X200 and SSD

Are there any major issues with SSDs and the X200/win7 64bit?

I was thinking of getting a SF1200 based drive.

Let me know what BIOS/SSD you are using if you have such a setup. Thanks.


Edit: See post #30 for my choice.

 

I'm using an SSD with my X200T. The drive is a fairly mundane Samsung PM800/PB22J, and the BIOS should be the latest. Never had any problems with the setup.

 

Yeah, you're going to get 50% performance in low power mode. There's a tweak for the registry over on StorageReview.com. It give an error in device manager and battery life takes a hit, but the performance goes back to normal.

 

ZaZ your comment seems very far out of context. I will look around at storage review later, but perhaps you could clue me more in.

"Yeah" seemed like you are replying to something.

 

He's responding to this part of your OP:

For reference, you can see these two threads:
http://forum.notebookreview.com/lenovo-ibm/510791-ssd-slow-down-low-power-mode.html
http://forum.notebookreview.com/har...upgrades/511684-ssd-slow-down-power-mode.html

 

My comment is on point. It's due to the Intel 4, 5 and series chipsets. In low power mode the chipset lowers performance to save battery life. I have not only seen this on my X200 and X200t, but other laptops as well. Like I said you can do the registry tweak and get back to normal performance, but you'll be left with a device manager error and less battery life if you care.

 

Is this the tweak you're talking about?

How To: Improve Low SSD Performance in Intel Series 5 Chipset Environments | StorageReview.com

 

Yes, sort of middle of the page.

 

So basically it might slow down 50% on random reading if you assume the storage review results when on battery power.

I don't understand the remarks about might as well have a mechanical disc though in some of the threads/discussion.

These numbers are still about 3-4x faster for seq long reads, 4-8x faster for 512K and ~50x faster for random 4K reads... Seems like almost an total non-issue other than the feeling of being cheated out of performance and throttling this not being ideal yet when trying to save power.

 

When SSDs have been hyped as the best upgrade to the silicon chip and you're laying out a good chunk of change to get one, I don't know that one can help feeling cheated. In low power mode, you're getting platter drive like performance and interestingly, no review I've ever seen mentions this. I have one in my desktop, which is plugged into the wall. It offers noticeably better performance. I like my SSD because it runs cool and is quiet. My machine doesn't shield much noise. I also run mostly in low power mode because my machine runs quiet that way. I may have still went for it because of those reasons, I just wish I knew what I was getting. It would be nice if there was a way to turn this on/off without having to hack the registry. I myself don't need much battery life. I mostly use the four cell and could live with the lower battery life.

 

I have no such problems. My drive feels as snappy on Max Battery Life as on Maximum Performance. That was both with OCZ Vertex 2 and Mushkin Callisto.

Maybe you should get a sandforce drive.

 

Interesting. I'll be plopping in an OCZ Vertex 2 the second my X201t gets here (on Monday). I'll post my experience.

 

The numbers don't support this statement. In the storage review link random reads 4K are 14 MB/s in low power mode, in contrast my standard X200 HD gets 0.2 MB/s, so it is 70 times faster... Even in the area that suffers the worst slow down. At least according to the synthetic benchmarks. Do you have other quantitative data to support your claim that it is no faster than a platter based drive in low power?

 

Do the same issues effect the X201?

 

I don't care what the numbers say. That's my real world experience on more than one ThinkPad and I've seen it on other machines too. I would expect the X201 to be the same, but I've not seen it for myself.

 

I upgraded my X200 with an OCZ Vertex 2 90GB a few weeks ago, and I'm very happy with the performance and battery life. During CrystalMark I get about 150mb/s (should technically be 270mb/s), but I get correct benchmarks on other software. I couldn't care less though, since the real life performance is waaay better than my previous 7200rpm disk - startup is FAST, applications open FAST, and my battery life is better.

 

Soooo, you "dont care" what a repeatable benchmark says, and we should instead just take your word for it with no other qualifications?

Howabout "no".

Until you can provide a logical explanation and follow it with repeatable benchmarks that OTHERS can perform on their own for verification, your story doesn't hold water. To include with my own IdeaPad Y460, with a Vertex 2 drive in it, with my own benchmarking on the lowest power state.

 

X201t:

Benchmarked my OCZ Vertex 2 with AS SSD Benchmark in a variety of power states/options. Biggest difference I was able to observe was a 20% drop in 4K write speed--not a big deal.

Haven't used it long enough to establish a qualitative feel for system speed in low power mode, but don't think the SSD is to blame for any sluggishness. The results even in low power absolutely kill the raid0 raptors on my desktop.

 

Well, if the laptop feels like you're on HD instead of SSD, it's pretty obvious even without any benchmarking. It would be a pretty huge difference. I would definitely notice that FAST.

As to why it happens for him, no clue. These are my benchmarks:

Power Savings profile:


Maximum performance profile:


This is Mushkin Callisto 60GB. From what I remember, I had similar numbers on OCZ Vertex 2 when it was in.

 

Did I not just say that?

I'm not going to repeat myself, but if you want to wade through my 15 page thread about my SSD, here it is. There's plenty of information and benchmarks in there. To sum it up, it is related to the Intel 4, 5 and 6 series chipsets, which reduce SSD performance in low power mode to conserve battery life. Not only have I seen this on ThinkPads, but on other machines as well. There is a BIOS tweak which largely fixes the issue, but gives an error message in device manager and battery life does take a hit.

 

So, finally my SSD arrived. What a joy! cannot describe it... it's so silent that I have an impression that it hanged Everything is so much faster.

In Linux I'm getting 245MB/sec seq read.. still looking for a better tool to benchmark it though.

Battery life got higher, at least with 30-50min more.

PS: It's a g.skill phoenix pro

 

@oct, I am planning on throwing in an SSD on my x200 too and noticed that you are using a Gskill.

Is this the model number? FM-25S2S-120GBP2
I'm about to order one.

 

Yes, that's it. Ordered from newegg during Black Friday sale...

The only thing I'd like to mention is that for some reason it didn't fit 100% into hdd caddy, it seems like ssd is a bit wider, I couldn't put 1 screw, but that's a minor thing since rubber rails are holding them together pretty well. And may be only my case...

 

I also ordered one of those GSKILL FM-25S2S-120GBP2 R from NeweggBuisness ($174 cyber Mon) for my X201.
I'm drooling in anticipation (^_^)!

I can never go back to HDD since my VAIO Z but if I get crappy results as some others mention, I can always reinstall the 72K HDD my X201 came with back in.

EDIT: I just noticed your response about the SSD not fitting perrfect. Did you try the hardware that comes with the SSD?

 

Nice! I've got mine for $186

I'm not sure there is anything there that I could use.

 

That's odd. Did you make sure the holes are all aligned with the caddy? Or tried different screws - the ones that came with the drive.

 

somebody who bought this SSD, wrote a newegg review and he did mention that he is having trouble fitting the screws.

I might hold ordering this one.
anyway, thanks for the info

 

I installed my SSD without the HDD caddy, just using the rubber guides. There have been several reported incidents of random freezes/crashes with people who install SSDs using the caddy on thinkpads (the issue seems to be that in some cases the SSD when in the caddy doesn't insert fully into the SATA slot).

If you run into issues, just remove the caddy an put the drive in using just the rubber guides. You might also want to make a pull out tab with some tape, so you can get your drive out if you need to.

 

Never thought to put it w/o a caddy... it's good to know, thanks!

I managed, with a bit of force, to screw all 4 of them... since I'm not planning to take it out any time soon. Will see how it goes.

Thanks!

 

I decided to acquire a X25-M 120GB which I obtained for 170 dollars after rebate. (pending).

Given the concerns posted here about battery performance (a mode a rarely use in any case), I thought it wise to test it both synthetically, subjectively and simply.

My stock 160 GB Hitachi 5400 RPM disc that came with my X200 was the comparison point. I made it the "worst case" by running the HD on AC power and the SSD on battery. The SSD crushes the HD. As it should of course, but claims like you might as well use a mechanical disc by ZaZ would seem to strongly indicate he has not used a mechanical disc in some time.

The sequential speed of the SSD is a factor of 3 faster than the very latest 7200 RPM notebook mechanical drives. Synthetically the drive, on battery performs almost 2 order of magnitude faster on random reading, which is the SSD's strong point vs mechanical drives. As I said still a factor of 2 to 3 faster for sequential.

These synthetic numbers are attached. They are nice to know but essentially pointless. A simple test of how much frustration the SSD will releave can be found in simple launch times of things.

To the welcome screen from a cold start:
X25-M 120 -- 29s; HD -- 87s; This agrees nicely with the synthetic tests.

Launching Photoshop CS4:
X25-M 120 -- 8s; HD -- 25s; This shows there is more random access involved here.

So yes, my HD was slow to begin with but only at most by a factor of 2 verse the very best mechanical drive. The SSD would still have won every test I have just mentioned by factor of about 2 or more.

Presuming TRIM keeps the drive performance up, I think I will be most pleased. I had considered a SandForce drive but I read too many comments about power management issues from bleeding edge people with who knows what firmware. I guess I have out live my youth, a small performance advantage is not worth all that extra time in one lump.

Next year we will see 25 nm flash, however, I caution all of you, in this case smaller is worse. Although you will see controller improvements to reduce writing to cells, the cells themselves will not last as long. This makes me think if you can find a good deal waiting is a mixed bag. Cell size is only decreasing by 20% at the 25 nm node and so they also have to add more levels per cell which again will decrease performance/endurance. Still there will be some good engineering to help workaround these issues. Nevertheless, the physics isn't looking great. Luckly, for me 120 GB is quite enough for all of my work and then some.

To sum up, yes on batteries it is slower. But then so is the CPU, that is the trade off to power manage devices. You should expect that. If the CPU took no hit on battery it would always be throttled, the reason there are different power states is there is an overhead to throttling. I am glad they made some sort of engineering compromise and looked to trade some performance for power, because when I am on batteries I care mostly about how long my computer will run doing low performance things like web, power point, etc.

So my initial conclusion is very positive. I will post again if I run into issues, power or otherwise.

 

This is why I didn't buy the GSKILLs it was a common complaint in the reviews that the case tolerances were horrible. Almost went for a Mushkin SandForce 120, but the Intel was a better deal. Best drive to have would really be Crucial C300 for a Win 7 laptop or a well made SandForce drive... But the cost premium on the C300 was just way to much (90 dollars! -> 50% more)