sblantipodi

My two cents to the site reviewers...

10 posts in this topic

Hi,

I use this site to choose what drive buy next...

Unfortunantly I never see a real comparison betweeb drives of two or three or four previous generation.

Never seen a reviews that let readers to easily understand if it's the time to upgrade.

Imho a great reviews with modern drives vs previous one is the must for people who are seeking for an upgrade.

Is it the time to upgrade?

1 person likes this

Share this post


Link to post
Share on other sites

It can be difficult: For example:

1)Each situation is different.

Gen 4 HDD may be faster than a Gen 1 HDD, but a Gen 2 SSD might be a better if someone wants speed and does not store much. And a Gen 3 HDD will be cheaper than a Gen 4 HDD and will likely be a good upgrade for someone who has a Gen 1 and doesn't want to spend a lot. So each review will have include just a half of page trying to explain that which may get convoluted(espically with different types of users/computers).

2)Tech moves fast.

Comparing past two generation may not be necessary and may be wasted effort. For example, Seagate is on their 4th(maybe 5th generation) of SSHD(hybrid drivers), while Toshiba is on their 2nd generation. A comparison of the Seagates latest gen drive(2013) to their first(made 2010 or 2007 depending on where you define it) is not as important as comparing it to Toshiba's second generation to see how it stacks up against competition.

However, it is still a good observation and it is one of the reason for the 'What Drive Should I Buy?' forum and another reminder that a dynamic chart of benchmark is needed(so you can easily compare new drives with old gens, competitors, etc side by side).

Also, a plug for Storagereview's reference guide: http://www.storagereview.com/storage_reference_guide

It contains guides and how-to articles.

Share this post


Link to post
Share on other sites

We compare against prior generation in almost every case. For hard drives the changes are minimal, but the scale when comparing SSDs for instance gets pretty wacky pretty fast. And since most old drives aren't in the market any more and our testing platforms continue to evolve, it's difficult to make apples to apples for much older drives. You can still reference the prior reviews though to at least get a ballpark comparison.

Share this post


Link to post
Share on other sites

We compare against prior generation in almost every case. For hard drives the changes are minimal, but the scale when comparing SSDs for instance gets pretty wacky pretty fast. And since most old drives aren't in the market any more and our testing platforms continue to evolve, it's difficult to make apples to apples for much older drives. You can still reference the prior reviews though to at least get a ballpark comparison.

I don't understand why you say this for SSD.

I bought a Force GT 240MB 18 months ago, 18 months for technology is a lot, can you say that if you add a Force GT the "scale gets pretty wacky"?

I don't think so.

From the reviews it seems that there is absolutely no news on SSD.

500MB/sec two years ago, 500MB now.

85K IOPS two years ago, also lesser IOPS on modern drives.

Where do you see that the scale gets pretty wacky?

Share this post


Link to post
Share on other sites

If you want to go back 4 generations to original X25's it starts to become an issue. At that point you're also comparing different SATA interfaces on different testing platforms too. After 4 years many SSDs that we get in for testing are also dead or near dead after being abused aggressively in our lab. So it's not as simple as re-benching on a new platform. I was simply suggesting that if you want to go back 4 years on reviews that it's going to be easier to pull up the benchmarks from those reviews and compare them to modern reviews. Too much changes in our testing too quickly to go back more than a single generation for SSD.

Share this post


Link to post
Share on other sites

If you want to go back 4 generations to original X25's it starts to become an issue. At that point you're also comparing different SATA interfaces on different testing platforms too. After 4 years many SSDs that we get in for testing are also dead or near dead after being abused aggressively in our lab. So it's not as simple as re-benching on a new platform. I was simply suggesting that if you want to go back 4 years on reviews that it's going to be easier to pull up the benchmarks from those reviews and compare them to modern reviews. Too much changes in our testing too quickly to go back more than a single generation for SSD.

4 years is too much, but 2 years at least.

there is no differencee in the drives of two years ago with the modern drives.

My Force GT 240GB (two years old drive) has the same interface SATA 6GB/sec of the new Crucial M500 and it is also faster.

What I would like to say is that there should be an evidence that newer drives does often not add any performance improvements over previous one

or too few to worth an upgrade.

Can't understand why you say that SSD is improving faster than HDD, from my Force GT I haven't seen a drive that it worth an upgrade.

There plenty of new drives slower than mine with reduced lifetime also.

This things should be in evidence, there are many people who needs to buy new drives, but there are many other people that simply need to upgrade,

this sites lack an "upgrade section" :D

I love your site tought and I will always make adverisement of this site in external forums :D

Share this post


Link to post
Share on other sites

I don't understand why you say this for SSD.

I bought a Force GT 240MB 18 months ago, 18 months for technology is a lot, can you say that if you add a Force GT the "scale gets pretty wacky"?

I don't think so.

From the reviews it seems that there is absolutely no news on SSD.

500MB/sec two years ago, 500MB now.

85K IOPS two years ago, also lesser IOPS on modern drives.

Where do you see that the scale gets pretty wacky?

Brian is referring to the generational changes between SATA I, SATA II, and SATA III SSDs. The Sandforce SF-2200 controller in the Force GT is still a current generation drive, although one one of the first to be released in this cycle that has spanned a couple of years now.

We like to keep our charts easy to consume and cause as little confusion as possible. With the rapid spread of Sandforce controllers to the market, this means we will generally keep only one device that uses that controller on our charts (such as the Intel SSD 520) since so many of them will perform near identically with one another. There is little point to showing the performance of 50 different brand/model SSDs if they all use the same controller, same firmware, and minor differences in NAND. The performance of SSDs in the past two years in the consumer market hasn't changed a significant amount since they first hit the market since there has been no radical change to the interface like there was from SATA II to SATA III.

Right now even the latest generation consumer and enterprise platforms still have SATA or SAS 6.0Gb/s as the bottleneck. SAS 12Gb/s is starting to filter out, with one HBA family, one shipping SSD, and a few working their ways through development status. Most performance you see right now is pretty close to what you will see from the interface being the limiting factor on the consumer side, with some interesting stuff on the enterprise side (with a huge cost associated with it).

Share this post


Link to post
Share on other sites

The other point that gets lost when comparing SSDs is that the relative performance between the slowest and the fastest SSD, is usually quite insignificant in actual system performance.

In a blind test most PC users would not be able to differentiate a slow SATA II SSD from the fastest benchmarked SATA III SSD. Part of the reason is because the benchmarks show the optimum performance and not the typical performance once an SSD is load with files and other debris. The benches should be used more as a reference point but not the total basis for deciding if an update is necessary or a reasonable value.

Most PC builders update the hardware when they build a new box so they have matched, contemporary hardware. There are some folks who just like new toys who change HDDs or SSDs or CPUs or GPUs more frequently, because a PC is more for entertainment than as a useful tool. It all depends on the person and the situation. Some folks don't update for 5 years and they are completely happy.

Edited by Beenthere

Share this post


Link to post
Share on other sites

I'm hoping the StorageRivew is reading this, I signed up mainly for this reason. since theres a similar thread I will just say it here. I hope all future reviews, especially for Hard Disk Drives, should definitely include a 512byte or 4KB sequential read/write. main purpose for this is because copying a ton of little files will be extremely slow, and not knowing it will be difficult for me to make a decision to purchase a device.

an example:

WD 750gb Black scorpio 2.5" HDD is rated one of the fastest HDD on the market for notebooks even till today, great access time, great 64/128k sequential read/write and decent random 4k read/write, HOWEVER the sequential 512b and 4k read and write are extremely slow at around 5~34MB/s.

where as compare to some not so famous 500gb old seagate HDD top out at 110MB/s sequential 64/128k read/write with quite high access time, the 512b and 4k speed is 12~65MB/s.

as you can see if we simply use it for loading programs, or transfer large block file size, WD black scorpio would win due to low access time and high sequential read/write. HOWEVER for small files transfer including text files, window files, documents, tiny game files, and pretty much everything else, seagate easily comes out on top, and easily DOUBLE performance of WD black scorpio.

please, PLEASE include 512b or 4k (either one is fine) sequential read/write test in your future HDD reviews, thank you.

Edited by unityole

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now