|
On Fri, Nov 6, 2009 at 15:55, Thorbjoern Ravn Andersen <ravn@xxxxxxxxxx> wrote:Our core business legacy product is essentially disk-bound. The exact speed of memory and cpu does not matter much if they can keep up with the disks (which they usually can).
I was at a Java performance talk where it was mentioned that the x86
processors are SO fast these days that hundreds of instructions can be
done in the time it takes to get a single value from memory. Every
time. Having more cores does not help this.
Power 6 is still stuck with dual channel DDR2-Memory at 667 Mhz, while
Intel is currently using Triple-Channel DDR3 Memory at 1333 Mhz. Intel
also uses much faster On-Die L3 Cache (currently at 8MB in the
entry-level CPUs). Intel also uses a much more efficient 45nm process
to create their hardware, and the next generation to be launched in Q1
2010 will be in 32nm process.
IBM is planning to release Power 7 in 45nm process somewhen in 2010 -
at a time when Intel will already by bringing out 32nm CPUs.
IBM seriously missed the bus, and is lagging behind more every year.
There's a reason why Apple switched from Power to Intel - Intel isI believe the transition is not as much a technical issue but a monetary issue. Apple could get more value for their money with mobile x86 cpu's instead of being a grumpy customer in the IBM shop, where mobile cpu's are very low priority.
better.
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.