > >not run FreeBSD. > > All in all, this was a really long and stupid email that had > nothing to do > with answering the simple question asked, which was just about IA-64 > support in FreeBSD. I would say this, if Linux is going to support it, > then it'll be a compiler option in gcc. If that is the case, > then FreeBSD > will be supporting it as well, unless something drastic happens > in the near > future. > > -- Proud owner of a 600MHz Athlon, 512MB of memory, all Ultra/160 SCSI, A > 19" monitor, a GeForce256, and other misc options that all together still > cost me about 1/4 what a similarly equipped Alpha would have cost. -- > > > > > -------signature file------- > PGP Key Fingerprint: > 446B 7718 B219 9F1E 43DD 8E4A 6BE9 D739 CCC5 7FD7 > > "I don't think [Linux] will be very successful in the long run." > "My experience and some of my friends' experience is that Linux is quite > unreliable. Microsoft is really unreliable but Linux is worse." > -Ken Thompson, Interview May 1999. > > http://www.freebsd.org > FreeBSD - The Power to Serve > > http://www.rfnj.org > Radio Free New Jersey - 395 streams - 96kbps @ 44.1khz > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?005001c045c4$8b05bed0$15c6ddd1>