Intel Quick to Respond to X25-M 'Slowness' Claims
Posted 02/20/09 at 02:30:10 PM by Paul Lilly
Comments
Print
Email
According to a report published by review site PCPerspective, Intel's advanced sector remapping and wear-leveling algorithm used in the company's X25-M SSD is causing the drive to suffer serious performance degradation over time. In some cases, the site noted reads had been reduced to a pokey 22MB/s. The only solution PCPerspective could come up with to restore the once speedy SSD back to its original performance level was to use a dated version of HDDErase.
Not so fast, says Intel in response to PCPerspctive's claim that the X25-M had become, well, not so fast. Despite the review site having found a drop in performance in all three of its SSDs, Intel claims it has not seen the same type of degradation in its own labs.
"Our labs currently have not been able to duplicate these results," Intel said. "In our estimation, the synthetic workloads they use to stress the drive are not reflective of real world use. Similarly, the benchmarks they used to evaluate performance do not represent what a PC user experiences."
Intel went on to say that it's completely normal for a PC's drive, whether it be an HDD or SSD, to exhibit reduced performance when filled up, but that PCPerspective's results are higher than what Intel would generally expect. Hence the reason why Intel questions the methodology that was used.
Any SSD owners, Intel-brand or otherwise, notice any performance slowdowns over time? Hit the jump and let us know what your experience has been.
Image Credit: Intel
Do some research
Submitted by mesiah on Thu, 03/26/2009 - 11:55pm
As far as the SSDs go, a great deal of research has been done by anandtech and a few others. Some people have been able to reproduce the drive failure, but only by throwing massive amounts of IO at it for long periods. Something the drive will never see in daily consumer use. As far as Intels benchmark numbers being fudged, they are. The drives are actually faster than intel states they are. And compared to any other drive on the market (including other SSDs) they blow them out of the water, especially when it comes to random writes. Intel SSDs are 10x faster than velociraptor drives at random reads, which is what your hard drive spends the majority of its time doing. Alot of SSDs on the market are flawed right now, especially the ones based on jmicron controlers. Intel has made a very solid product, although it obviously has some inherent issues when stressed to its max. And as far as the worries of the drives sectors burning out from overuse, the drives are estimated to have a span of 5 years under heavy continuous (24hours a day) load, which is impressive.
As for complaining about manufacturers skewing benchmarks, are you serious? Does anyone pay any attention to manufacturer claims? If you did you wouldn't be online reading reveiw sites. The bottom line, comming from someone who owns both phenom and core2 products, no matter what manufacturer benchmarks say, every single person running "real world" benchmarking tests (i.e. ones based on games and software commonly used in the consumer marketplace) can tell you that Intels top proccessors destroy AMDs. Thats all there is too it (damn that was a long run on sentence.) So seriously, I get tired of everyone hating on Intel, McDonalds, and Wal-Mart because they are the big boys, when they play by the same rules as everyone else They just get put in the spotlight when they do the same shit everyone else is doing.
/endrant
Submitted by damicatz on Sun, 02/22/2009 - 9:00am
I'll stick with my Velociraptor, thanks.
I never thought I'd see the day where I'd be saying that hard drives are more reliable than flash memory. But I don't have to worry about my velociraptor slowing down over time or burning out sectors due to exceeding a write count.
This makes no sense at all.
Submitted by Keith E. Whisman on Fri, 02/20/2009 - 11:51pm
This makes no sense at all. Why is Intel trying to act like there is nothing wrong when obviously there is? There has to be a simple firmware fix that Intel can make. Those people at Intel defending a possible defect as one that does not exist are hurting Intel greatly. I hope they wise up and just look at the problem and fix it. Quit being little kids.
"Our labs currently have not
Submitted by gatorXXX on Fri, 02/20/2009 - 1:50pm
"Our labs currently have not been able to duplicate these results," Intel said. "In our estimation, the synthetic workloads they use to stress the drive are not reflective of real world use. Similarly, the benchmarks they used to evaluate performance do not represent what a PC user experiences."
Well, well. So basically intel, your saying that BENCHMARKS do NOT represent real world performance? So if that's the case, how can YOU claim YOUR CPU's are so much faster and better than AMD's accurately? Because you say so? Because your benchmarks designed by you favor your CPU's say so? In essence Intel, your saying your shit is the bomb BECAUSE YOUR benchmarks say so and if OTHER benchmarks provide another analysis not in your favor, your quik to point out that YOUR benchmarks say differently and that everyone else is wrong. Your crap is no better than AMD's or any other company you bully. You said it yourself....benchmarks do not represent what a PC user experiences......Have a great day and FREE WILLY!!
English Fail. Intel is
Submitted by TheMurph on Sat, 02/21/2009 - 1:48am
English Fail. Intel is saying that PC Perspective's benchmarks aren't indicative of real-world performance... which is fair, to an extent, given that PC Perspective is slamming the drive with continuous IOMeter workloads over the course of a few hours or so.
I'm not entirely convinced that the fault is at the SSD level just yet, but I can't really speak more to that until I've looked at a few more things.
Submitted by gatorXXX on Sat, 02/21/2009 - 6:14am
Your right. It might not be all the SSD's fault. But that particular benchmark discovered a possible flaw that Intel, at this moment, won't announce because they say that they haven't reproduced the problem. It doesn't matter who did or did not discover it, the problem is still possibly there. Most or all benchmarks slam the hardware with continuous loads whether it be on GPU's, CPU's, and HDD's. According to most, the fastest wins whether it's a 5 minute benchmark or a 5 hour benchmark! But all I'm saying is that there is a HUGE difference in synthetic benchmarks vs. real world performance. But Intel here is saying OH, um, well, the tests they use aren't real world performance meters, but is intels benchmark system real world or is it synthetic also?? They won't tell you. If it's real world benching, and I totally say that it is not, then it wouldn't be fair to intel of this analyisis by PC Perspective or anyone else. But on their CPU's, they use synthetic benchmarks and gloat HOW much faster they are compared to AMD's!! So if those benchmarks aren't fair, then NO synthetic bechmarks are fair because side by side, in real world performance in day to day activities, I personally see NO difference between my wifes NEW core i7 system and my Phenom II system.
Submitted by TheMurph on Sun, 02/22/2009 - 3:03am
I don't know why you would believe Intel, AMD, or anybody when they gloat that their own product is faster than the competition's. Find me a company that says "our new product is actually slower than our competition," and I'll give you a hat*.
*Okay, so in not pushing against Core i7, perhaps AMD is a little guilty of this. But you know what I'm talking about.
But while your paragraph is long and hard to decipher, you do make an interesting point. Journalists talk all day about statistics and facts when it comes to the products they review, but are you *really* ever going to notice a 10 MB/s difference in hard drive speeds? Or a 15 fps difference on your videocard? Sometimes yes, sometimes no. Regardless, benchmarks exist as a (quasi) scientific way to show differences between products, giving you more ammo to make an informed decision at the store. I know that, as a consumer, performance is a huge part of how I buy products... but so is price. While Core i7 might stomp all over Phenom II processors in the benchmarks, I know exactly what I'd be picking up were I an AMD owner of average financial resources.
But I digress. PC Perspective has illustrated a flaw in the SSD system. But you have to consider a number of factors that, honestly, might not even make the problem the SSD's fault per se. Have you stopped to consider that perhaps this is a software or usability issue that's common to all SSD's, not just Intel's?
Submitted by gatorXXX on Sun, 02/22/2009 - 3:43am
Well, if it were common to all SSD's, don't you think they would have posted the results of the other brands? I'm sure they didn't just benchmark Intel's and if they did, shame on them for not making a complete analysis. But I am sure they have benchmarked all SSD's on the market.
I'm not against Intel per se, nor am I for the little guy, I am all for progression and tech advances that stem from competition. Without competition, there would be no advances or price checks. Am I wrong? While I believe no one is innocent, including AMD, Intel seems to want the monopoly in every aspect of computing by lying about benchmark scores and undercutting the little guys that try to make competition a little more fair. And yes, intel was caught just a year or so ago (check back issues of MPC or google it), fudging with other vendor's benchmarks to make it favor Intel CPU's! They all have done it, but Intel got caught.
And by the way, you bring up some interesting points as well but this isn't English 101 nor is it a journalist haven for correct pronunciation. This is a forum.
Submitted by TheMurph on Sun, 02/22/2009 - 5:42pm
"But I am sure they have benchmarked all SSD's on the market."
I'm sure they're doing that now. But the initial testing was done on Intel SSDs only, as I understand it from various conversations I've had with people in-the-know.
Again, I'm not sure why you're suggesting Intel "lies" about its benchmark scores. There are a few trains of thought on this matter: companies can overinflate their performance based on specific testing patterns that favor them; companies can tweak their products to show maximum, unnatural performance on common benchmarks; and companies can outright lie about what their product can do. The first two are more common in the tech industry than the latter, as that would be in violation of FTC statues -- and subject the company to punishing legal ramificaitons.
And no, I'm not attacking your grammar. I'm just trying to make sense of all the points you make in a single, huge paragraph of text.
Submitted by gatorXXX on Mon, 02/23/2009 - 3:32am
Ok, so your telling me that since SSD's have hit the market, Intel's the only one that has been benchmarked in that fashion? I do kind of find that hard to believe. If that is the case though, then they jumped the gun to point out that Intel has a problem as opposed to all SSD's in general.
I'm not saying they are flat out lying. I'm saying that Intel said themselves that those benchmark scores don't reflect real-world performance usage. They don't but No synthetic benchmarks do in that matter. But they found a potential problem by using a synthetic benchmark and Intel denies it with by saying "Our testing can't reproduce the problem so therefore there is no problem" attitude. If that's the case also, then everything needs to be benchmarked "real-world" to get an accurate measure.
I agree, my grammer is not the best. I went and re-read my paragraphs and they are atrocious!
Submitted by Master123 on Fri, 02/20/2009 - 2:30pm
Intel is saying THEIR benchmark testing does not represent real world performance, not all benchmarks. They claim that their testing methods will not represent what a real user does.
Submitted by gatorXXX on Fri, 02/20/2009 - 4:52pm
UM.....most or if not all benchmarks are NOT indicitive of real world performance. Including Intel's. So intel's way of benchmarking is the only right way? It's almost like intel saying...."NO, you can't test it that way....here, use this! It's what WE use!" With evil grin pursuing.......
Must Read Articles
Feature
Review
Feature
Feature
Feature
This Month's Issue
FEATURE How to Get FREE Programs, Services, Software & MoreFEATURE Digital Photo Printer RoundupHOW TOBuild a 3D CameraFEATUREDIY Arcade PCWHITE PAPERHow TRIM Works
Technology News
Computer Cooling Fans
Computer Cases
PC Game Controllers
PC Games
Computer Hardware
Headphones
MP3 Players
Stream Video
Computer Mouse
Monitors
Motherboards
NAS Storage
Networking
Laptop Computers
DVD Burner
Digital Cameras
Portable Storage
Computer Accessories
Smartphone
Antivirus Software
Sound Cards
Speakers
Computer Systems
Thumb Drives
Video Cameras
Video Card Reviews
Water Cooling
Gadgets
- Keyboards
© 2009 Future US, Inc. All Rights Reserved.