bruggink@neep.engr.wisc.edu (Dennis Bruggink) (03/23/91)
Evidently 68040 Mac's were demo-ed at MacWorld in January; any word on any release dates?? Are there any Mac's (e.g. an 040-fx) even close to production? We are close to ordering IIfx's, but wouldn't mind a short wait for a CPU w/ ~double the speed potential (as long as price wasn't prohibitive); otherwise, should we buy an fx now, hoping an upgrade to an 040 is made available, and at a reasonable price?? --dennis bruggink [bruggink@ccc.nersc.gov or bruggink@vms.macc.wisc.edu]
jo0e+@andrew.cmu.edu (Jared M. Oberhaus) (03/29/91)
WAIT! According to my sources ;-) a new Mac in a Tower version is expected in June sometime. This will be a monster, with 5 internal SCSI connectors for the 5 internal drive bays, will accept 1, 4, and 16 meg SIMMs (for up to 65 megs RAM), and a lot of slots. With the 68040 at only 25Mhz, the damn thing will be 4 to 12 times faster than a 68030 IIfx at 40Mhz, almost twice the clock speed! The secret to this is that the 68040 is really a RISC chip with some CISC chip features (for downward compatibility) overlayed on the chip. Another new Mac will have 3 NuBus slots, a 25Mhz 68040, and be shaped like a IIci. Apple's plan (I think) is to lower the price of the IIfx (which has already been done in the Universities at least) a few thousand or so (it needs it!) and price the IIci-size 68040 Mac where the fx is now. The tower machine will probably be priced like a Compaq SystemPro (i.e., out of my price range!), but who needs 5 drive bays anyhow? Wait for the 68040--you'll kick yourself if you don't.
mslater@cup.portal.com (Michael Z Slater) (03/30/91)
>WAIT! >According to my sources ;-) a new Mac in a Tower version is expected in >June sometime. This will be a monster, with 5 internal SCSI connectors >for the 5 internal drive bays, will accept 1, 4, and 16 meg SIMMs (for >up to 65 megs RAM), and a lot of slots. With the 68040 at only 25Mhz, >the damn thing will be 4 to 12 times faster than a 68030 IIfx at 40Mhz, >almost twice the clock speed! The secret to this is that the 68040 is >really a RISC chip with some CISC chip features (for downward >compatibility) overlayed on the chip. I don't believe these performance figures. The 68040 is at best about 4 times the speed of an 030 AT THE SAME CLOCK RATE. A 25-MHz 040 is unlikely to be much better than twice the performance of a 40-MHz 030. Also, this "RISC chip with some CISC chip features" is misleading. The 040 uses many of the implementation techniques that were first widely used in RISCs, but that doesn't make it a RISC. With this type of statement, RISC becomes a synonym for "well designed." Michael Slater, Microprocessor Report mslater@cup.portal.com 707/823-4004 fax: 707/823-0504
6600dtam@ucsbuxa.ucsb.edu (Marc "Tae-Kwon" Tamsky) (04/03/91)
In article <40672@cup.portal.com> mslater@cup.portal.com (Michael Z Slater) writes: >I don't believe these performance figures. The 68040 is at best about 4 times >the speed of an 030 AT THE SAME CLOCK RATE. A 25-MHz 040 is unlikely to >be much better than twice the performance of a 40-MHz 030. >Michael Slater, Microprocessor Report mslater@cup.portal.com I heard that the 68040 was almost 10 to 50 times faster in FPU intensive applications.. Is this true? -- = Marc Tamsky 6600dtam@ucsbuxa.ucsb.edu 6600dtam@UCSBUXA (bitnet) =