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.
3.33 6core vs 2.66 12 core
- Thread starter allupons
- Start date
- Sort by reaction score
When a person presents something as if to sound "like" a fact, you should be prepared to seriously back up your claims. While it may seem simply intuitive to make the assumption you did above, your statement smells a lot like someone who thinks they know stuff but in fact does not.
Then please do. As of yet you've only spouted typical SSD myth ignorance.
Come back when you've actually tried it.
"spouted"? no, I referred directly to the diglloyd testing being done.
"tried it", how about "try" making a more inane comment? No kidding initial use appears to be better, that's not what is being discussed and is therefor irrelevant. It being super fantastic in the first phase of use is completely irrelevant to the discussion.
You come in here sounding like a know it all with zero sensibility in your approach - and you have been using one for how long and how intensively?
I do not doubt the actual tests that have read about. If you want to be useful then supply the proof as well as what I had asked above rather than what sounds like a basic "just trust me, in know" response.
-
and did you even read the thread?
"tried it", how about "try" making a more inane comment? No kidding initial use appears to be better, that's not what is being discussed and is therefor irrelevant. It being super fantastic in the first phase of use is completely irrelevant to the discussion.
You come in here sounding like a know it all with zero sensibility in your approach - and you have been using one for how long and how intensively?
I do not doubt the actual tests that have read about. If you want to be useful then supply the proof as well as what I had asked above rather than what sounds like a basic "just trust me, in know" response.
-
and did you even read the thread?
Ah, such as your worthless ignorance driven drivel?
ALL use. The only disadvantage an SSD has is in $/gb.
3 years in daily use. I started with a 25GB Samsung MCBQE25G5MPQ and I still use it as my primary drive to this day without the need for a wipe.
Due to it being an older SSD, my much newer Barracuda 7200.12 outspeeds it in sequential writes. But in random reads and writes, the SSD still rapes it in every category.
Drive Type SAMSUNG MCBQE25G5MPQ-0VAD3
Disk Test 120.87
Sequential 117.90
Uncached Write 92.21 56.61 MB/sec [4K blocks]
Uncached Write 113.97 64.48 MB/sec [256K blocks]
Uncached Read 104.67 30.63 MB/sec [4K blocks]
Uncached Read 210.41 105.75 MB/sec [256K blocks]
Random 123.98
Uncached Write 47.85 5.07 MB/sec [4K blocks]
Uncached Write 109.14 34.94 MB/sec [256K blocks]
Uncached Read 2572.93 18.23 MB/sec [4K blocks]
Uncached Read 551.11 102.26 MB/sec [256K blocks]
Drive Type ST31000528AS
Disk Test 60.51
Sequential 179.58
Uncached Write 188.55 115.76 MB/sec [4K blocks]
Uncached Write 205.00 115.99 MB/sec [256K blocks]
Uncached Read 123.29 36.08 MB/sec [4K blocks]
Uncached Read 251.15 126.23 MB/sec [256K blocks]
Random 36.39
Uncached Write 10.78 1.14 MB/sec [4K blocks]
Uncached Write 266.93 85.45 MB/sec [256K blocks]
Uncached Read 119.09 0.84 MB/sec [4K blocks]
Uncached Read 198.07 36.75 MB/sec [256K blocks]
A modern SSD would do far better in sequential testing and wipe the hard drive off the map.
If you had actually read this thread, you'd have seen I've been a part of this thread 65 posts before you even came along.
One other off hand comment apparently without serious regard for the situation or even the various ways 6 vs 12 can be compared, hardly constitutes "part of this thread".
However, 3 years of ssd use sounds like a great test of time ....Why do you have to be pushed so ***** hard to speak up?
So if you have been using these for so long, that's great, but it still says nothing about how intensively.
I'm sorry if it bothers you to be questioned but my business depends partially on the work that has to go through my computer so I am not just going to take the word of anyone who just claims this or that.
I don't get paid to know what random reads/writes are vs sequential, and I couldn't care less either. I am not a computer geek and do not have time to be but I want to learn about the components that are important for my work.
This is specifically reliability first and performance second. Compared to most people, I probably don't care so much about the cost, to me, quality warrants the cost because it supports my work.
Let me ask then, if you are so certain about the ssd, why are established folks going on and on about the ssd degradation issues? Furthermore, why can't anyone answer my simple question, "who cares if they get slower over time, can't you just reformat / swipe them and start fresh and fast again?" -every month if necessary?
Is there any way the ssd reliability is less than the best hdd? in terms of file integrity but also failure rates?
As for performance, I thought I read that tests of 4 fast drives would easily yield better results do to the limits of the ssd - currently. How about an array of 10k rpm drives either in sata or sas type? Doesn't the controller for them have a lot to do with ultimate speed limitations?
I should stay away
the 4 disc tests ? head over to digilloyd mac performance guide and look up his scratch disc stuff
s
Because trolls like yourself lack the frontal lobe capacity to understand the point the first time.
Wow, you're pushing for "proof" yet you don't even understand the subject you're talking about.
An SSD is supreme at this. It has no moving parts, makes no heat (relatively) and isn't affected by movement/vibration or magnetism.
They don't know what they're talking about. The TRIM feature, once supported by OSX, will eliminate the small write penalty.
It doesn't work that way. Nobody answered because a quick goggle search will answer for you.
Being more compact and tightly integrated, hitting an SSD with a hammer will destroy it easier. A hard drive could have its platters installed into another drive for data retrieval.
It doesn't matter if you have 100 drives in RAID, they all have mechanical heads that must physically move to retrieve data. That means the RAID will be faster in raw throughput but an SSD will still wipe the floor with them in random read response.
Thats how an SSD makes any machine feel faster, they have a 0.01s seek time vs 0.1s.
I should point out hitting a spinning HDD with a hammer will destroy it as well. I'm talking shattering into magnetic dust! I'm speaking from experience here.
For the record a troll is not someone trying to learn more.
Secondly you continue to spout your opinions as if they are the gospel, thus my reason for questioning your accuracy. Other folks here with more credibility than yourself will post references to claims, etc. You just expect people to think you know it all because you say you do. Which makes you sound even more like a regular dork.
None of this stuff is even slightly hard to understand, the issue is more that you are clearly just a total jerk.
not really wanting in on this thread...but diglloyd's disktester has a "Recondition SSD", command. He does state that it isn't needed for drives such as OWC as they don't show performance degradation overtime. (This may or may not be true but probably is true for most use cases.)
http://diglloydtools.com/manual/disktester-recondition.html
I think this is a different issue to simply wearing the SSD out, which could happen if you are constantly writing a high percentage of the SSD's capacity.
I wouldn't use diglloyd's recondition tool... I'm not sure what exactly it's doing but he says "The recondition command helps restore solid state drive performance by rewriting free space on the drive in very large chunks."
This is, in fact, the opposite of what you want to do to recondition an SSD. When all the nand blocks on a drive get full, which is what this method assures, drive performance starts to deteriorate because of the write amplification... writing just 8 bytes (for example) will require the reading and writing of a full 512K byte block of NAND.
Here's the correct process using the ATA command that all SSD drive controllers recognize:
https://ata.wiki.kernel.org/index.php/ATA_Secure_Erase
Unfortunately, this is not trivial to perform on a Mac so myself and others have found a few ways to do this successfully:
https://forums.macrumors.com/threads/841182/
The key to the process is the issuing of the ATA command that instructs the drives controller to mark all NAND blocks as free. Which means that writing a small amount of data will no longer require the reading and writing of a full 512K block... thus performance returns to factory values.
But don't take my word for it... here are some reputable sources:
Intel:
AnandTech:
Mac Pro 3.33 GHZ 6 Core vs. Mac Pro 2.66 GHZ 12 Core
I have enjoyed reading everyone's opinions regarding the purchase of a 6-Core 3.33 GHZ Mac Pro vs. a 12-Core 2.66 GHZ Mac Pro. From what I gather it seems like the consensus of the posts leans towards the 6-Core 3.33 GHZ.
I will mostly be using this machine for Final Cut Express and Photoshop CS5. I don't waiting for that final render, but appreciate when a machine is super fast during the editing process.
I don't mind spending up to $5,000 on a new machine, but I would rather put the money into more RAM, solid state HD or other components and not unnecessary cores. I realize this thread is a few months old, but if anyone has any last minute advise for a fellow Mac enthusiast I would appreciate it!
Thank you!
Seth
I have enjoyed reading everyone's opinions regarding the purchase of a 6-Core 3.33 GHZ Mac Pro vs. a 12-Core 2.66 GHZ Mac Pro. From what I gather it seems like the consensus of the posts leans towards the 6-Core 3.33 GHZ.
I will mostly be using this machine for Final Cut Express and Photoshop CS5. I don't waiting for that final render, but appreciate when a machine is super fast during the editing process.
I don't mind spending up to $5,000 on a new machine, but I would rather put the money into more RAM, solid state HD or other components and not unnecessary cores. I realize this thread is a few months old, but if anyone has any last minute advise for a fellow Mac enthusiast I would appreciate it!
Thank you!
Seth
I just went through that ... got my 6-core just over a week ago and its awesome. 24GB memory, and my two 2TB Caviar Green will arrive tomorrow. And that external 2TB drive for backup.