Forgot your password?
Apple Hardware Technology

Osborne 1 vs. IPad 2 249

Posted by CmdrTaco
from the there-actually-is-only-one dept.
On Saturday we ran a story about the 30th Anniversary of the Osborne Computer, and today we have an amusing head-to-head: Osborne 1 vs the iPad 2. StormDriver starts: "At first, they seem to belong in completely different weight categories. Osborne 1 is just under 11 kg, enough to pull your arm out of the socket, if you're a skinny geek. That's roughly 20 times more than an iPad, or about the same as whole suitcase of them But what about the processing power? Osbourne 1 was sporting a Z80 CPU, running at a stunning frequency of 4.0 MHz. You cannot compare the different architectures directly, but iPad's CPU is a dual core A5, clocked at up to 1 GHz. That's approximately three hundred times more, not counting in the vastly superior architecture. Z80 CPU was supported by whooping 64KB of system memory. Surprisingly, it was enough to run databases, word processors and complex, professional software. Today's iPad is equipped with 512MB of RAM (roughly one thousand times more), and some reviewers complain it's a bit on the low side."
This discussion has been archived. No new comments can be posted.

Osborne 1 vs. IPad 2

Comments Filter:
  • by Anonymous Coward on Tuesday April 05, 2011 @10:42AM (#35720472)

    Well, look at the order:
    Rubber tires vs wooden.
    Model T vs 2011 Kia.
    LEDs vs Candles.

    I just assumed that the Model T was considered the obviously better one.

  • Re:1,000? (Score:3, Interesting)

    by hattig (47930) on Tuesday April 05, 2011 @10:50AM (#35720604) Journal

    Indeed, it's 2^13 (8192) times more.

    And the RAM in the Osbourne 1 was probably eight 8KB chips, whereas (IIRC) it's two 256MB dies in the iPad 2, on the same chip as the CPU and GPU and more.

    But in the end magnitudes are all that matter when the differences are so massive. A Z80 took between 4 and 11 clock cycles to perform an instruction (8 or 16 bits typically) - let's say 0.1 MIPS/clock, whereas a 1GHz ARM A9 can do 2.5 MIPS/clock. That's 25 times more instructions per clock, and 250 times the clock, and twice the cores, and then we have to consider the ARM is 32-bit - so you need even more instructions on the Z80 for 32-bit operations. It's probably not too far off 20,000x faster to compute something on the integer cores of the A5 than on the Osbourne's Z80 - and that's before we consider the Neon vector units, the dedicate hardware for security, graphics, video, ...

panic: kernel trap (ignored)