Tell us why this is a Land of Confusion
Help us, help Intel, HP and Sun
Reader quiz Given the tumultuous events in the server and processor worlds over the last couple of weeks, it seemed appropriate to take a moment and clear up a couple of nagging details surrounding the strategies of HP, Intel and Sun Microsystems. And to make things fun, El Reg is asking for reader help with these server/processor conundrums.
Houston, we have an Itanic problem
Getting straight to the point, HP has pulled its customers through one of the most confusing product launches of all time by throwing both Xeon (now enhanced) and Opteron at users within one week. HP did this while still claiming Itanium is A) thriving and B) an industry standard.
Let's be clear here. After more than a decade of marketing and development and years of delays, it's simply not accurate to declare Itanium a success just because Intel manages to ship 100,000 processors in a year. Itanium makes up the tiniest fraction of the overall server market - one eclipsed by the younger Opteron - and has largely been a laughingstock of the industry.
With this in mind, Intel and HP must go to great lengths to promote the notion that Itanium is an industry standard. Yes, a number of vendors sell Itanic servers, but HP owns close to 95 percent of the market. To help make our point, we turn to HP's server marketing manager Don Jenkins.
"We actually lost (Itanium) market share," Jenkins said, in a joyous tone, during a speech at last week's Intel Developer Forum. "The Itanium market grew 500 percent in 2003, and our growth was 400 percent. That meant our share fell."
This brings us to challenge number one. Can any Reg reader point to a single instance in which HP boasted about losing market share with any other product? Extra points will be awarded to those that can find evidence of HP bragging about losing ground in an "industry standard server" race.
And to prove, we're not alone in this quest, our friend at Illuminata Gordon Haff shed some light on the situation.
"That's been HP's great quandary for the past couple of years," Haff said. "Touting their great dominance and special position with Itanium runs exactly counter to Itanium being an 'open standard.' They've really never resolved that that fundamental contradiction - not that they really can."
What to call The Beast's bastard child?
Contest number two is just to the side of the Itanic - you know, the bit where the water is rushing in and customers' panicked screams rule. Intel likes to call it Xeon (now enhanced), while AMD calls it AMD64.
Linus Torvalds brought up the Xeon (now enhanced) naming problem in an e-mail earlier this week.
Actually, I'm a bit disgusted at Intel for not even mentioning AMD in their documentation or their releases, so I'd almost be inclined to rename the thing as "AMD64" just to give credit where credit is due. However, it's just not worth the pain and confusion.
Any Intel people on this list: tell your managers to be f*cking ashamed of themselves. Just because Intel didn't care about their customers and has been playing with some other 64-bit architecture that nobody wanted to use is no excuse for not giving credit to AMD for what they did with x86-64.
(I'm really happy Intel finally got with the program, but it's pretty petty to not even mention AMD in the documentation and try to make it look like it was all their idea).
Go on, Linus. Tell us what you really think.
The Linux man, however, makes a great point. What are we to call this thing from Intel?
Our suggestion uses the names of the Titanic's sister ships - Britannic and Olympic. If Xeon (now enhanced) is a failure, it will take the Britannic name, since the boat crashed like its most famous sister. If, however, the chip is a success, it can be dubbed the Olympic - the Flopteron killer.
In a perfect world, however, we'd like to see some Soviet themes thrown into the naming mix. This would help point out Intel and HP's see no evil, hear no evil approach to x86-64-bit computing. So, a plethora of points go to any reader than can pull of this boat, gulag, five year plan mix in style.
The software makers can't hear you, Sun
Sun enters the picture with its introduction of the UltraSPARC IV processor. The company bills the dual core processor as a single chip and expects software makers to do so as well. This would give Sun users a nice edge in per processor software licensing models, since both IBM and HP count their dual core processors as two chips.
Oracle, however, tells us that they count each core as a CPU, meaning Sun users are in the same boat as the rest of the RISC world. (But clearly not the same boat as the Itanic crowd, which won't see multicore chips for quite a while - Ed.)
As Sun moves to put tens of cores on each processor ahead of rivals, this should leave the company, its customers and software makers in a confusing state.
We'd like to hear from our readers about who should buckle first - the hardware or software vendors. How do you think a processor should be defined? Big points for the most coherent answers here.
We'll add up all the points in a totally arbitrary manner and award Reg kit to the top scorers. All questions must be answered to qualify. Well, except for the first one. That will be tough. Please send your answers right here. ®