macbook pro 2011. The best value for money upgrade?

[temen]

Hi,

I'd like to summarize my research.

Objective: Better performance. Options to upgrade with the best value for money at this (or +2 months) time.
(So, 16 gb of ram for $1600 is NOT an option for 1 or 2% increase of speed.)

If somebody _needs_ for example 16gb of ram, Im sure they have money to spent without blinking an eye, and also they know they need it. So leave this scenario.

Also leave question "If I need it", please. Let's just focus what is the best for getting better performance and spending $2000 for upgrades.


1. Ram
What: 2x4GB DDR3 1333MHz CL7 ~$120
Example: http://www.avadirect.com/product_details_parts.asp?PRID=20108

1. Hard drive?
What: 128-256gb ssd 6gb/s ~ $320 (for 128gb)
Example: http://eshop.macsales.com/item/OWC/SSDMX6G120/
Why: Yes, it's a bit pricey, but if you're looking for performance there's no reason to buy not so much cheaper ssd 3gb/s which is two times slower than 6gb. If you don't have $320 - wait till price goes down.




(At the beggining, when I opened "new post" link I just wanted to just ask for your opinions, but decided to tell what I know/think I know, then ask for correction if Im wrong.. )

So. Am I right?

 

[simsaladimbamba]

RAM is good for running several memory hungry applications concurrently or running one very memory hungry application. With After Effects (and many other applications) I can never have enough RAM.
SSDs are good for faster access times, then even less RAM might not be so bad, if one needs more RAM.
For many uses 4GB RAM is enough.
Am I wrong?

 

[Ccrew]

Best performance will be in an upgrade from hard drive to SSD, that's the hands down winner.

 

[Hellhammer]

RAM - 83$

120GB for 320$ sounds quite a lot. 120/128GB SATA 3Gb/s drives can be had for less than 200$. I wouldn't be able to justify spending +50% more $ on a faster SSD, especially when I know that there is no difference between the two in real world.

 

[temen]

RAM is good for running several memory hungry applications concurrently or running one very memory hungry application. With After Effects (and many other applications) I can never have enough RAM.
SSDs are good for faster access times, then even less RAM might not be so bad, if one needs more RAM.
For many uses 4GB RAM is enough.
Am I wrong?

Yeah, but I thought I asked for leaving questions if upgrade is needed..?

Best performance will be in an upgrade from hard drive to SSD, that's the hands down winner.

Oh, sorry, then I wasn't clear enough.
It's not a question "ram OR hdd". Is more question "which ram" AND "which hdd"

 

[simsaladimbamba]

Yeah, but I thought I asked for leaving questions if upgrade is needed..?


Oh, sorry, then I wasn't clear enough.
It's not a question "ram OR hdd". Is more question "which ram" AND "which hdd"

Then I didn't understand your OP correctly. I will leave the questions out of it then.

 

[cluthz]

Hi,

1. Hard drive?
What: 128-256gb ssd 6gb/s ~ $320 (for 128gb)
Example: http://eshop.macsales.com/item/OWC/SSDMX6G120/
Why: Yes, it's a bit pricey, but if you're looking for performance there's no reason to buy not so much cheaper ssd 3gb/s which is two times slower than 6gb. If you don't have $320 - wait till price goes down.


...

6gb/s SSDs aren't twice as fast as a 3gb/s SSD. The real world difference is more like 10%. In most tasks you won't notice difference between any SSD at all.

In syntetic benchmarks SSDs are often 100x faster than normal HDs, but once an application is launched there often aren't much difference at all.

 

[Fugue]

RAM - 83$

120GB for 320$ sounds quite a lot. 120/128GB SATA 3Gb/s drives can be had for less than 200$. I wouldn't be able to justify spending +50% more $ on a faster SSD, especially when I know that there is no difference between the two in real world.

The RAM you linked - does it work flawlessly with the 2011 13" MBP? What's the difference between a more expensive RAM, and this one? What about installation? Can I simply turn off the machine, remove the old RAM, install the new RAM, and start it up again and have it functioning perfectly?

Thanks

 

[temen]

6gb/s SSDs aren't twice as fast as a 3gb/s SSD. The real world difference is more like 10%. In most tasks you won't notice difference between any SSD at all.

In syntetic benchmarks SSDs are often 100x faster than normal HDs, but once an application is launched there often aren't much difference at all.

Well, Im working on macbook (software engineer), and I have a lot of i/o operations, so I think I will notice the difference.. But "10%"? Benchmarks are showing that 6bg are really ~500mb/s fast. Could you elaborate?

The RAM you linked - does it work flawlessly with the 2011 13" MBP? What's the difference between a more expensive RAM, and this one? What about installation? Can I simply turn off the machine, remove the old RAM, install the new RAM, and start it up again and have it functioning perfectly?

About this specific model - I don't know. It was more like an example for which hardware parameters are good (cl7, 1333MHz). Yes, you can simply turn off, remove old ram, install new, and start.

 

[Hellhammer]

The RAM you linked - does it work flawlessly with the 2011 13" MBP? What's the difference between a more expensive RAM, and this one? What about installation? Can I simply turn off the machine, remove the old RAM, install the new RAM, and start it up again and have it functioning perfectly?

Thanks

It works fine. I've seen plenty of 2011 MBP owners buying that specific RAM. There is no noticeably difference between that and more expensive RAM, other than the price tag. Personally, I always go with the cheapest name brand RAM (pretty much whatever NewEgg or other reliable reseller sells). You just swap the RAM and reboot, that's all.

 

[JasonH42]

1600MHz RAM rather than 1333MHz gave me (and others) a 10% performance boost if you believe benchmarking tools.

The same benchmarking tools told me my C300 and Vertex 3 SSDs don't benefit much at all from the 6GBps SATA3 interface over 3GBps SATA2. SSDs are amazing though, more speed, far less heat - so no fan noise generally, no vibration, best upgrade I ever bought.

 

[temen]

1600MHz RAM rather than 1333MHz gave me (and others) a 10% performance boost if you believe benchmarking tools.

The same benchmarking tools told me my C300 and Vertex 3 SSDs don't benefit much at all from the 6GBps SATA3 interface over 3GBps SATA2. SSDs are amazing though, more speed, far less heat - so no fan noise generally, no vibration, best upgrade I ever bought.

Good to know 1600MHz runs on macbooks, thanks for that.

About SSD: from what I see on some tests C300 have ~150mb/s (for example: http://www.anandtech.com/show/3812/the-ssd-diaries-crucials-realssd-c300/2 )
And 6gbps: http://www.barefeats.com/ssd6g.html

 

[Hellhammer]

1600MHz RAM rather than 1333MHz gave me (and others) a 10% performance boost if you believe benchmarking tools.

In real world, there is very little to no difference in performance. RAM bandwidth is not the bottleneck, thus faster RAM won't help. IMO 1600MHz RAM is waste of money.

 

[adrian1480]

In real world, there is very little to no difference in performance. RAM bandwidth is not the bottleneck, thus faster RAM won't help. IMO 1600MHz RAM is waste of money.

pretty much this.

 

[JasonH42]

In real world, there is very little to no difference in performance. RAM bandwidth is not the bottleneck, thus faster RAM won't help. IMO 1600MHz RAM is waste of money.

Possibly. But Xbench says otherwise and there's little if any price differential so why not?

SSD benchmarks are so variable and speed claims are based on different sorts of data. My experience with the C300 was fantastic but it wasn't much better on SATA3 and nor is the Vertex 3 despite the grand claims. In fact, I bought a second Vertex 3 to put in the optical bay in order to try them in RAID0 configuration. The optical bay is only SATA2 but I reckon as a RAID0 pair they'll perform much better than a single Vertex 3 on the SATA3 port.

 

[adrian1480]

if you need more ram you have to get the ram. it's all about how much RAM you use for your work.

the SSD will give the best increase, but you can still be bottlenecked in your work if you don't have enough RAM. Caching--even to an SSD--sucks and slows everything down.

 

[Hellhammer]

Possibly. But Xbench says otherwise and there's little if any price differential so why not?

SSD benchmarks are so variable and speed claims are based on different sorts of data. My experience with the C300 was fantastic but it wasn't much better on SATA3 and nor is the Vertex 3 despite the grand claims. In fact, I bought a second Vertex 3 to put in the optical bay in order to try them in RAID0 configuration. The optical bay is only SATA2 but I reckon as a RAID0 pair they'll perform much better than a single Vertex 3 on the SATA3 port.

Xbench is synthetic, its results are NOT comparable to real world. Sure, 1600MHz DDR3 has ~20% higher maximum memory bandwidth but again, that does not mean that it's 20% faster in every task. Like I said above, memory bandwidth is very rarely the bottleneck, thus it doesn't matter whether you have 1600MHz or 1333MHz because they will perform the same in real world.

The same thing applies to SSDs. The only way you are going to notice any difference is when you launch Xbench and look at its synthetic numbers. To put this straight, you won't notice any difference between single Vertex 3 and two of them in RAID 0, especially because random speeds do not scale up like sustained speeds do. In most of the every day tasks, the usage consists of random reads and writes, not sustained.

 

[DarkFlame]

Hi,

I'd like to summarize my research.

Objective: Better performance. Options to upgrade with the best value for money at this (or +2 months) time.
(So, 16 gb of ram for $1600 is NOT an option for 1 or 2% increase of speed.)

If somebody _needs_ for example 16gb of ram, Im sure they have money to spent without blinking an eye, and also they know they need it. So leave this scenario.

Also leave question "If I need it", please. Let's just focus what is the best for getting better performance and spending $2000 for upgrades.


1. Ram
What: 2x4GB DDR3 1333MHz CL7 ~$120
Example: http://www.avadirect.com/product_details_parts.asp?PRID=20108

1. Hard drive?
What: 128-256gb ssd 6gb/s ~ $320 (for 128gb)
Example: http://eshop.macsales.com/item/OWC/SSDMX6G120/
Why: Yes, it's a bit pricey, but if you're looking for performance there's no reason to buy not so much cheaper ssd 3gb/s which is two times slower than 6gb. If you don't have $320 - wait till price goes down.




(At the beggining, when I opened "new post" link I just wanted to just ask for your opinions, but decided to tell what I know/think I know, then ask for correction if Im wrong.. )

So. Am I right?

HERE WE GO - this is my thread. OP I know EXACTLY what you mean. I only but computers every 4 years or so, I want the BEST value for money. I challenge ANYONE to beat this:

MARCH PURCHASE:
$999 (1049.00 w/ tax) Microcenter 2011 13" MBP
120GB Vertex 3 (received March 28th) $275 Shipped
8GB DDR3 1333 Ram for $65 (G.Link - works perfectly).
50 dollars for optibay caddy (could've paid 25 for an asian one off ebay but that quality is questionable). Moved my 320gb OEM over to the optibay slot!
1TB External Seagate GoFlex Ultraportable for $67 shipped - Seagate also announced that they are currently making a 20 dollar thunderbolt adapter!

Grand total INCLUDING tax: $1456.00

This is by far the best computer I have ever owned. It has changed my life, I don't think there has been anything as significant as SSDs in the computer industry as 4X CDRs 10 years ago. I remember the day I burned my first mix cd!

 

[scottlinux]

newegg has a ton of SSDs:

http://www.newegg.com/Store/SubCategory.aspx?SubCategory=636&name=SSD

 

[JasonH42]

Xbench is synthetic, its results are NOT comparable to real world. Sure, 1600MHz DDR3 has ~20% higher maximum memory bandwidth but again, that does not mean that it's 20% faster in every task. Like I said above, memory bandwidth is very rarely the bottleneck, thus it doesn't matter whether you have 1600MHz or 1333MHz because they will perform the same in real world.

The same thing applies to SSDs. The only way you are going to notice any difference is when you launch Xbench and look at its synthetic numbers. To put this straight, you won't notice any difference between single Vertex 3 and two of them in RAID 0, especially because random speeds do not scale up like sustained speeds do. In most of the every day tasks, the usage consists of random reads and writes, not sustained.

But you concede the 1600 has higher bandwidth; so if the price differential is next to nil then why wouldn't you go for it? It would be daft to buy the 1333.

 

[Hellhammer]

But you concede the 1600 has higher bandwidth; so if the price differential is next to nil then why wouldn't you go for it? It would be daft to buy the 1333.

If the price difference is close to inexistent, then of course it's better to get it.

 

[leman]

Best performance will be in an upgrade from hard drive to SSD, that's the hands down winner.

Well, this is simply not true. SSD result in great "subjective" performance, that is, the machine will be more responsible and a pleasure to work with (after i got my SSD I never want to part with it). But in reality, the SSD won't improve the performance of most programs you actually want to improve. It won't speed up filter processing in Photoshop, it won't speedup compilation of big projects, it won't speedup LaTeX typesetting and it won't speedup statistical simulations. All these things are still CPU(and RAM) bound. An SSD is more of a convenience thing (except if you actively use one of few programs out there which are actually HDD bound, like - I don't know - web servers )

 

[Ccrew]

Well, this is simply not true. SSD result in great "subjective" performance, that is, the machine will be more responsible and a pleasure to work with (after i got my SSD I never want to part with it). But in reality, the SSD won't improve the performance of most programs you actually want to improve. It won't speed up filter processing in Photoshop, it won't speedup compilation of big projects, it won't speedup LaTeX typesetting and it won't speedup statistical simulations. All these things are still CPU(and RAM) bound. An SSD is more of a convenience thing (except if you actively use one of few programs out there which are actually HDD bound, like - I don't know - web servers )

Agreed - to an extent. But when you take into consideration the time to launch any of the applications and the I/O associated with any reads/writes from them then you're still going to get a huge improvement from the SSD versus memory. Just to use P'shop as an example, while you're correct in that the filter operations themselves aren't going to be enhanced, the launch of the program and any sub-functions are, as is performance to scratch and anything that requires drive access. In that respect it is overall the better performance bang for the buck. I'm not sure you're figuring how much an app actually writes to the disc...

 

[leman]

Agreed - to an extent. But when you take into consideration the time to launch any of the applications and the I/O associated with any reads/writes from them then you're still going to get a huge improvement from the SSD versus memory. Just to use P'shop as an example, while you're correct in that the filter operations themselves aren't going to be enhanced, the launch of the program and any sub-functions are, as is performance to scratch and anything that requires drive access. In that respect it is overall the better performance bang for the buck. I'm not sure you're figuring how much an app actually writes to the disc...

Ok, you got me there - I actually don't use photoshop and have only limited knowledge about this software You are probably right that SSD is a great update for such software. In my case — I usually do statistical simulations/data mining/LaTeX — SSD does not help a bit as all these tasks are CPU bound (I actually timed it). It does improve the worst case performance though