Price cut coming?
In May 2006, after the dual-core Power5+ chips came out late and didn't offer the kind of performance boost customers were expecting and the Power6 chips were pushed out into 2007 instead of their expected fall 2006 launch, IBM offered customers using System p 570 (16 core), 590 (32 core), and 595 (64 core) machines upgrade paths to the future Power6 machines that preserved serial numbers on Power5 and Power5+ boxes. The delays with the Power6 chips and a nine per cent drop in Unix server sales in the first quarter of 2006 also compelled Big Blue to cut prices on Power 590 and 595 machines.
IBM has not cut prices on its current Power6 and Power6+ servers, but if the upgrade guarantee doesn't grease the sales skids, that will be the next move.
In addition to talking about the upgrade path guarantee, Handy also provided a few more details about the future Power7 chip, which will be implemented in a 45 nanometer chip process and which will be manufactured in IBM's East Fishkill, New York, foundry. Back in December 2008, when discussing the "Blue Waters" supercomputer and its data center, IBM had confirmed that Power7 chips would have eight cores.
According to Handy, IBM actually plans to offer Power7 chips with four, six, or eight cores when they begin shipping in systems some time in the first half of 2010. He would not be more precise about the launch schedule, except that it would be a staggered announcement as was the case with the Power5 and Power6 rollouts, starting on a few machines and eventually spanning the line.
The Power7 chips will come with a variety of clock speeds, too, some designed to give the most performance per core and some designed to deliver the most performance per watt. The word on the street last year was that the Power7 chips would clock at between 3 GHz and 4 GHz, but IBM has not confirmed this (even now). Handy also added that the Power7 chips will support up to four threads per core, up from the two threads per core of the Power5 and Power6 generations.
What IBM did say is that depending on the chip, the Power7 variants will offer anywhere from two to three times the performance per watt of a Power6 chip at a given wattage. In the jump from Power5 to Power6 (which was actually two steps including the Power5+), IBM delivered about twice the oomph in the same energy envelope. The jump from Power5+ chips to Power6 chips was more like 50 per cent more oomph by my estimate, and the move from Power6 to Power6+ was more like 15 per cent.
Clearly chips with a plus don't fare well at IBM, since the Power4+ was not much to write home about, either. Anyway, provided that the energy envelope of the Power7 machines remains more or less the same, what this means is that customers can expect two to three times the performance.
IBM also confirmed that the Power7 chips will support DDR3 main memory; Power6 and Power6+ servers use DDR2 memory, and Power5 and Power5+ machines used DDR1. IBM currently uses double data rate InfiniBand links between the Power6 and Power6+ chips out to remote I/O drawers (which it calls 12X I/O), and it is not yet clear if IBM will boost this up to quad data rate InfiniBand with the initial Power7 machines, according to Handy.
Customers using the older Remote I/O drawers and their High Speed Links (IBM's funny name for Fibre Channel peripheral links to external I/O cages) will have to upgrade to 12X I/O before moving to Power7 chips, since Remote I/O will not be supported with Power7 servers as it was with Power6 and Power6+ machines.
On the logical partitioning front, the Power7 machines will be able to host up to 1,000 partitions per system, up from the current maximum of 254 partitions on Power5 through Power6+ machines. I have been told in the past that the Power6 and Power6+ chips could already, in theory, support 1,000 partitions, but for whatever reason IBM did not activate this latent capacity. Presumably the Power7 chips theoretically support something akin to 4,000 partitions on boxes using eight-core chips; IBM didn't say.
Handy said that IBM will patch its AIX 6.1 Unix variant as well as its i 6.1 proprietary operating system so it can run atop the Power7 chips and make use of its features. Some time in 2010, whole new AIX 7 and i 7 operating systems will be delivered by IBM, not only with full Power7 support, but presumably with other capabilities.
IBM plans to talk a bit more about the Power7 chip and server designs at the Hot Chips conference hosted by the IEEE at the end of August at Stanford University. We'll keep you posted. ®
IBM lifts the veil on Power7 chips
COMMENTS
Anonymous Coward
"Though, I have a feeling you're exaggerating quite a bit here since a "major UK bank" experiencing the frequent outages you describe would switch to something else ASAP."
Oh, there are politics involved. The best tech doesnt necessarily wins. How many uses sucky Windows, and never switch to Linux? In my company, large fortune 500, Ive recommended new technology that would make enormous savings and better up time. But to no avail. The management has decided on some technology and that's it. I can do nothing, despite showing better calculations.
@AC Not Found
The link worked for me, and just tried again and it works
" IBM Power 595 and 570 Servers
CEC Concurrent Maintenance
Technical Overview
April 14, 2009 "
ftp://ftp.software.ibm.com/common/ssi/sa/wh/n/pol03021usen/POL03021USEN.PDF
"I can say that popping ram chips, MCM's and RIO books happens so frequently that I'd love it if I could hotswap with the system on line"
Something must be wrong with your environment, we have a dozen of p570s here untouched for months and a couple with 1 year+ of uptime.
"overcommitted as per IBM's own best practices please explain how I remove 4/8 cores from a system already running at capacity?"
Another IBM best practice in these cases is to have Live Partition Mobility-ready servers, so you can move partitions easily without disrupting your apps.
Best regards! Cheers!
@AC 'Please, update yourself before posting. '
Same shit, different badge - you did notice I said 'rs6000/system P/pSeries' didn't you?
I use p570's on a daily basis, and I can say that popping ram chips, MCM's and RIO books happens so frequently that I'd love it if I could hotswap with the system on line. Yeah I know the new flexy interconnect helps coldswapping a node out, but with a virtualised load which is overcommitted as per IBM's own best practices please explain how I remove 4/8 cores from a system already running at capacity?
I followed your link in case I missed something from last weeks IBM sales propaganda sessions... I guess not...
ftp://ftp.software.ibm.com/common/ssi/sa/wh/n/pow03023usen/POW03023USEN.PDF
Not Found
The requested URL /common/ssi/sa/wh/n/pow03023usen/POW03023USEN.PDF was not found on this server.
IBM_HTTP_Server/6.0.2.13 Apache/2.0.47 (Unix) Server at ftp.software.ibm.com Port 80
@AC RS/6000 misinformed dude
Dude, RS/6000 brand is gone not yesterday but 8-10 years ago.
After that, it was pSeries and then POWER servers.
Please, update yourself before posting.
Try starting reading, this:
http://www-03.ibm.com/systems/power/hardware/whitepapers/595_570_cec.html
That is related to concurrent maintenance of high-end servers that ONLY 'mainframe' class systems are enabled to..
Best regards! Cheers!
'That means, you can repair live machines without powering it off.'
I nearly wet myself.
I can't recall the last time I did any system maintenance at a major UK bank that didn't require an outage on rs6000/system P/pSeries (still sounds like 'pessary' to me). pSeries stuff is so far behind 'mainframe' class systems for in-situ repairs. You still can't replace memory or cpu while the box is up which is such a PITA since the parts IBM sources fail constantly.
Oh you meant disk swaps? Yeah, you can do those without shutting an rs6000 down... usually.
