25 posts • joined 13 Aug 2008
The king is dead; long live the King....
Although IBM isn't really dead, they certainly need to make their recent investment start performing in the market. The way that IBM laid-off so many, so quickly, they won't be able to respond to a rebounding market very quickly.
With the disastrous investment in Itanium evaporated, HP has made a marvelous recovery with Intel's x64 Windows servers. In my shop, the Windows servers are still prone to problems that make them difficult choices for long-uptime medical applications. Yet HP is doing great with that configuration.
Keep an eye on the market, though. Powerful undercurrents bring change from all directions.
MSNBC and FOX News had announced that "all android platforms together" outsell iPhone, at this time. Clearly, iPhone has a larger installed user base, though. The key question is whether flash-less iPhone will simply churn its existing clientele with the new release OR whether they will tap into new populations with it.
Android 2.2 came out this week and has Flash 10.5. It also runs non-flash existing applications over 100% faster.
The market will show whether Apple continues its fabulous growth OR if there really is a competitor now that sells more new phones than Apple.
We did stop NASA from further manned-space exploration after Skylab. The USA economy wasn't that strong in the 70s. The presidents of the 80s never had much use for manned spaceflight. Certainly is got larger in the 90s, culminating in the ISS from 1998, completed in 2011. We are expecting its planned-deorbit to be move out from 2015, but that is not definite yet.
Clearly, NASA rebalanced itself towards unmanned spacecraft. There were, after all, a history of unmanned probes to the Moon, Venus and Mars before Apollo took wind. NASA has had unprecedented success with Pioneer, Voyager, Galileo, Magellan, Cassini, Martian Rovers, Phoenix, MRO, and New Horizons is enroute. There is substantial leadership and technology-gain from these pursuits.
The message that is not penetrating is that our technology growth favors robotic missions, for the most part. I do agree that a new generation of boosters is needed, but Ares was (at best) a warm-over of Space Shuttle boosters with little scientific or economic advantage.
Taser use in the USA
Police in training in the USA get experience in the use of the Taser. Depending on the state, 60-80% of police cadets volunteer to be hit with a Taser as part of that training. From what I am told, it sucks really bad for about 10 seconds and you remain very sluggish for 1-3 minutes.
Taser use is allowed when taking someone into custody has deteriorated to a physical 'activity' (where tackling, nightsticks and other techniques might otherwise be used). It can also be used in a vehicle pull-over if the officer deems it necessary. Tasers definitely place the safety of the officer over the safety of the apprehended person.
There have been deaths with Tasers, but they have been where the person was quite frail or has flame-accelerant on their skin. There have been far more deaths from nightsticks and physical constraint of people. Messing with the police is never for the frail.
More mixed terminology...
Ok. I went to the TPC website. IBM is doing a one-server, 8 socket, 4 core/socket "turbo" 4.14 GHz server. Total is 1 server, 32 cores, 128 threads, score 1.2 million.
That is a max-configured version of 780, with half of its cores turned off and raising the clock-rate.
The Sun Server to which the author compares it to a 12 servers, 384 cores, 3072 threads cluster. Score was 7.6 million.
You only compare clusters to single servers when they have lower costs. The 12 cluster configuration is hideously more expensive, its TPC/$ is 342% of IBM's and the maintenance of a 12-cluster, 384 core systems is stupefying compared to a single 32 core system.
The Dell machine cited has less than 9% the throughput, so its TPC/$ was higher.
I like to see clearly written stories on this kind of topic, but this one was not-so-clear.
Power 7 sockets and cores
I agree with the coward that the terminology is a bit vague. "Two four-socket Power 7 processors" is mixing metaphors.
Most Power 7 'sockets' have 6-8 cores. IBM 770 and 780 servers have four sockets in each 4u 'block'. A given 770 or 780 can have up-to-four of these blocks plus expansion I/O drawers, if needed.
On top of all that, sockets on 780s can be configured to turn off half their cores and raise the clock rate on the remaining ones (not all sockets have to do it at once, either).
Probably, the writer's language meant that the 780 had two-blocks. Each block with four sockets. Perhaps with raised clock rates, too. That is about half of the max-configuration permitted by IBM for the 780, so it is a curious benchmark to publish.
Everybody form a circle, hands to the right...
Wow, fiction doesn't come together this fast by accident. X-37 isn't a gunship. There are even pictures of it on wikipedia. It might deliver a payload (not clear) but its main function is to get into near-LEO quickly and cheaply.
It isn't about to leave junk in orbit, nearly 30% of ALL the junk in LEO are the smithereens from that horrible Chinese satellite-explosion test.
The German thingie ran on a boiler...
Geez. If the USAF is catapaulting a boiler into space to do some 50,000 mile mission then we should fire those Generals.
The WWII stuff was trying to do the same stuff as the NASA stuff, but is isn't anywhere near the same technology.
There was a move a few years ago called Wild Wild West (from the 60s TV show) where they show how one could have done modern things with the steam-engine technology of the mid-1800s. The WWII stuff compares similarly to the USAF stuff.
The Obama adversaries are flourishing here...
I agree with Obama. The revisit-the-moon logic was stupid. Space stations are where-its-at for the next decade. The new-capsule for LEO-insertion was usable, but the Ares manned-booster was too-little-too-late (and too-expensive). The USA doesn't have the kind of money-in-the-bank to bankroll a 20X-100X uptick in NASA manned-activity now. We DO need to reduce military and civilian spending for a decade to pay down the horrible situation we are in today.
Getting to Mars requires interplanetary rockets that are nowhere-in-sight. Current technologies could probably get a crew there in ~1.3 years (Mars/Moon distance ratio=55M/.238M, 3 days transit for Apollo, a little faster now). Martian blastoff would have to be 4-10 months later (or else the flight-time will be much-longer). The return flight would also be 1.3 years. The whole thing would take 3-3.5 years end-to-end.
WAAAAY too long and waaaaay too expensive. We can get there with good planning (which we haven't had for a decade or two, or three).
Many ulterior agendas here...
Unless someone wants to argue that the species needs to get back to the Moon, then we have no arguments there. It is nonsensical to build Moon bases to stage towards Mars. Space stations make far more sense. I can't tell if the Cognoscenti believe that we should use an asteroid instead of a space station, though.
Interplanetary rockets are clearly a decade away from any 'build phase'. I agree that spending NASA's scarce resources on heavy lift with current technology is unwise. Sorry about the job-losses, though.
Privatizing LEO space lifts is an interesting gamble. Since it is the path, let the competition begin!
Isn't this a technology decision?
Too much stuff here is largely fiction. X38 landing capability was extremely limited (only on extremely-flat ground) and used a parafoil to soak up speed after re-entry cooled. It was icky.
The shuttle cost too ever-lovin' much. That is proven. The ridiculously heavy reusable has been too trouble-prone to be cost-effective. It is well-past its 'sell by' date, too.
The USA evidently don't have anything that competes with Russia's 1960s-era Soyuz for cost and reliability. NASA attempted to define 'human rated' lifting systems (rocket plus return-device) to establish new high-reliability standards. NASA just didn't get enough funding early-enough [or implement an efficient-enough development plan] to deliver any solution today.
Let the Chinese and Indians (and the French and the English) go to the Moon. There is little there for commercial exploitation. It has almost-no military value, either [other then ransoming the alteration of the moon]. Parking stuff in the Moon's gravity-well for interplanetary missions is just stupid. No one has the interplanetary thrust systems yet. Some things (like ion-rockets) have shown promise, but need a lot of work.
Space stations DO have value. If commercial lifting systems can bring enough technology and reliability for the price, they are the USA's first choice. NASA needs to decide what its role and support can be in that situation.
For now, NASA is about launching unmanned probes with technology unmatched by any other space-going nation. If we want NASA to resume heavy-duty manned-missions we will have to replace some of the current emphasis.
No more growth for Itanic.
Everyone here can agree that the loss of Windows and the decline of HP/UX leave no growth markets for Itanium. It could take a decade, but the sunset of Itanium is certain. Now we have to see if the PC/Power discussion is another sinking ship. Desktop PCs may be on the wane, smart handhelds (and various thin-clients) are clearly encroaching on their market.. If you believe in cloud computing, user-configured servers are also on the wane.
We can only guess where this goes over time, but expecting x86-x64 PCs and Power servers to dominate to market would be ignoring these leading indicators.
Wow, lots of spinning in the first few sentences. It looks like IBM is trying to capture the larger-server business. Perhaps they expect it to be high-margin business.
What I don't get is how-well Windows (or Linux) will leverage all the additional CPUs and threads. Hopefully, there will be benchmarks soon.
What else does it do?
Unsure about this note. Does the firmware update do anything useful to the customer. The only item listed seems not-useful to the customer.
Do they give Trumpets to everyone?
Another exciting day to hear the trumpet blasters. First off, HP calls the Integrity 'Superdome' (http://h20341.www2.hp.com/integrity/w1/en/high-end/integrity-high-end-servers.html) so someone can just SHUT UP about IBM being condescending with that.
Complain about the benchmarks if you like, but they are fair ways to see if two things perform the same.
I LOVE to hear about folks comparing chips to systems. Some people are comparing IBM 750 systems (32 cores lists at $190KUS with 192 GB of memory) to a handful of chips that can be assembled onto a Supermicro motherboard (street price ~$10K). Those are not the same thing at all. The Supermicro cannot hold 192GB of memory, the Supermicro doesn't have the uptime and partition-ability of the 750. It would be amusing to see what kind of benchmark scores the Supermicro would have, they would be less than the x64 machines above. You cannot even get eight 8Gb SAN cards and four 10Gb Ethernet cards on any Supermicro that I have ever seen.
With all of that, IBM clearly is still charging more. I just haven't seen a credible report with apples-to-apples performance. This article certainly isn't pretending to do that, either.
People that compare VMWare with PowerVM just don't understand what PowerVM does. PowerVM is firmware-assisted for both Ethernet and Fibrechannel operations. You implement NPIV so that the WWN follows the partition when you move it about. You implement HIE to the VIO servers so that twin VIO servers share the low-cost 10Gb Ethernet ports. You implement internal-VLANs so that firmware can route traffic once it is inside the server to whatever-host needs them. There is nothing like this (yet) in VMWare. Multi-CPU clients are highly firmware assisted in PowerVM. They are high-maintenance issues for VMWare. Perhaps VMWare will get there, but VMWare must invent everything as CISC software.
Lastly, power. It is clear that many datacenters have their backs against the wall on AC. They would prefer to stretch-out overtaxed infrastructure by purchasing devices that use less power, even if they will only last 24 months. It just isn't a good choice in the long term.
IBM knows that they are not the only ones selling Nehalem. They don't want to have non-competitive machines at the top of their x64 machines. It is troll-like to assume that they would.
Power7 is about UNIX. Nehalem is about Windows (with some Linux, but its market penetration is small, compared to Windows).
Until the 16 socket windows machine comes out, it is troll-like to behave as if you know how well it will run. With 4-8 cores per socket, that would give 64-128 cores. Do you believe that Windows is ready for that? Not even Linux pretends to be ready for that.
Too much of your facts are actually value judgements that can easily be contested, yet you state them as irrefutable.
So you think an Asus board will beat P7?
I love it when people compare Intel servers, built with commodity hardware against enterprise-class Intel or IBM servers. NO ONE will take stock parts and come anywhere near the performance, uptime or maintainability of the Power 7 (or a similar Intel server).
These machines have special inter-processor communications, special memory configurations, special I/O designs and special north/south bus configurations to give extraordinary throughput.
If you compare a four-core blade server against a six-core Power 7, you could gloss over the speed gap with a cost ratio of 10 or 20 to One. The Intel box wins all of those sales. If the machine needs to stay up 350 days in a row, Intel doesn't win those sales. The first market is MUCH larger, the per-server profit is MUCH larger for the latter.
IBM also offers a special partition-only cut-down AIX that provides services to other AIX/Linux partitions within the same hardware-machine. This software (called VIO) enables SCSI or FC sharing, partitioning of hard-drives (so multiple systems can use a 300 GB drive for ROOTVG, for example) and other hardware sharing. VIO certainly takes MUCH more than 1% overhead, depending on how much hardware infrastructure can assist VIO. VIO gives some of the 'software hypervisor' style magic without being a hypervisor at all. Some functions that are VIO-only on Power 5 have gone to firmware assists on Power 6 (don't yet know if more got there on Power 7). Shared Ethernet Adapters used to be high-overhead in software-only Power 5, but don't involve VIO at all in Power 6 (where did the overhead go?, it didn't ALL go away). Some think that VIO-itself is a stopgap until IBM can push as-much-as-it-can into the hardware virtualization.
PowerVM is firmware for IBM Power 5, 6 or 7 systems. It comes in three levels, the highest includes Dynamic LPARs plus an x86 software emulator for running Intel-Linux on Power. IBM's UNIX also has a Solaris-like 'software hypervisor' that permits booting multiple AIX/Linux images under AIX. Those AIX/Linux images have special code to support drives/NICs/queues/etc. in this environment.
For performance, PowerVM is well better than VMWare. You lose about 1% of each non-dedicated core when using PowerVM (the firmware redispatches such cores 100 times a second). There is a VMotion-like function called LPM (Live Partition Mobility) that permits movement of any-size partition from one hardware platform to another, without users getting signed-out. Power-architecture also permits Concurrent Microcode changes, so system/HBA/whatever can be kept up-to-date without downtime.
Cool stuff, but Power systems start at $8K and go up into the Millions. Will they compete-forever with x86? Geez, if I knew that, I wouldn't be wasting my time reading tech-web.
There is a lot of supposition going on here. I can help a little on the Power 7 stuff:
- IBM had advised that Power 7 servers were a March intro for several months. Moving the announcement up to February is a surprise.
- The allegation the IBM will bring out small machines with defective cores FIRST does not match with IBM open announcements. IBM set expectations that mid-high servers would come earliest.
- IBM's 4 GHz RISC core will well-outperform INTC's 2 GHz RISC core. Since IBM packs 8 on a chip, it would be reasonable to expect that 8 *4GHz is four times the throughput of 2*2GHz from an Itanium chip. We will see.
- IBM is announcing SYSTEMS and INTC is announcing CHIPs. There isn't much doubt about that. Itanium built-out Systems should be announced quickly after INTC's announcement, though.
- 32nm or 25nm Itaniums have not yet been announced by INTC. There are IBM and INTC announcements for 32nm chips. INTC will be first (6-15 months). IBM will be later (24-36 months). This is only relevant if it affects high-core offerings or high-GHz offerings that deliver important performance advantages.
- Itanium hardware virtualization hasn't appreciably improved in over six years. There needs to be INTC guidance for that. IBM and INTC 32-bit hardware virtualization has increased significantly/dramatically in the past five years.
- Oracle matters. We will have to see what they do. They are 'playing defence' with all the Sun stuff that they bought, though.
- Even Itanium fans are on edge with this delayed intro. Itanium needs a smooth launch to have any chance at a further 5 year's existence in the CPU-wars.
Gagging on self righteousness...
Wow, let us review what actually happened. US trustbusters clear the deal in 45 days, so it goes to the EC. The EC the holds down the deal for six months while they extract new promises from Oracle to deliver (for free) MySQL facilities for European companies. Promises that Sun never made.
OK, I got it. When is that next EU-based merger going through the US. I feel a 12 month vacation coming...
NASA, Russia and the ISS
The USA is developing a human-rated (recent term!) rocket system that is MUCH less expensive than Shuttle. Support, consumables, etc. cause Shuttle launches to be over $200MM (some say $700MM) when Russia can launch for $30MM. The Russians have cornered the market for economical launch systems (congrats).
The tight economy has accentuated the need to pursue this path. The USA will do its best to give the Russians a run-for-their-money as quickly as we can.
Whaddaya want from the SEC?
The SEC is not in charge of Sun running its company well. This IS Capitalism, ya know. The SEC is only about following a process of reporting what is being done. I suppose that you want the newpapers to be responsible for preventing crime too, eh?
Saved by the EC?
My goodness. The EC is saving the world from some scary Oracle slaughter of MySQL. The longer this takes, the more MySQL support staff will be lost. It could easily be that 50% of the EOY 2008 support staff for MySQL won't be there when the EC does its approval.
Good thinking there...
We should probably use the EU vehicle, instead
Where was that picture that I saw....