Show us your SSD performance

[ignatzatsonic]

Am I fighting the wrong war here.

Absolutely not.

I know that an 80 GB Intel SSD shows as about 93% of that: 74.5.

We know from Anandtech that a 300 GB Intel SSD shows as about 93% of that: 279 GB.

I think Hoppy said his 120 GB Intel showed as 111 plus in Windows; about 93%.

What about other Intel SSD sizes?

Do all other Intel SSD sizes happen to show as 93% of stated capacity in Windows?

Is it merely a coincidence that Intel allots 7% of advertised space for over-partitioning?

I have no idea.

 

[essenbe]

As I understand it, all SSD manufacturers over provision their SSDs. That is why wear leveling and Replacing bad nand cells comes from. There are nand cells on the SSD that are not accessible to the user.

 

[Hopalong X]

Exactly Essenbe.

Intel must have figured about 7% is the sweet spot for their drives.

Other brands do vary slightly.

 

[whs]

As I understand it, all SSD manufacturers over provision their SSDs. That is why wear leveling and Replacing bad nand cells comes from. There are nand cells on the SSD that are not accessible to the user.

Steve, that is correct - but that is another matter. The other discussion we had was a matter of pure mathematics (Decimal vs Binary)

 

[Hopalong X]

Oh well.

Good conversation while it lasted.

 

[ignatzatsonic]

111.79 x 1.024 x 1.024 x 1.024 = 120.033


111.79/120.033 = .9313272

1 - .9313272 = .0686728

.0686728 x 100 = 6.86728%

A coincidence?

 

[A Guy]

Read the 1st 2 paragraphs in this pdf file from Intel

http://cache-www.intel.com/cd/00/00/45/95/459555_459555.pdf

The performance benefits of over-provisioning are most
relevant to the Intel® X25-M SSD because the SSD already has
a larger spare area set by design. For this MLC device, which is
architected with 7% spare area, increasing the spare area can
boost performance by 10% or more depending on the amount
of spare area assigned

A Guy

 

[whs]

Read the 1st 2 paragraphs in this pdf file from Intel

http://cache-www.intel.com/cd/00/00/45/95/459555_459555.pdf

The performance benefits of over-provisioning are most
relevant to the Intel® X25-M SSD because the SSD already has
a larger spare area set by design. For this MLC device, which is
architected with 7% spare area, increasing the spare area can
boost performance by 10% or more depending on the amount
of spare area assigned

A Guy

Bill, I am a bit lost. What is this quote telling me - other than that they set sufficient spare area aside.

Does that mean that there is no need to leave extra unallocated space on the X25?

 

[Snakeyeskm]

whs I have a few Intel drives and a little extra OP(10% plus) has always helped. The toolbox helps in establishing the trim command, but the OP is really where the SSD executes the sub-processes that clear the blocks identified by the trim command. The extra OP can always be recovered by expanding the partition but there is a catch 22. Unfortunately SSD drives that out over 80% full tend to operate in the "mapped" steady-state and under heavy usage (read/write/delete) will frequently enter the "hammered" state.

 

[A Guy]

Read the 1st 2 paragraphs in this pdf file from Intel

http://cache-www.intel.com/cd/00/00/45/95/459555_459555.pdf

The performance benefits of over-provisioning are most
relevant to the Intel® X25-M SSD because the SSD already has
a larger spare area set by design. For this MLC device, which is
architected with 7% spare area, increasing the spare area can
boost performance by 10% or more depending on the amount
of spare area assigned

A Guy

Bill, I am a bit lost. What is this quote telling me - other than that they set sufficient spare area aside.

Does that mean that there is no need to leave extra unallocated space on the X25?

It confirms the 7% number for OP that was mentioned, and people were questioning if that was binary to decimal ad nauseum

A Guy

 

[Britton30]

Total confusion reigns - LOL

Divide 100 billion (100GB in decimal) three times by 1024 and see what you get.

The first one to figure it out wins the Speedy Gonzales prize.

93.13225746154785...maybe

 

[whs]

It confirms the 7% number for OP that was mentioned, and people were questioning if that was binary to decimal ad nauseum

I don't think so. These 7% are purely coincidental. The 7% we were discussing (for the GB range) are a purely mathimatical phenomenon. If it were in the MB range, we would be seeing appr. 5% and in the TB range appr. 10%. But I think we have discusses that to death. So no point to start all over again.

 

[whs]

Here is some SSD wisdom from Ed Bott who is one of the industrie#s Gurus: Windows 7 and SSDs: Setup secrets and tune-up tweaks | ZDNet

 

[linnemeyerhere]

Get out the Amex Black and or Master the possibilities or do like I'll do and drool and wait for them to fall with time !

 

[fseal]

Frankly... that's more resonable than the 48k Apple ][ expansion that cost $2250 when it came out

 

[whs]

First day with my new M4. Nums look a bit strange but what really counts are the 4K nums. The thing is lightning fast with a 7.8 WEI.

 

[A Guy]

First day with my new M4. Nums look a bit strange but what really counts are the 4K nums. The thing is lightning fast with a 7.8 WEI.

A Guy

 

[essenbe]

WHS, those numbers look prety good. Crucials have lower write speeds. Are you in AHCI mode?

 

[whs]

WHS, those numbers look prety good. Crucials have lower write speeds. Are you in AHCI mode?

Yes, AHCI and it is on a Sata3 port.. What is interesting is that the nums and the performance experience do not always match. I have one OCZ Vertex 2 in a laptop which has a WEI of 6.9 (don't ask me why), but this is my fastest system.

 

[essenbe]

well, I think that says an awful lot about the WEI system, don't you?