Interesting Results Testing Robson Intel Turbo Memory on F3Sv-a1

hmmmmmm
becuase they are all the same numbers!!! within 1% of each other. So i took the average!!! what confuses you about this? i really dont get it. if you run xyz loop and test one gets 3000 points, test 2 gets you 3030 points and test 3 gets you 3015 points. wouldnt it make sense to just say that it scored 3015 points. its not rocket science!

 

So you...tape the module...into the computer? I guess it works, but I didn't expect it to be that ghetto.

 

No you cover it for its own protection, kind of like insulation and put it into the port where it is posed to be. nothing ghetto about it. its a mod like any other, like many others i have made over the years. if you found something better please enlighten me.

 

http://www.zdnet.com.au/news/hardwa...l-s-Turbo-Memory/0,130061702,339278180,00.htm

 

I find some things in that article very interesting. particularly their mention of adding extra ram to boost performance. If they did thorough testing they should have found that tests that show robson is working are the same tests that would be unaffected by adding system memory. That statement is the equivalent of saying "we will not offer engines with more cylinders because adding a 6th gear to the transimission effectively increases fuel efficiency" robson is not a compliment to system ram and system ram is not a compliment to robson. pcmark tests show the same within the hard drive tests, the only tests to show that robson is actually working "under the hood". forget system ram. it has no effect on robson, robson has no effect on it.
On any day iam going to take the word of intel and microsoft over hp and sony any day. can any one actually name one product that hp actually designs and manufactures? i dont even think they manufacture their own printers. maybee that they do.

 

Isn't it easier to just post the results rather than argue?

 

The article is not saying that, its talking about overall system performance and windows start up time right? Its saying adding more ram does more than turbo memory would. I like what it said about having the option to increase memory in the future using readyboost via sd or usb, whereas you are stuck with the amount of memory you get as turbo memory.

 

Hey no offence but 1% taken away from 4600 is 46 points
1% taken away or added to 3600 is 36 points. i didnt do the calculations just now on a calculator but i am pretty sure i am right. something tell me you are going to try to proove me wrong here though
Edit
Well i guess you figured that out on a calculator as you just removed your post. it would be easy to post all the scores 3 times over but as i said before it would only show the same score with 1% difference 3 times over. Why would i do that? please tell me your logic. why do you want to see almost identical scores 3 times over?

 

I understand that adding ram to a computer may boost system performance. that is a given!
How can you argue against a technolgy by siting evidence of a completely sepaprate technology benefit? i just think it does not belong in a conversation about robson because robson works on a whole separete level not benefiting from the ammount of system ram. one has nothing to do with another.

 

And watch as he ignores this simple concept, which I have been basically yelling at him for pages.

 

I hear ya Alex, like ive posted before..

"Why not post all individual results for the test you did originally? I have asked this 3-4 times now, do they not exist, did you not keep them, at least say something on this matter."

 

I already posted them. figure it out how to add or subtract 1%. use a calculator if you have to.

 

Here is the video of F3SV-B1 boot time, both laptop came with 1GB RAM, let's see which one boots faster by adding Intel Robosn or upgrade the RAM:


<object width='425' height='350'><param name="movie" value="http://www.youtube.com/v/0kgwF_djBSo"></param><param name="wmode" value="transparent"></param><embed src="http://www.youtube.com/v/0kgwF_djBSo" type="application/x-shockwave-flash" wmode="transparent" width='425' height='350'></embed></object>

 

We don't want the damn PCMark scores Eddie stop playing dumb and make with some solid real world benchmarks.

EDIT: Sweet vid Ken, but could you find both at 1Gb with and without robson?

 

Yes I have, will upload in a minute.

 

eww ^^; both are slow. Im used to my desktop booting up in like 30 seconds with my single core opteron

I think upgrading the ram over getting the robisin is a clear choice and always has been, what we need to do now is determin if the robison itself gives any benifit esp in a system with 3gb of ram in it already (or 4)

So another test request, besides the 1gb stock vs 1gb + robison

can we get a 3gb stock vs 3gb + robison.

If its extra performance then hey, why not get it. If its no diffrent than its a waist. Also while I havnt really dug into the details I know I have heard of accounts of performance loss with it instead of gains.

 

F3SV-B1 1GB RAM with and without Turbo Memory:

<object width='425' height='350'><param name="movie" value="http://www.youtube.com/v/2qI3qLfcf9g"></param><param name="wmode" value="transparent"></param><embed src="http://www.youtube.com/v/2qI3qLfcf9g" type="application/x-shockwave-flash" wmode="transparent" width='425' height='350'></embed></object>

 

Hmm well thats no good now is it??

sort of makes the 3gb test pointless if it dosent even help with 1gb of ram.

im sure the counter argument to this will be somthing like it may boot slower but system performance as a whole will be better, or that on first boot the cache is clear so it needed a first boot to cache the boot data for a quicker boot next time.

Not quite sure what it will be, so imma go grab some pop corn and enjoy the show.

(not really im at work and we dont have any popcorn)

 

Conclusion:
I will spend my money on more RAM even just 1GB more, not only it helps on boot time without install any drivers or wait for drivers to load, it also helps on game play which Robson can never provide, adding more RAM almost help on everything you doing, period.

 

Why are people even considering getting a laptop with only 1gb of ram, especially with vista these days.

 

Yeah also the RAM are so cheap now.

 

exactly, im only interested in what diff turbo mem makes ona 2gb system, this article is pretty interesting though...

http://www.zdnet.com.au/news/hardwa...l-s-Turbo-Memory/0,130061702,339278180,00.htm

 

"The greatest improvement came as a result of adding more actual RAM to the system. "We added 1GB of RAM and saw a much higher improvement in performance compared to using any of the ReadyBoost or Robson technology," Doddridge said. He added that: "If you have enough system RAM in the system already, ReadyBoost doesn't give you a lot."

My video shows the exact same result. another disadvantage I think is, Turbo Memory going to ruin a fast high capacity 7200rpm drive in a laptop if you have one.

 

If thats the case im definetly not getting turbo memory then, im currently looking at a t7500 2.2ghhz laptop, 2gb 2x1gb dual channel 667mhz ram, 160gb hd 7200rpm.

 

So case closed.

GentechPC and everyone else: 1
Eddie: 0

 

@AlexOnFyre and Gi1bo : You have already taken it far enough, Eddie is one of the most knowledgeable dealers on the forum who has helped many users on the forum and has been here for a long time, if you don't stop your aggressive attitude and can't have a proper argument, I will be forced to use my moderator powers here.

Now on the subject :
The Robson is a flash memory, if you want to compare the power of the flash memory compared to a hard-drive. The seeking (Random access time) on the flash memory is waaaay lesser than a hard-drive. You can see it here : http://www.anandtech.com/storage/showdoc.aspx?i=2982&p=4
(SSD is flash memory too)
Now the function of the Robson here, is to act like a proxy between the hard-drive and the computer memory (see DMA for more info, the data always travels from RAM to HDD and from HDD to RAM). Now, what this proxy function will reduce is the speed of sequential write and read. It means that if you are reading/writing big files with minor changes and seek operations, the Robson will actually reduce the performance like Ken demonstrated here.
But there are some operations and applications that need continuous read/write operations to a hard-drive. Like a database that is continuously updated. In this case (specially if the database is rather small, like 200MB) it means that you will see some good performance improvements in this area using the Robson memory.
Yes, these improvements might range from nothing to even reduction for certain users, but bashing the technology altogether and accusing the dealers just because your (and many other user's usage) is different is not a proper thing to do at all.
And if you have read Eddie's first post, you will notice the part he ran through many benchmarks to find the point of this technology. And the PC Mark is much more of a real-life benchmark than the 3D Mark, PC Mark uses some scripts and simulates the usage of the user on different programs. And It can be used as a proper measure for the performance of the system.

 

phew, thanks mujtaba.

ok, does anyone know if the Robinson tech will be released in the form of a Xpress card / PCI-e card , or mini- PCI card? So that users with silghtly older notebook can make use of it. Or is it only soldered onto the motherboards. Reason I'm asking is that I heard that intel are working on a PCI-e version for desktops.


2nd question, does this Robinson tech only work on vista, or can it work on linux and XP too?


thanks

 

Ken
i ran multiple 3dmarks, sandra and all the utilities out there including pcmark o5. In a test environment i found yor claim to be comopletely unsubstantiated. I dont want to thorw it out there but i will. i prooved you wrong with the full sized card in the g1s. now you know it CAN be done. I prooved your wrong with the a8js bluetooth prooving it can be done in under a few minutes not hours. If you feel the technology is not good then why the heck do you offer it on your website.? Up until now i had no problem with you. I considered you an assett to this comunity and never felt the need to rub your nose in anything. Please dont bump into me in the future and we will remain on different boats sailing to the same location. okay? The controlled pcmark test i ran showed absolutely no negative effects of file writes. controlled tests mean there is no background activity.

wrong!!! the ammount of system ram and performance as a result of the ammount is unaffected by robson. Unnafected because the results are lowered by less then 1% in some tests. the 50% increase in transfer rate as a result of robson is worth a mention. the 1% change in memory or another aspect is not. Go back and retest. this time make sure you nderstand the driver and the technology first. if you dont believe in it then remove it from your website. that is a little contradictory dont you think?

 

I belive fujitsu will be one of the first to come out with hybrid drives and pump money into technology that immitates robson. you will see it on desktops and laptops alike. i am not too sure what the future if itm is but you can be sure that it will grow and the modules will get larger. Intel and ms will also continue to include support for it in future chipsets and it will only work better as time goes by.

 

Has nothing to do with it. Please explain how when pcmark resusts clearly show opposite. your video shows good results when adding ram. if you havent read it many times.;....addint robson does not compliment results that are complimented by ram. Robson benefits in completely different tests where ram does not have any effect.

 

Ken is quoting very early results from toms or another reviewer which did not do proper testing. Anand basically revised his early review so will toms and many others. PCmark prooves conclusively that with robson a 7200rpm drive benefits in the same 3 out of 5 hd tests as does a 5400 rpm drive by about the same %%. I dont think he meant ruin i think he meant "dimish performance" which is 100% not true. Also he is quoting an article which insinuates that there is a connection between the ammount of system ram and robson. Under controlled environment i found exactly the opposite. The testing i did was a bit more thorough

 

Ken's video said exactly what everyone else is trying to tell you, but you are too stubborn to listen (or desperately grasping at straws to save your argument).

Regardless of how well Robson works in the hardware, it won't improve real world performance since it is badly supported by software.
Your tests proved the BIOS works well with it.
Ken's test proved that Vista doesn't.
Robson is a software (read: Windows) activated feature, ipso facto if the software (read: Windows) doesn't support it, it isn't going help the software (read: Windows) go any faster.

I'm not saying that you are wrong in your testing. I am saying you are wrong by suggesting that Robson will enhance your real world performance on this notebook or the G1S.

Lastly, it cripples a 7200 RPM drive because the 7200 RPM drive (high capacity), has close to a 3Gb/s transfer rate, and the latency between Robson and the RAM and between the HDD, that it is essentially working only just as fast as the HDD is, but making the path between the HDD and the RAM longer.

 

Hi Geared2Play interesting topic as always.. First of all to clarify from what i read in the posts i really dont know what peoples problems are? I think the point of this thread is to show if the robson tech does what it claims to do which was clarified in the first few pages of this post.. Second I dont know why people keep comparing ram to robson both have different benefits and different price ranges.. Its like saying i wont buy a new video card with dd3 because it doesnt speed up typing documents in MS word.. And i dont get why people are questioning the benchmark scores to represent real world benefits(thats the point of benchmarks in the first place not unless you wanna spend an entire week to a month timing everything with a stop watch).. And if the man doesnt want to display each freakin result which varies by 1% he has the right too if u dont like it go do ur own.. If another site disagrees with his result then its up to you which one you want to believe.. But if one of the users (especially AlexOnFyre) claims that it has no real world benefits, then id want to see the benchmark he did to prove his claim rather than pointing me to a site or an article that proves him correctly..
In conclusion from what i read in the post(which i dont get why people keep arguing) If u have enough money to buy ram and a faster hdd, buy it; if u have extra buy robson; if u dont have enough money to buy ram or a better hdd and u just cant wait to save up and somehow want to improve ur system buy robson. Yeah you might call me biased well i am. I side with the person who has concrete results not with people who keep shooting down people who give time and effort(without payment i might add) based on other sites or articles.. If people from toms hardware or something were arguing with geared2play that would be a different story.. Pls next time play nice and be more polite thank you.. If you dont agree with me or think im wrong then dont reply quote by quote just say it once (politely please) and never read this thread again simple as that.. Alot of people in this site benefit from people like geared2play who try to test the waters of new technologies. that would be all thank you

 

Alex please read carefully, whatever tests ken ran with robson and 3dmark are flawed, I had to rerun my own test becuase I got negative results only when the module is in but turned off or simply not working, I reran multiple tests multiple times for a reason, I repeat again when robson and drivers are working right robson has no effect on 3dmark scores. Kens tests are flawed if infact they show otherwise, it doesn't matter how much ram is present, stop rubbing other peoples tests in my face, they are flawed, just like anands early tests were due to driver issues and bios issues, I feel like I have to yell that robson does not have any adverse effects, atleast it did not on the g2s and f3sv

 

What are you talking about with 3DMark? I am talking about the videos posted on the last page, which show two identical systems, one with Robson, one without, and they both load at the same rate (even seems like the one with Robson is a couple seconds slower). That isn't a benchmark, that is straight up showing, physically and beyond any doubt, that Robson does not provide a real world benefit.

 

"I had to rerun my own test becuase I got negative results only when the module is in but turned off or simply not working."

And I really don't see why you bother arguing. He's not telling people to BUY Robson from him or from others. He's just stating that Robson, when operational and properly installed, decreases HD load times.
He's not arguing that it is better than getting more memory, or getting a HD with faster RPM (Rather, he concedes to this). He's saying that Robson, when drivers and stuffs are finally fixed, will have its benefits.
And all you're saying is "No no no! Your tests are bad! Robson doesn't work because [insert source] says/shows it!"
He's saying that "[insert source] are doing flawed tests because they're testing for the wrong thing OR Robson is malfunctioning."
I'm not sure how someone can check to see if Robson is working properly though.
As for giving him negative ResellerRatings because you disagree with him? WTF? His technical support and stuffs are grand.
I do think that Eddie needs more tact though. But oh wells, he is who he is.

 

Alex those tests are flawed as well. Most tests out there are flawed. I have no doubt that the videos you are referring to did not have the module working correctly in addition did not have a propper test environment, and ideal test would be one that I will make some time in the near future. Flawed results are nothing to point to, the only reason I sepnt so much time tinkering with robson is becuase I was so sure that most articles out were flawed in some way and I have no doubt that they were. Lazyguy you are right, I curse, drink, smoke, argue and have a temper, hope ur notebook is doing well after its last rma. My tact is not what fixes our products, it is what gives me my personality. I take my work very serious sometimes I even take it home. As I said above to alex, before you accuse me of anything find a single dealer that has accomplished more then we have and then ask him what he thinks of me. Alex if u don't like it then stfu or come back with your own results instead of quoting someone elses flaws. You have not pointed to anything but flawed or improper tests thus far. Stop babeling "the path is too long between the ram and the hdd so it cripples the 7200rpm" bla bla bla. Just shut up you don't even know what you are saying your self, how do you expect others to get it?. Point me in the direction of a proper test with proper results because this technology is going to be a huge success, robson and every variation of it in the near future. Stop telling me about results that are already buried in the ground.

 

Lazy the module drivers are flawed ina few ways. The biggest problem I had which is the same as anand in his phd thesis that really says nothing but the fact that he is an expert linguist was enabling the module, it is not as simple as turning it on and off. Iam not entirely sure how or why it malfunctions but once it starts working it doesn't stop, thus the 3 loops of each test, adverse effects were encountered mainly when both readydrive and readyboost boxes were not both checked, checking and rebooting repeatendly was one way to fix it, even after both checked it took a few reboots before robson kicked in. Once it kicked in the results were predictable and identical every loop, I suspect meost people did not spend a whole week testing and retesting and throwing out faulty data, I suspect most ran one or 2 tests, I ran dozens every day, everything from 3dmark, to pcmark to 3dmark again. An easy way for me to tell that the module was on and working was mainly by running loops of pcmark over and over. I removed the module and put it back so many times that I lost count how many hours I spent testing retesting and throwing out flawed data. anand in his article also came to the same conclusion however his tests on another notebook for some reason revealed slightly lower writes while my tests on two notebooks did not. I did not notice any decrease in boot times, but I did notice problems in hibernating when the module was malfunctioning. A review elsewhere mentioned this as an adverse effect when in reality its just flawed data that should be thrown out just like much of the data out there. Another article I read mentioned that the software has not matured yet where vista takes full advantage of the technology. I find this statement right on!

 

some people just dont like me. must be my big grill and personality

mostly access and read. write was unafected in my tests but anand found otherwise. More tests need to be done.

 

watch this: http://channel9.msdn.com/Showpost.aspx?postid=242429

This is an interview with the guy who created superfetch, at about 34:40 he goes into talking about the internal nvram that can be used as extra cache for vista. Might be interesting to some of you.

 

Fine dig your grave deeper...
I reject most of your statements becuase
1. most of them were wrong and inaplicable in this thread
2. they were based on flawed and falty data.
This thread was about new data that has not been presented yet anywhere. You come off as a negative sceptic that bashed me. I dont have to lord my experience, you should have recognized it from the get go. I ve been doing this for many years. Data collection and troubleshooting is my area. few other review sites can claim the same based on all the flawed results that are going around. Notebookreview just happens to be the first site to offer a new view and evidence of the contrary. i was glad to offer my services. I am very loyal to this site. I have been here for many years as mentioned above. If you dont want to work with me then you should not even be here

The most inteesting part of this interview is when he gets into random and sequential i/o
If you look at the 5 hd test results in pcmark. the top 3 are atleast partial or completely random while the botton 2 are sequential. the top 3 have a significant increase in throuput while the bottom two have none what so ever. pcmark was right on according to the interview with the man from microsoft from what i gather. he also mentions that sequential i/o have much greater throughput then random in the first place. pcmark shows the same comparing the top 3 vs the bottom 2. right on!

 

Alex, if u really believe that the 7200rpm HDD has a transfer rate close to 3GB/sec then I would like to say u need to get back to ur drawing boards and do some research or read more on hard disks.. 3GB/Sec transfer rate is just theoretical and on paper. Real world tests wont even give 10% of that transfer rate let alone 1 GB/sec!!

Iv been reading this thread for some time and noticed that all that is being spewed by you or G1b0 is just theories and quotes from sites here and there while Geared2play has been giving benchmarkmark results to support his claims..

Dont get me wrong here, im not trying to take sides. Just try to settle a discussion peacefully instead of trying to be big man around here.. Peace out!

 

Unfortunately this thread has degraded from its original purpose into a less than civilized discussion about people rather than merits of Robson. Its not serving NBR well and should be closed in my opinion.

I'll appreciate Geared2Play's work in showing some promise to Robson and I'm sorry that he was personally attacked after making a post that should have remained about Robson and not about personalities.

 

Thats right. potentially there is potential for 3gb/s but realistically the spindle, needle, rotors, and everything else that is in the way will not allow even a fraction of that. potentially if you strip a lawnmower down to just the engine you can make a helicopter.
I am not the big man around here. the mods and owners here are. I am someone who takes it very offensively when a sceptic newbie who pretends like he knows his stuff because he read a book or 20 tries to rub my face in complete garbage which took someone 1 hour to come up with. I worked on these results for the better part of every work day for a week just so that there were no errors or flaws. Now that you (alex) know what my exprience is what my degree is and what the exprience of my partners are why dont you enlighten us what it is that makes you so sure that technology that was created by intel and microsoft is never going to work. Also what is that made you ask big daddy Ken to come here and negate a topic that neither of you know anything about...

 

Thank you. i agree
It is my fault more then anything else really. I should have asked the mods to remove alex from this thread when he first started getting on my case but i am not one who can resist an argument. I have issues too

 

I have deleted the offensive posts in the last two pages, please keep it under control in here or this thread will be closed. Thanks.

 

I didn't say 3GB, I said 3Gb, huge difference, a factor of 8 difference. That is only 375MB/s =P

In any case, this gives me a wonderful reason to quit this thread (which I have wanted to do for 15 pages, but didn't want people to think buying Robson right now would be magic in box.) So I bid all of you gentlemen ado.

 

Haaah, playin with words now, are we??

Anyhow that is a good one! Hehe..

I didnt realize we started mentioning transfer rates in terms of bits nowdays..

 

Bits or bytes it really does not matter. This is called interface or potential burst transfer. Like mentioned above it has nothing to do with real life transfer rates, sequential reads, random or whatever. Alex your bottleneck logic that says robson is bad becuase it takes this much to get here and the path is longer is absolute rubish. it is in the ground as most of your negative remarks about robson. it does not apply even if it exists. You can not negate a beneficial tehcnology just becuase it has limitations. It is clearly beneficial and it clearly has limitations. You would make a horrible lawyer, that is all you have prooved in this thread