Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Sandy Bridge vs. Arrandale
- Thread starter prism
- Start date
- Sort by reaction score
Not to sure, but according to activity monitor, with numerous programs open, Im sitting at 85% idle. So its not like I'm taxing out my CPU anyway. Im sure most people wont notice a difference. Everyone seems to just get worked up with the release .They seem to think that all of a sudden their MBP won't be able to do much.
Ok, thanks for being so vague but I guess it was up to me to formulate the question more precisely so here is my second attempt:
How much performance gain would SB bring to the table compared to a similarly configured arrandale system. Similarly configured as in same clock speed, same number of cores and assuming the same GPU?
When comparing similar clocked CPU's ( i7-975 vs the high end i7-2600k ), there is hardly any difference between the 2 in performance.p
On certain Apps the SB does perform really good like Photoshop. But for most stuff like Office, it performs the same ( ok, 0.9 second faster over a 11 second operation, but I find that insignificant )
So unless Apple will come with Quad cores, I'm not interested
When comparing similar clocked CPU's ( i7-975 vs the high end i7-2600k ), there is hardly any difference between the 2 in performance.p
On certain Apps the SB does perform really good like Photoshop. But for most stuff like Office, it performs the same ( ok, 0.9 second faster over a 11 second operation, but I find that insignificant )
So unless Apple will come with Quad cores, I'm not interested
Hmmm, so what is all the hype about? Anand seems to be super excited!
It has a GPU and offers high-end Nalehem performance for a lower price. It's also more energy efficient since it's 32nm. Maybe later this year new Sandy Bridge will come however that are more impressive.
But I think Anandtech forgot that you need to buy a new Motherboard for PC users who wish to upgrade to Sandy Bridge, so it's not exactly cheap.
edit: If you got an old computer and it needs to be replaced ( also motherboard ), then it's a pretty good upgrade for the money you pay for it.
Last edited:
The fact is, Sandy Bridges are faster but your usage must be able to take advantage of that extra processing power. It's up to your usage.
Average of course. Anand has been saying that since his early tests. SB brings you 999$ performance for <300$, that's why it's cool.
In the mobile market it's going to be a different story I think. The Sandy Bridge dual cores are clocked at 2.5 ghz up to 2.7 ghz while the Arrendales are clocked from 2.4 ghz to 2.8 ghz.
Since we see from Nalehem that clock vs clock, it can keep up with Sandy Bridge in most things ( such a Nalehem CPU costs alot, but not for Arrendales which are also on 32nm ).
So if you got an i7 2.8 Dual core MBP, I don't think the 2.7 ghz i7 2620M will offer a big jump of improvement at all.
A 2.3 ghz Core i7 2820QM Quad core however, is going to be a big jump in performance.
Interesting, so if the updated MBP with SB continues to use dual core chips, the average performance gain (similar clock speeds) should be minimal. On the other hand, if the dual core is replaced by a quad core, performance gains will be significant. Surely, switching to quad core seems like a no brainer unless...there is a material difference in power consumption. Right?
Look here (in German):
http://www.notebookjournal.de/praxis/sandy-bridge-intel-core-i7-2620m-i5-2540m-i5-2520m-benchmark-check-dual-core-cpus-prozessoren-nkrt-nkst-136/2
The Sandy Bridge Core i5-2520m (dual core) is faster than the high-end Arrandale Core i7-640m (dual core). Not by much, but still
http://www.notebookjournal.de/praxis/sandy-bridge-intel-core-i7-2620m-i5-2540m-i5-2520m-benchmark-check-dual-core-cpus-prozessoren-nkrt-nkst-136/2
The Sandy Bridge Core i5-2520m (dual core) is faster than the high-end Arrandale Core i7-640m (dual core). Not by much, but still
I was a bit confused by all the hype around SB as well. I purchased my first MBP 15" last year in August, it is the i7 2.66 High Res. I was considering upgrading this year thinking that I would need to to use the new Lion OS and keep up with current trends. But after some advice from forum members and after doing a bit of research it would not be worth while for me to upgrade. My current use is average. I use photoshop a little bit and pages, powerpoint, iphoto, downloading and uploading emails and images for my art work as well as an average amount of apps itunes youtube. So to sum it up I am an average user and to be honest the i7 was probably a bit extravagant for my needs but I had the money at the time and I beleive it will keep up with future upgrades and tech changes longer in to the future.
I personaly will not be upgrading this year maybe not the next either as what I have suits my need now and for the next few years.
Regards
Jenko
I personaly will not be upgrading this year maybe not the next either as what I have suits my need now and for the next few years.
Regards
Jenko
Half. Arrandale had a 45nm GPU on the same package. Sandy Bridge has the GPU on the same die, and it's all 32nm.
Sandy Bridge is much faster for something like, compressing a large video file.
I've seen speed tests that show the 2630QM to be MUCH faster compressing video than the 640M.
I've seen speed tests that show the 2630QM to be MUCH faster compressing video than the 640M.
That's quad vs dual. Clarksfield was faster than Arrandale in those tasks as well as it had more cores If you just compare two duals, the difference is much more negligible. QuickSync will make SB fly but I'm not sure is there any software for Mac that supports it.
Well you can compare a dual to a dual and the SandyBridge is still faster at these tasks because of QuickSync. It is specific tasks like video rendering I am talking about. The new Sandy Bridge i7 duals are faster than the current i7 offerings, we are talking about 15 - 20% faster at these tasks.
Sandy Bridge has a much much faster memory access and Cache than Arrendale. It gets a new turbo and by default higher clocks.
That is good for a much bigger speed up than you see from old to new quad core SB on desktops.
Here this is the best comparison that you can get.
http://www.anandtech.com/bench/Product/289?vs=143
That is basicly comparing an Arrendale and a Sandy Bridge Dual Core at very similar clocks without any Turbo.
The new Turbo only helps when opening or starting something as a very short boost thus the difference cannot really be put in numbers it is more of a subjective thing that you probably wouldn't be able to distinguish.
I just wish Anand would over relative values or svg graphics or some java script script so could switch to relative. You can calucalte the realtion by hand. If you want a average mean take into account how important each bench is for your real life usage.
With winrar you can see what fast memory access can do with apps that need it.
That is good for a much bigger speed up than you see from old to new quad core SB on desktops.
Here this is the best comparison that you can get.
http://www.anandtech.com/bench/Product/289?vs=143
That is basicly comparing an Arrendale and a Sandy Bridge Dual Core at very similar clocks without any Turbo.
The new Turbo only helps when opening or starting something as a very short boost thus the difference cannot really be put in numbers it is more of a subjective thing that you probably wouldn't be able to distinguish.
I just wish Anand would over relative values or svg graphics or some java script script so could switch to relative. You can calucalte the realtion by hand. If you want a average mean take into account how important each bench is for your real life usage.
With winrar you can see what fast memory access can do with apps that need it.
Are we talking like GT28RS or more T3/T4E? Ball-bearing perhaps? Will it be optimized for nitrous oxide?
Hold on, im confused, are you saying that the performance gains with sb for say comparable dual core arrandale and sb setups are substantial or not? Or does it depend the specific application?