replacing CPU on laptop

I'm trying to customize a Toshiba Satellite U500-ST6344 which will arrive with an intel i3-330m. in the future i would like to upgrade it to a i5-520m.

QUESTION: is the CPU soldered to the MB or does it have a socket that you put the processor in and tighten a screw or lever?

link for details on Satellite U500-ST6344: http://www.toshibadirect.com/td/b2c/cdetland.to?poid=467540&tab=customize

PS. If anyone owns a U500 series laptop please chime in on the battery life. i plan on getting the 12 cell. how many hours of use am i looking at?

THANKS

 

No, its not soldered, it's socketed. I believe it's PGA988.

 

I'll bet however that once you get familiar with the capabilities of an i3 vs an i5 cpu, you'll never make the switch. About the only diff is turbo-boost which is unproven to do any real good for real (non-benchmark) applications.

'standalone' Intel mobile CPUs are notoriously hard to find and the good ones cost an arm+leg, far more than when you order one built-in to a new laptop.

Buy what you need now even if it means you need to pause and save up for another month.

 

Yes, I believe the CPU on the U500 is replaceable. However, I would recommend against doing so, because you will likely not feel any difference in performance in every day tasks. The only times you would see a relatively large improvement are when the CPU is maxed out at 100% on CPU-bottlenecked tasks, such as video conversion.

A much better use of your money would be a nice SSD (very noticeably improves boot time and application launch time), or just saving up for your next computer

 

i would get the CPU i5-430m at least... it will be way better than i3 because of hyperthreading so u virtually have a quad...also , ur CPU will most probably be upgradeable but doing so will void ur warranty... ur really better off just paying $120 now and getting it upgraded by toshiba for u...

 

doesn't the i3 do hyperthreading as well?

 

oops my mistake... i3 does hyperthreading but has no turbo boost... anyways , i5 would be better with the extra power needed to play vidoes etc...

 

what 'extra power'? Remember for that every second an i5 or i7 gets into turbo boost the cpu will throttle back *below* the nominal clock rate to help cool things off.

outside of synthetic benchmarks, turbo boost hasn't been proven to do a damned thing.

 

Seriously for that price it is not worth it.
Core i3 already has enough power IMO.
Currently the prices are still steep if upgrading is a must wait until the prices fall.

 

Extra clock speed is the most obvious and direct way to get a performance increase, and that's all turbo boost is. Whether or not the CPU will need to throttle back is decided dynamically based on the temperature; if your laptop has decent cooling, it won't be a problem.

With Turbo, the i5-520M should be 25% faster than the i3-330M on two cores, and 37.5% faster on one. Whether or not that's worth it is up to the OP. I would say it probably isn't worth it, but I'm not the OP.

 

transitory clock speed increases will only speed up a system when the necessary data sets and instructions/code are already in ram ready to be slipped into the cpu cache.

people who are paperwork fans of turbo boost need to read intels white papers on the subject. even intel says that tb will show real-world improvements ONLY for workloads that are tuned in ADVANCE for the specific cpu and cache layout that will be running turbo boost.

Intel also says that it is trivially easy for synthetic benchmarks to game the 'rating' system by deliberately using workloads that are pre-tuned for turbo boost.

Hyperthreading has much the same problem. Run a lot of cpu speed increases including out-of-order code execution without pre-tuning the workload and you quickly run into things like cache invalidation that forces the cpu to get into MORE wait states as the memory subsystem tries to refill the L1/2/3 cache with needed data and code.

It's also why Intel is making a huge investment in compiler technology and contributing a lot of code to the GCC development trees. IBMs execution speed and i/o throughput advantage with their Power cpus vs the past three generations of Intels CPUs (including the Itanic) are mostly due to the superiority of the supporting compilers, not from the raw power represented by the cpus. This is also why HP has brought PA-RISC back from the dead and wishes that they hadn't killed Alpha.