What makes you a geek?

[Treesh]

Ooooh, forgot, I watch a lot of anime! Anime = geekiness, yay! Bleach, Ghost-07, Full Metal Alchemist, Full Metal Panic, Claymore, Berserk, Neon Genesis Evangelion, Death Note and many others I've forgotten, watched them all. In Japanese!*

*with subtitling

Anime is still considered geeky? I have been misinformed! Damnitall!

BTW, I really recommend Code Geass: Lelouch of the Rebellion. =) I really enjoyed that one. Denno Coil was rather interesting as well. Subtitles are the only way to go for anime. =P

[Kevin]

Anime is still considered geeky? I have been misinformed! Damnitall!

BTW, I really recommend Code Geass: Lelouch of the Rebellion. =) I really enjoyed that one. Denno Coil was rather interesting as well. Subtitles are the only way to go for anime. =P

Agree with LeLouch for sure. Eden of the East is another I'd recommend for ones that I didn't see mentioned here.

[LavCat]

I'm replying to both Lissa and Concillian here. We all know what application is driving this technology: we have to load our textures faster (err, I mean files). As I understand it, WoW stores textures (and other static data) compressed into large .mpq files. As a texture is called for, I can't believe the game engine starts doing a read through the .mpq file. Common sense tells me the texture data must be stored in a manner that it can be accesssed randomly. Does WoW decompress the .mpq files at load time? How exactly does it work? Is random read performance the parameter that we want to maximize?

I could buy an ST3300655LW today (if I had $475.98, which I don't), plug it in, and without any worries about incompatibilities, reasonably expect that it would last five years (I've only ever had one Seagate drive fail in warranty). Or, for about the same price, I could order a small OCZ Colossus that (following the "buy" link to Amazon) "usually ships in one to three months." Would I really see a difference?

Reading the customer reviews for SSD's on Newegg does not give a warm fuzzy feeling.

[Vandiablo]

You mean cop right? :whistling:

Yeah, yer right. A lot of cops want a better feel.

-V
(cop right 2009 VDP, Inc.)

[Vandiablo]

I didn't even manage to pass page 3........making an integration function just didn't go anymore, let alone solving it. I guess it has been all the drinking.

Yeah, I settle down with a math book these days, read half a page and:
.... zzzZZZzzz ... zzzZZZzzz ... zzzZZZzzz ...

-V

p.s. The following symbols were used in this post:
<blockquote>... -------- Ellippsessess, my Precious, to denote a pause
.... ------- Denotes a pregnant pause
: --------- Colon. Mine has been acting up. It has performed as an understudy and gotten crappy reviews
---- ------ Dashes. Used with p.s.
p.s. ------ Pepper and Salt.
zzz ------ Denotes a sound roughly equivalent to a garbage disposal
ZZZ ------ Denotes a sound roughly equivalent to a 10km asteroid striking the Earth
-V -------- Denotes a negative sanity vector, or -5, depending
</blockquote>

[Crusader]

Anime is still considered geeky? I have been misinformed! Damnitall!

BTW, I really recommend Code Geass: Lelouch of the Rebellion. =) I really enjoyed that one. Denno Coil was rather interesting as well. Subtitles are the only way to go for anime. =P

Already seen all of Code Gaess. It had it's moments. Denno coil I didn't really like so I stopped after 3 episodes.

Agree with LeLouch for sure. Eden of the East is another I'd recommend for ones that I didn't see mentioned here.

Eden of the East seems interesting. Might pick it up.

[Concillian]

Articles I've seen from Tom's Hardware, Anand Tech, and a variety of other hardware websites that have performed true non-sequential tests comparing SSDs and Spindel drives have shown that in non-sequential reads/writes, SSDs are little better than their spindel based brethern showing speeds usually 3 to 4 orders in magnitude less than their sequential read and write speeds.

non-sequential reads / writes are "little better" than rotating storage? Did you even read the AnandTech article I linked to? 30-50x faster in both reads and writes is a "little better"? No, that's more than a little better. 30-50x faster...

sequential read and write speeds are 3-4 orders of magnitude less? In that article the SSD drives are 2x faster than a 10k RPM SATA drive in sequential reads and about half the performance in sequential writes.

so 1/2 the speed of the spindle drive is 3-4 orders of magnitude less? 3-4 orders of magnitude is 0.001 to 0.0001 (1 / 1000 to 1/ 10000)....3-4 orders of magnitude is how far off your reading comprehension is from reality. I linked the article, you chose not to read it, and then formulate an argument based on the hope that I didn't read it either?

Yeah, they lose on sequential writes. Who cares. The whole basis for my argument is that sequential throughput barely matters because normal users don't do large sequential transfers. Normal usage is closer to random than to sequential. It's the whole basis of my argument that RAID is a minimal performance benefit, and it's not something that's changed since the RAID article I linked was written. If anything, growth of file sizes and drive sizes would drive the spectrum further towards fragmented files.

[LochnarITB]

What makes me a geek? I continue to read a good hearted self-degradation thread even though it devolves into a battle over hard drives and controllers.;)

[Lissa]

non-sequential reads / writes are "little better" than rotating storage? Did you even read the AnandTech article I linked to? 30-50x faster in both reads and writes is a "little better"? No, that's more than a little better. 30-50x faster...

sequential read and write speeds are 3-4 orders of magnitude less? In that article the SSD drives are 2x faster than a 10k RPM SATA drive in sequential reads and about half the performance in sequential writes.

so 1/2 the speed of the spindle drive is 3-4 orders of magnitude less? 3-4 orders of magnitude is 0.001 to 0.0001 (1 / 1000 to 1/ 10000)....3-4 orders of magnitude is how far off your reading comprehension is from reality. I linked the article, you chose not to read it, and then formulate an argument based on the hope that I didn't read it either?

Yeah, they lose on sequential writes. Who cares. The whole basis for my argument is that sequential throughput barely matters because normal users don't do large sequential transfers. Normal usage is closer to random than to sequential. It's the whole basis of my argument that RAID is a minimal performance benefit, and it's not something that's changed since the RAID article I linked was written. If anything, growth of file sizes and drive sizes would drive the spectrum further towards fragmented files.

Maybe I wasn't clear or you didn't read it the way it was intended. The loss in read on random reads is for the drive itself. Let's take a standard SSD now which is doing around a sequential read I/O of 200 ms. During a random read, the I/O on said SSD drops from 200 ms to around 50 to 60 ms, that's where the 3 to 4 magnitude loss occurs. In spindle drives, this loss is usually around twice at most. From what Tom's Hardware showed (and I can't find the article at the moment), doing random read I/O between an Intel SSD (s'posedly the best on the market) and a pair of WD Raptor drives in a RAID 0, the Intel SSD was only slightly faster than the RAID 0 Raptors.

The other thing that all SSD manufacturers tell you not to do is defragment your SSD which means it will get fragment far more than a spindle drive over time as you work with it. Spindle drives, you're told to defragment on a regular basis as part of normal maintenance. So as time goes by, your I/O on an SSD is going to go down as it fragments more and more while the spindle drives will likely stay about the same as you defragment them as part of normal maintenance.

[Concillian]

During a random read, the I/O on said SSD drops from 200 ms to around 50 to 60 ms, that's where the 3 to 4 magnitude loss occurs.

Again, where are the orders of magnitude? 50/200 is 1/4, not even one order of magnitude. Why are you using numbers that don't even support your claims? You disprove yourself in your own writing, if an attempt at humor, it's not terrible, but an attempt at serious claims? At least it's still humorous.

And your claim is that rotating storage drops less from sequential to non-sequential than SSD? That's simply not true. That's the whole advantage of SSD, non-sequential transfers are fast due to seek times that are an actual order of magnitude (not your incorrect definition, but the real definition of order of magnitude) lower than rotating storage.

Again, the AT link shows rotating storage both dropping more and having overall performance less than SSD when comparing sequential and random throughput. If you wish to convince me otherwise, you're going to have to stop making things up and show me a link that has rotating drive and SSD benchmarks demonstrating a counterexample to the AT review. Because the Tom's SSD reviews don't have rotating drive comparsions, so don't support what you've said and the most recent AT review certainly doesn't support what you've said. Where are you seeing data to support what you're arguing, I don't see it. Show me.

[Lissa]

Again, where are the orders of magnitude? 50/200 is 1/4, not even one order of magnitude. Why are you using numbers that don't even support your claims? You disprove yourself in your own writing, if an attempt at humor, it's not terrible, but an attempt at serious claims? At least it's still humorous.

And your claim is that rotating storage drops less from sequential to non-sequential than SSD? That's simply not true. That's the whole advantage of SSD, non-sequential transfers are fast due to seek times that are an actual order of magnitude (not your incorrect definition, but the real definition of order of magnitude) lower than rotating storage.

Again, the AT link shows rotating storage both dropping more and having overall performance less than SSD when comparing sequential and random throughput. If you wish to convince me otherwise, you're going to have to stop making things up and show me a link that has rotating drive and SSD benchmarks demonstrating a counterexample to the AT review. Because the Tom's SSD reviews don't have rotating drive comparsions, so don't support what you've said and the most recent AT review certainly doesn't support what you've said. Where are you seeing data to support what you're arguing, I don't see it. Show me.

Then how do you explain tests performed by a reputable hardware site showed that when doing random reads that an SSD was little better than two RAID 0 spindled drives? The numbers quoted by Tom's was literally a less than 10% difference between the SSD and the two RAID 0 spindled drives (WD Raptors in this case). How do you sit there and say that the difference is huge? I remember the numbers shown by Tom's was something like an I/O of about 110 for the SSD and the RAID 0 Raptors was about 100 to 105. Care to revise you claim that SSDs on random reads/writes are light years ahead of RAID 0 drives?

Simply put, there just isn't the price to performance ratio at this time to warrant SSDs. Maybe when SSDs dropped by about half their present price per G they'd be worth it, but right now they're just too expensive to warrant it when putting some drives in a RAID 0 can keep up.

[Concillian]

Then how do you explain tests performed by a reputable hardware site showed that when doing random reads that an SSD was little better than two RAID 0 spindled drives? The numbers quoted by Tom's was literally a less than 10% difference between the SSD and the two RAID 0 spindled drives (WD Raptors in this case). How do you sit there and say that the difference is huge? I remember the numbers shown by Tom's was something like an I/O of about 110 for the SSD and the RAID 0 Raptors was about 100 to 105. Care to revise you claim that SSDs on random reads/writes are light years ahead of RAID 0 drives?

What article? What benchmark (many benchmarks are extremely throughput centric, often sequential centric, where IO meter, with intelligent weighting of the various types of transactions is usually the most representative of actual usage)? Were these Intel SSDs before the major firmware revision 6 months ago that significantly boosted they're non-sequential performance? Were they drives so old that they aren't representative of current SSD drives?

Show me.

At this point I'm pretty certain that you aren't knowledgeable enough on the subject to have a valid opinion, I just want to see what article fed you misinformation so well that you're willing to ignore all that I've presented.

[Lissa]

What article? What benchmark (many benchmarks are extremely throughput centric, often sequential centric, where IO meter, with intelligent weighting of the various types of transactions is usually the most representative of actual usage)? Were these Intel SSDs before the major firmware revision 6 months ago that significantly boosted they're non-sequential performance? Were they drives so old that they aren't representative of current SSD drives?

Show me.

At this point I'm pretty certain that you aren't knowledgeable enough on the subject to have a valid opinion, I just want to see what article fed you misinformation so well that you're willing to ignore all that I've presented.

Here's one of the articles from there, take a close look at it before you keep touting that SSDs blow spindel drives out of the water.

SAS RAID vs. Mitron and Samsung controller SSDs

I can't find the other right now that was directly an Intel 80-M vs. the pair of Raptors, but as you can see from the above article, more often than not, the spindel drives had higher I/O than the SSDs when placed into a RAID configuration.

[Concillian]

Here's one of the articles from there, take a close look at it before you keep touting that SSDs blow spindel drives out of the water.

SAS RAID vs. Mitron and Samsung controller SSDs

I can't find the other right now that was directly an Intel 80-M vs. the pair of Raptors, but as you can see from the above article, more often than not, the spindel drives had higher I/O than the SSDs when placed into a RAID configuration.

Did you read the article and realize that they're formatting the drives to ~1/10th their full capacity in order to make the rotational drives approach SSD performance? So you're arguing that we should not only use RAID, but also cripple the capacity in order to compete with SSDs? The full capacity formats don't compete with SSDs, except the 4 15kRPM drives in RAID0. So yes, they can compete if you like buying 4 15kRPM SAS drives and a SAS RAID controller.

Mostly this demonstrates exactly the points I've been trying to make. Look at the 4x 7200 RPM drives in RAID-0 (at full capacity format). They barely hang with a single 15kRPM drive at full capacity, and they don't hang with the SSDs at all. You can't RAID your way to lower latencies and RAIDing 7200 RPM or 10k RPM drives only gets you so far.

When seek times are very low (15k RPM 450GB drives formatted to 20 GB capacity) then, yes RAID-0 scales well. Because latency is reduced to the point where throughput has a much higher weight on overall performance. You can also see this in the "random access" IOmeter profile (web server). Here the SSDs dominate. Consequently, most users have a really large number of small files that they access on a regular basis (ever looked at the distribution of file sizes on your drive?). Actual performance for normal users is approximated with the Workstation profile, but in my experience, this still weights random small files too lightly. The delta in the small random profile (web server) is so large that even if you give it only a 5% weigting compared with workstation, it almost doubles the SSD scores. So even a small amount of small file random type of activity will be perceived as a benefitby an end user. I don't pretend to know an exact %, but in my experience, it's enough to be a noticeable win for SSD in actual perceived performance from the person sitting in front of the computer while doing actual things on the computer which is really difficult to quantify in a benchmark. It's like the single core / dual core transition a few years ago. Dual core gave real perceptible benefits that just weren't well captured in bencmarks.

So yes, if you're willing to de-stroke your drives to an abysmal capacity, then rotating storage can hang with SSDs if they're in a RAID. This article does a good job of demonstrating how much hinges on reducing latency, which is what I've been saying all along... latency is the dominant factor in end-user performance.

I do thank you for finally linking something, it was an interesting read.

I'll leave the thread with an interesting quote from Anand about how deceptive benchmarks are with SSDs:

[comments are based on a pretty old SSD that was sold with the Macbook Air]
The benchmarks for that drive didn’t really impress. Most application tests got a little slower and transfer speeds weren’t really any better. Application launch times and battery life both improved, the former by a significant amount. [...] It benchmarked faster than hard drive, but the numbers didn’t justify the cost. I pulled the drive out and sent it back after I was done with the review.

The next time I turned on my MacBook Air I thought it was broken. It took an eternity to boot and everything took forever to launch. Even though the benchmarks showed the SSD shaving off a few seconds of application launch time here and there, in the real world, it was very noticeable. The rule of thumb is that it takes about a 10% difference in performance for a user to notice. The application tests didn’t show a 10% difference in performance, but the application launch tests, those were showing 50% gains. It still wasn’t worth $1000, but it was worth a lot more than I originally thought.

It was the MacBook Air experience that made me understand one important point about SSDs: you don’t think they’re fast, until I take one away from you.

These are the exact kinds of comments people were making about dual-core processors. They were faster in benchmarks, but nothing to write home about... until you experienced one and then turned off one of the cores. There's more small file random reads than you think in a normal user's usage profile. Also consider that this is an experience with one of the first consumer SSDs, 2 generations older than current SSDs, which have made very significant advances.

[LennyLen]

I think Lissa and Concillian have demonstrated quite well why they are geeks.

[Zingydex]

Han Shot First.

That only matters when geek points are being considered.

I'd say it matters quite a bit when you consider the plausiblity of the scene from a professional reviewer's standpoint - and they love to pick stuff like that apart to use as ammo for giving thumbs down to all things of interest to geeks. Greedo couldn't possibly have missed at that range, nor could Han have outright dodged the shot. The only way he could live to continue being part of the story is if he shot first.

[kandrathe]

I'd say it matters quite a bit when you consider the plausiblity of the scene from a professional reviewer's standpoint - and they love to pick stuff like that apart to use as ammo for giving thumbs down to all things of interest to geeks. Greedo couldn't possibly have missed at that range, nor could Han have outright dodged the shot. The only way he could live to continue being part of the story is if he shot first.

I'm ok with that. Sometimes, knowing what you know about someone else, shooting first is a great defensive strategy.

[Vandiablo]

I'm ok with that. Sometimes, knowing what you know about someone else, shooting first is a great defensive strategy.

And sometimes, what the first shooter "knows" is based on dubious "facts" filtered through paranoia and hatred.

Having a blaster pointing at you is a fact not left to misinterpretation.

[KingOfPain]

I guess I can claim some geekness if I have to, but you know, I just have to try and be different. What makes me a nub?

Years ago when I was working in R&D on virtual reality, I had to install a dial-up modem on my traveling computer but failed at the attempt. I called in a technician without giving it a second try because I figure I just saved the company a bunch of money. It's way cheaper to pay him $50 an hour than paying me $/hour to RTFM.

I purchased a digital answering machine 2 years ago and it's still not hooked up. The fine prints are too small for my failing eyes to read comfortably, so I asked the kids to set it up and record a message and get it online. It's still sitting there unused. My eye problem probably started when I was debugging programming codes on a TV screen - trying to discern if it's a. or,

When the timer on the vcr (yes, vcr, you know, tape) blinks I will ask whoever is around why it's blinking, because I am too lazy to reset it. That's what kids are for, right?

[LavCat]

I decided to try to answer some of these questions for myself. It still doesn't look like I will ever be able to afford a new SCSI drive, but I recently received a little Intel X25-V SSD that cost less than $125. The drive worked out of the box in Windows and was great for WoW, but getting it optimized has been a nightmare.

Intel says to update the firmware to the latest version, but the firmware update software didn't find the drive. Disconnecting other drives did not help, nor did ripping out the SCSI controller. I eventually updated the firmware by connecting the SSD to the first SATA channel. This was trial and error. Oh, and the SATA controller has to be in IDE mode, can't be RAID either.

Next I tried to align the partition. Windows has a tool, DISKPART, for creating partitions. I could create partitions all day long, but the alignment didn't work. Turns out it doesn't work in the XP version of DISKPART. In desperation I booted from my Windows 7 upgrade DVD, clicked through a bunch of stuff and got to a command prompt. I typed DISKPART and it worked. I could create an aligned partition. Went back to XP and was pleased to see the partition recognised. I could then format, and I used 64K allocation units.

Then it was time to run Intel's SSD Optimizer. Of course it didn't work. Had to go back into the BIOS and set up SATA in IDE mode. Then the software pretended to work, but just sat there for hours and hours. Many hours. Turns out it didn't like 64K blocks. Went back to 4K and the Optimizer worked. Also tried 16K and that worked too. All of this is trial and error, no information from Intel helped (except the part about having to be in IDE mode).

I left out a whole bunch of other experiments that didn't work, but I may be seeing light at the end of the tunnel. In very limited testing, WoW textures load noticeably faster. Anyone else have experience with WoW on an SSD?