Slashdot Mirror
Alpha Lives! But Who Will Market It?
chriton writes "The Inquirer is running articles about HP's and new "Marvel" server which will arrive Tuesday, Jan 14th and the expectation that HP will try to keep it's performance quiet. Not because it's bad like Itanic I, but because it's too good! It's built on Alpha EV78 processors connected by a switched fabric and promises blazing performance. "Marvel has, apparently some rollickingly good benchmarks that HP wants to underplay, just in case people start comparing the performance of the Alpha Marvel architecture with the Itanium 2 it also sells, and perhaps more importantly, the SuperDome machines." Alpha offers the kind of choice and competition the processor market will sorely miss when it goes. The FTC was sleeping when they allowed HP to acquire it."
no, compaq sold alpha technology to intel, but they still had it for themselves and were still selling alphas. When HP bought compaq, they inherited the alpha line.
and while they're great machines that perform well, they're very limited. It's difficult for us to get many of the applications that we use for the Alpha, and if the app is available, the vendor usually provides poor support for it. Sure you can compile OS software on the alpha, but the commercial world overwhelmingly uses traditional closed software. HP decided to stop production of the Alpha because they had a competing product (pa/risc) that was in higher demand. They even plan to eventually lose PA/RISC in favor of itanium, as the article mentions. As far as price goes, one of our clients purchased a wildfire gs320 because of the low price. They found that while it offers acceptable performance, it's very difficult and expensive to find the expertise needed to properly maintain this equipment. We run a primarily Sun shop not because it's necessarily the best, but because it's what everyone else runs, and thus easier to maintain and cheaper in the long run.
Go away, or I will replace you with a very small shell script.
IIRC, some of the associated technologies like the switching architecture and some of the NUMA features were not licensed but held by Compaq for their Itanium servers (to give them an edge).
Yey, 500th post, on a subject I like :)
I remember when Compaq bought the Alpha technology. I was invited to a demo for their new workstation machine, that was back in the late 90s, I remember the workstation they were demoing in front of everyone, nice audience, people that worked on the movie Titanic were there to explain how they used the alpha technology to render those huge datasets, manipulate large 3d models, etc etc...
They were so EVASIVE when people would specifically ask them to compare the Alpha Workstation to intel workstation. I mean everything looked professionnal up to that precise moment. Why on earth are you getting yourself in so much trouble to advertise your alpha workstation, invite people to costly hotel floor, serve them good food, etc, if you don't want to address the PRIMARY concern of your target audience? What "non-alpha" people (new customers) want to know is why would they go alpha if it's not for the proprietary software?
(In this case, Lightwave was one of the tools and it was cross-platform, every Lightwave users KNEW that the alpha crushed the PC in rendering, so hiding this fact looked very suspicious for this small portion of the people that were there. Then you add the fact they they didn't want to give any comparing numbers, being evasive and all. The only positive thing they mentionned is the FX32 emulator and the fact that they could run non-native software like photoshop in their alpha workstation. Now who the hell would buy a workstation like this if it doesn't show any appeal outside from the people that already know about it? If you say "3x faster rendering, only 1.5x the price" now there's an apeal! They didn't! How on earth are they going to gain sufficient marketshare with mouth-to-ear strategy, where amiga, for example, failed. With a CPU R&D buisness, you need a LOT of sales to cover you expenses, they had a bomb on their hands, and while I understand that they had to play nice with Intel, they could have thrown the bomb at intel instead of blowing up with it.
This is another situation where Money and Monopoly is bad for evolutions and revolutions, try to find ONE SINGLE alpha user that bitched about the architecture (before it got left out dying, obviously), make a percentage (you'll probably get something close to 0%), then compare that percentage with Intel users. Not that Intel technology is bad, but it sure isn't revolutionnary, heck I'm still waiting to get that 7505 chipset board with 2 2.8Ghz Xeon on it, everything is back order or N/A yet. If compaq would have had a clue, I'd have a box probably 4x more powerful today with win2k support and good driver support for about the same price... shame.
--- Metamoderating abusive downgraders since my 300th post.
As an engineer that worked on EV7 it is sad to see such a wonderful machine fall by the way-side. When the SPEC numbers do come out not only will all the world will see that Alpha is again the world's fastest processor, but that Marvel systems scale linearly. We'll all eventually go over to Intel, which a lot of us aren't looking forward to, and hope not to get laid off
Alpha's have always had awesome specs, hell I think slashdot started on a UDB (early alpha unit, small compact case, built in sound). The alpha processor has long been one of the best performing and worst marketed main processors in the history of computers.
The fact is that DEC wasn't in a posistion to market it, and when they COULD have sold the chip to use in apples (instead of PPC) they declined (morons). Compaq bought DEC and had NO clue what the hell to do. It took them almost 2 years to wrap their head around the fact that the alpha servers where the only profitable product they had. (See service support contracts and high margins for the high end alphas). By then it was too late, they were working on the merger with HP.
No HP's here, and doesn't want to compete with it's own inferior equipment. Lines are being drawn and you can bet that the superiour technology of the alpha will again suffer. Remember that the EV78 is an OLD alpha design and it still kicks ass. Compaq basically stopped developing the alpha series AGES ago. (the EV8 was supposed to be out early last year according to one of the early compaq alpha ropadmaps)
Too bad the alpha is dead. It is taking years for intel and IBM to come up with a chip that comes close to alpha performance. Good thing that they are competing against old alpha designs and the EV8 has been killed. Otherwise those darn pesky spec numbers would have been embarassing.
cluge
"Science is about ego as much as it is about discovery and truth " - I said it, so sue me.
In a rational world, management would own up to their mistakes (whoops, Alpha IS better than Intel) and work to make things right. When I was in a business simulations class (we were grouped into teams, and our "yearly" decisions as to the mix of funds devoted to r&d, marketing, production amounts, pricing, etc. for each of our respective companies, were fed into a sim every week), my team made the mistake of trying to eliminate an existing product line in favor of a more profitable "premium" product.
While it was more profitable, the market was actually bigger in the more mature market - something that none of the teams had taken into account. However, because our team invested heavily into reducing production costs (retooling, R&D into improving production efficiency and unit quality, strategic partnership with suppliers) we were able to shift some of our capacity back into the "classic" product, price it lower than the competition, and royally kick ass in the simulated market in the following year.
What does this have to do with HP? Well, if you have a superior product, one that will dominate for a pretty good while, and you have the sole source for it, WTF would you want to sell an inferior, lower-margin commodity product in direct competition with a whole boatload of competitors? I mean, isn't that what is killing SGI? The fact that they're trying to compete in the commodity market, but without a superior selling point (either tech, or price), they're getting hammered.
Florina was death to HP. I'm going to miss their R&D and their printer line when they go under, and only can hope that HP's board members never sit on any other company's boards in the future. Well, any company except maybe Microsoft...
What if some company, for example Red Hat bought the Alpha technology. Just think how a premier hardware architecture could be marketed along with Linux, which has huge growth potential.
If Linux is to totally dominate, Linux vendors need to come up with some better hardware.
Then DEC kind of died. They didn't seem to market. The tech was good, but no marketing, and some issues. Before the compaq merger, DEC sold StrongARM, and all it's fabs (aging) to intel, in return Intel was supposed to fab the next generation alpha chips, and was prevented by the FCC/court (or a combination) from aquiring alphas (due to anti-trust, not that that mattered to the DOJ when they did...) Intel did not fab the next generation (21164@smaller process and 21264s) of alphas. They claimed that they couldn't because the chip was too complex. (There is no evidence that they ever did, and this was just before the compaq merger)
Compaq acquires DEC. It takes it's time, but releases 21264s (fabed mostly by Samsung, and some supposedly by IBM), They branch off the alpha tech to API (Alpha Processor Incorperated) which sets the EV7 (21364 = (21264 core w/improvements + RAMBUS controller) development back. (additionally, MANY alpha engineers were hired by AMD when DEC was merging, and the EV6 bus (and many other tech goodies) were licenced to AMD (Slot A was originally an Alpha slot) for inclusion in Athlons (who still run on an Alpha bus)
Compaq decides to inhouse the developers again and sets EV7 back more. (EV8 is reportedly mostly on schedule) Then Compaq decides to sell the alpha tech (or much of it) to Intel (DOJ apparently doesn't care about anti-trust at this point) and cancels EV8 (which was reported to include Hyper-threading, multiple cores on a die, Onboard Memory controller (like the ev7) (pretty much every "cool" thing Power4, Intel, and AMD were planning on having.)...and was due out this year) EV7 is phenominally behind schedule. Finally EV7 makes an appearance, or will (asuming the article is accurate) as basically an EV6 core with tweaks and a RAMBUS memory controller onboard the chip (256-bit dual channel, so it actually isn't a POS like many RAMBUS inmplentations. (for comparison some RDRAM implementations have been 16-bit, many currently are still that or 32-bit)
Now, Alpha is slowly slipping, but currently (aside from Power4) the only chip holding it's own against Intel/AMD. (based on a several year old core at that) The EV7 will be the last generation of alphas, without all of the features the EV8 would have had (and probably the performance crown for a LONG time)
(compiled from memory, it is 1:30, and post errors/debates as responses)
The fact that they've already spent billions of dollars on it doesn't necessarily mean that continuing to push it is the best business decision. Sometimes you have to realize that you made a bad choice, and write off the investment. Otherwise you may just lose many billions more.
If I were to believe that everything was as simple as you make it, my conclusion would be that I need an Itanium 2 + Pentium IV + PA-RISC or POWER4 to have an overall better machine than an Alpha Server. Is that what you meant? Incidently, the reason the Alpha 21264 has 3 integer units (providing integer scores that dominate Itanic integer scores) is to keep the 2 fp units fed. I'm not an expert on the Itanic architecture, but I'm led to wonder if the Itanic integer units are capable of providing all of the array indexing and loop-counting chores needed for many floating-point numerical anaylysis algorithms. I'd really like to know an answer to this, since the Itanic appears so unbalanced when looking at SPEC scores alone.
Furthermore, you're comparing prices on processor cards for different systems using full retail price. Have you ever bought this kind of equipment? I know I've never paid full retail price when I have. And you don't need to use Compaq's pricing utility -- just get the model number and search with google like you probably did for th Itanic board. Unfortunately, for the cpu board (KN610-EB) you're describing, there aren't many links. Another problem is that the price you quote for the Alpha board includes additional cpu licenses for Tru64 (for the KN610-EB) or VMS (not sure -- KN610-EC?). I don't see anything at the link you provide which states that the price includes a Windows or HP-UX license (and don't kid yourself, you'll need a license).
It's not clear why you mention the 3MB cache on the Itanium. The EV67 boards (KN610-BA, for instance) for the ES40 Model II (which is old) have 8MB cache, and the EV68 board (KN610-EB, for instance) have 16MB of cache (note that these are the big L3 caches for Alphas, not on-chip like I think the Itanium is). Also, it's not clear to me that the processor boards include the same functional components -- do you know that they do (you already seem to have missed the OS license issue)?
You haven't provided the price for the systems, nor stated their default configurations. This is certainly important when making a comparison. You'll also want to compare memory prices, since ram for these servers sometimes has a special form factor and costs a bundle. Again, never use list prices for any of this stuff, as you can occasionally halve the price with a bit of negotiation. That goes for cars and good office chairs, too.
-Paul Komarek