Oracle to invest in Sparc iron, clusters
If you were expecting a huge amount of detail on Oracle's plans for Sun processors, servers, and storage products at the five-hour mega-event held in San Francisco today, you'll be disappointed. But if you're a Sun customer, you'll be relieved to know that Oracle at least said it will invest in Sparc and x64 servers, storage, …
Thank goodness we have TPM to condense the essentials of the 5 hour Oracle telethon for us!
All in all Oracle appears to be taking a sensible approach to Oracle-Sun and the complete systems and 'gold standard' ideas are noteworthy.
Investment in Oracle-Sun and clusters may go some way to reassuring customers.
Would be good to see the skinny map no matter how skinny!
Fujitsu is key to successor SPARC64-VIII 8-way 'Venus' or another and the market will need clarification on this given solid Intel Nehalem and successors and IBM POWER7 solid performance and beyond maps. Unless Oracle Microelectronics intend to go it alone after SPARC64-VII? A joint processor approach may save 'Venus' that currently does not look well enough positioned to take on Intel and IBM's POWER7/8. Oracle should perhaps have made more of the Fujitsu partnership in the presentation, after all, Fujitsu saved Sun's credibility at the high end. Will be interesting to see if Oracle Microelectronics have SPARC plans of their own, independent of Fujitsu?
Customers as well as both companies may greatly benefit from an even closer working and cooperation between Oracle and Fujitsu. Both are most capable diverse companies. Both have tremendous unrealised potential. There is a good fit for cooperation and both companies need all the help they can get if they are taking on IBM, HP and others.
At least the major Oracle-Sun doldrums for customers should be over now that the Oracle-Sun deal has closed and yesterday's clarification should help customers be more certain.
Time will tell if Oracle's strategy is successful. With Oracle-Sun at least it ensures what can be saved can be saved and its good to hear of Oracle's investment commitment to Sun R&D and new staff.
Larry Ellison now turns his attention to the 33rd America's Cup Challenge on 8, 10, 12 February in Valencia, Spain. Certainly an interesting and perhaps an historic match?
Not much mention of the lower-end of the x86 range, even if used for clusters ...
although this came out from Sun today: http://uk.sun.com/emrkt/2010-q3-intel/part3.html
"Leave it to Dell" is fine - if only Dell or HP kit was as easy to manage (what do you mean, remote media or remote blablah need extra software licenses? Should just be bundled if you sell a server with "remote management features")
"[Y]ou'll be relieved to know that Oracle at least said it will invest in Sparc and x64 servers, storage, operating systems, and other technologies. That's something that Sun itself has not done in a very long time."
Huh? Sun invested billions in Millennium, Rock and other projects. Just because those projects were ill-advised--and then on top of that suffered due to spectacularly poor execution--doesn't mean the money wasn't invested.
Even after this lost decade, Sun still has some of the best engineering talent in the business. Maybe the Oracle boys will finally do what the old regime couldn't or wouldn't: finally put a stop to the infighting, the self-delusional project management, and the perpetual research project atmosphere.
I'm glad to see Fowler at the helm. But he's still got a long row to hoe.
The big disappointment is confirmation that there is no SPARC64 VIII (Venus)
There will be minor updates to the current chip and no plans to have a 45nm chip which Power7 and Nehalem EX will have soon. Not sure if there will be a SPARC64VIIIfx but that is only a HPC chip anyways
Sun needed to decrease the number of systems but cutting 50% of models immediately?
Sun fired most of their sales force last year, now it sounds like Oracle just fired all the business partners. How are they going to hire/rehire enough people to make up for the people that left for HP/IBM/NetApp/etc...
How does the chart say $100M of Exadata pipeline then Larry says $100's of millions of pipeline? Big jump in just a few hours. I bet HP is still sore about Larry killing V1 after only selling 25 boxes.
If they are killing the x86 product line how are they going to afford the development costs for the x86 systems in Exadata? You need volume to pay for the large fixed costs of development.
All in all it sounded like Oracle is only interested in the hardware to try to squeeze as much profit from the install base as possible as they slowly exit most of the hardware.
The only thing they seem interested in is the T processor which is going to be 16 cores => 16cores * 4 sockets * .5 factor * $47K/license = $1.5M Just the EE cost will pay for the $100K box. Maybe they will bring back the free oracle for sun box program but turn it around to free T box with new Oracle licenses.
Sun Cloud ==> dead
Sun loaner program ==> dead
Sun quick ship with distribution centers ==> dead
Fujitsu SPARC64 VIII ==> sounds a lot like UltraSPARCV
Sun volume x86 ==> dead
Sun x86 systems with windows ==> dead
Come on Matt you ass!!
Tell us again about the firesale for hardware.... Please, please provide a gibbering rant reaching several thousand words for your reply as well!! We're coming out of recession and this at the same time, it's like summer!
humble pie, humble pie, humble pie! How does it taste? :-)
SP&L? Hee hee hee... :-)
1) Fujitsu SPARC64 is end of life (very apparent in the presentation - no SPARC64 VIII)
2) Oracle will bring out another T chip, but regardless it will get crushed by Nehalem.
3) MILK MILK MILK those Sun customers as Oracle cuts people/partners/systems/projects/clouds/raises prices similar to WebLogic (+47%)
4) Oracle wants to own everything in a customer and sell only appliances, well the only reason they grew like they did in the past was because they were open. Safra knows that and will ensure they don't lose that to try and sell only double digit number of systems.
5) Larry will quickly get tired of the hardware business and the build to order model is the first move to get out of the hardware business. Expect Dell to make the hardware in Exadata V3 just like Teradata uses Dell systems.
6) "every employee needs to emotionally resign from Sun. Go home, light a candle, and let go of the expectations and assumptions that defined Sun as a workplace. Honor and remember them, but let them go." Ponytail
7) Ponytail got his $20M parachute...To all you people waiting for your package too bad. Now you will get quotas that are twice your current number and will be FIRED if you don't make your numbers.
8) Don't forget to order your t-shirt so you can remind yourself of the future of Sun
http://www.cafepress.com/SunRIP.428321275
baibai
Yeah, I do some work every now and again, don't look so shocked. :)
The biggest joker in the whole pack was Phillips admitting they are moving to a direct model, hence allienating all those Sun resellers that have slaved away at trying to keep the Sun installed base loyal. I'm not sure how it is in the US, but here in the UK the vast majority of the top accounts Larry is dribbling over attacking are served by multi-vendor resellers, and they will promptly switch their focus to IBM or hp rather than let Snoreacle go direct and steal their dinner. Those resellers have always owned the relationships with the customers, not Sun.
And how is Larry going to go direct when he doesn't have a salesforce skilled in direct enterprise solution selling? His current salesforce just flog Oracle licences and the Sun salesforce has been so reduced it's more like a salesnonforce! So who's going to step into the boardrooms and try and flog an "integrated stack" when they can't talk intelligently about even half of it? By the time Larry gets a salesforce with those skills assembled hp and IBM will have carved up that list of high-end customers between them, and will have been busy fighting with Dell over the low- and mid-range customers Larry just lost when he said x64 doesn't matter.
I must admit one failing of my own though. I did used to think Larry was a smart man.
/SP&L all the way to the pub!
I have no problems with your post. As long as you dont state incorrect things as facts (like Niagara is slower than POWER6), I have no problems. You present your beliefs and thoughts. Fair enough. I do not agree with you, but that is another thing.
Opinions are never wrong, but facts can be wrong.
> Opinions are never wrong, but facts can be wrong.
That sure helps me understand a number of your posts now. Especially where Sun T-systems vs. IBM Power based systems are concerned.
Sign up, sign up for Blocks and Files, The Reg's weekly storage newsletter
