Slashdot Mirror
Microsoft, Sony Clash Over Vista Turbo Memory
Anonymous writes "Sony is claiming that the current release of Vista does not support Intel's Turbo Memory technology, but Microsoft has dismissed the allegation. If Microsoft is telling the truth then all is well. But if Sony is right, Microsoft has opened itself to being sued for deceptive marketing practices."
That wouldn't be the first time Microsoft was sued. What does Sony have that the US-DOJ doesn't?
Sony is dying because of the way they've been treating their customers lately.
By attacking one of the few companies more hated than them, they're trying to re-direct some of their bad karma.
I am a free slashdotter. I will not be modded, blogged, DRM'd, patented, podcasted or RFID'd. My life is my own.
Another niche tech product backfires on its creators. I wonder if they'll ever learn.
vs. Microsoft's vague assurances:
Guess who seems more confident in their assertion?
Tags != Comments, and -1 (Troll) != -1 (I Would Respond Angrily To This Poster So They Must Be Trolling)
...must be one of most redundant statements in the English language.
We used to play a (DOS) joke called "Turbo Copy"
Just press "ALT" then "E" then "A" ... then hit the Deliver key (DEL).
Turbo Copy! 100% data loss, but it sure is fast!
Maybe that's how MS saves their roadmap.txt file.
"Sony is dying because of the way they've been treating their customers lately"
"By attacking one of the few companies more hated than them, they're trying to re-direct some of their bad karma"
http://finance.yahoo.com/q/bc?s=SNE&t=6m
was: Re:Its all marketing...
davecb5620@gmail.com
Sony: Vista doesn't support TurboMemory.
Microsoft: It does too. See? It uses the flash memory for...things. Vroom.
Sony: You call that support? It doesn't do what it's supposed to.
Microsoft: Just because you can't see it doesn't mean it's not happening. It's integrated and magical.
Sony: Yeah, it'll half-work, as long as you micromanage what files are cached.
Microsoft: See? Integration.
Sony: Um...no. Not quite.
± 29 dB
does not work in the way that it was marketed as giving a nice speed up and M$ just pushed it back to vista sp1.
There was a test recently in the german gamers magazine Gamestar, and they found that ''turbo memory'' did nios speed things up at all in a number of different set-ups they tested.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Microsoft? Using "deceptive marketing practices."
Huh.
Who'd of thunk it?
Must be a day that ends with "y".
(rolls eyes)
"The issue is that the OS needs to learn what to load into the Robson memory in order to increase performance," Sony said."
It sounds to me like Microsoft may have implemented it poorly so it's a feature that doesn't really help.
How many people here are old enough to remember the transistor radio? I remember the big thing was to get a five transistor radio. That was a radio with five (5) transistors. And they had five too but if you looked you might see that one of the three leads on two of the transistors were cut.
Unscrupulous companies were putting five transistors into their radios so that they could advertise that feature but they were using two of them as simple diodes not as transistors. What you paid for was a five transistor radio but what you got was a in effect three transistor radio. You couldn't really sue because the unit had all five transistors, just some of them weren't being used as transistors.
The race isn't always to the swift... but that's the way to bet!
Tests run on customer reference boards and preproduction latest generation Intel® Centrino® processor technology with optional Intel® Turbo Memory enabled against like systems without Intel Turbo Memory. Results may vary based on hardware, software and overall system configuration. All tests and ratings reflect the approximate performance of Intel® products as measured by those tests. All testing was done on Microsoft Windows Vista* Ultimate (build 6000). Application load and runtime acceleration depend on Vista's preference to pre-load those applications into the Microsoft ReadyBoost* cache. See www.intel.com/performance/mobile/intel_turbo_memor y.htm for more information.
Which in turn yields:
Performance measurements collected on pre-production Lenovo ThinkPad* T61 with pre-production BIOS. Detailed Notebook Configurations
PCMark05 Test from FutureMark is an application-based benchmarking tool used to measure overall PC performance. By using portions of real applications, this benchmarking tool can assess PC performance. (+36% improvement)
Google* Earth loading a fly through of a national park followed by Adobe Photoshop* Elements 5.0 creating a slideshow showing pictures from the same park. The input files for Adobe Photoshop Elements are 48 digital photos with a resolution of 10 MPel. (+127%)
Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Intel products as measured by those tests. Any difference in system hardware or software design or configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance of systems or components they are considering purchasing. For more information on performance tests and on the performance of Intel products, visit www.intel.com/performance/ or call (U.S.) 1-800-628-8686 or 1-916-356-3104.
But Sony is trustworthy, they'd never lie.
Is Turbo Memory technology hardware that is designed and built around an OS (Vista)? That seems to be a very peculiar (read:bad) idea. What does it mean for other users who intend to utilize different operating systems? Is there a loss of performance or just an added feature that cannot be used?
Microsoft has opened itself to being sued for deceptive marketing practices
Considering they got away claiming they were selling Operating Systems, I don't think this will be a problem.
Opus: the Swiss army knife of audio codec
Sony, look, if Vista is not using the turbo memory technology you could use the free space there to load your root kits even faster....
(rant)
Does anyone actually remember when "turbo" had a technical definition beyond "really fast?" Does anyone realize that, in the computing world, "turbo" is essentially meaningless? (Go ahead, demonstrate for me how you pressurize the incoming bitstream mix using the processor bitstream exhaust pressure...) Or has the influx of market-roids slapping a "turbo" badge on any slightly-faster-than-last-year's technology made this term utterly useless?
(/rant)
To keep it simple, it is flash memory (slower then ram (regular memory), cheaper then ram, not as long lasting as ram) that is added as an extra cache.
That is it, nothing more. Just a file cache. The OS controls it and has to tell it what to cache and what not.
Cacheing in itself is pretty simple and its speed increase is pretty damn obvious to all those of us who have lived through the age of the minimal/full setup for games. The game/data comes on the CD, with it being optional to "cache" it to the HD. The more you cache on the HD, the faster the game will load its data.
Now there are problems with cacheing. What to keep, and what to loose.
Take again a game. Say I a racing game. As I drive around the track new scenery comes up and has to be taken into memory. If it is full then old scenery needs to dropped out. Obviously the machine that has enough memory to take the ENTIRE track into memory will perform the best. Next will be the machine that can at least load it from something like an HD, preferrably a special cache file of the track that combines all the needed data in one handy arrangement, slowest will be the machine that is forced to read the track data from the CD as you drive around.
Turbo Memory(cache) is designed to load frequently used data(applications are data as well) into its memory, so that it can be loaded into main memory faster then if it had to be loaded from HD.
And there is its problem. HD's ain't slow, and it still got to be loaded from the cache into memory. The game engine itself barely benefits from this, it just might reduce the loading time IF your OS deems the game engine to be fit to be loaded. The game data itself will be too big to load. In a linear game you wouldn't even have much to cache, either stuff is needed constantly, and needs to be in main memory OR is used once, and there is no point in cacheing it.
This kind of tech ain't knew. Were it excells is in reducing the startup time of many small often run applications. Were it sucks donkey balls is when it comes to big run once, stay loaded type apps.
What is even worse, AI in OS'es generally just isn't very good and often gets it wrong. In trying to guess what you are doing it will often guess wrong and actually hurt performance.
Turbo Memory works with certain workflows were you would be better off with just more memory and or faster HD but can't have/afford that.
I am therefore not suprised at the Sony and MS reaction. Both are absolutly correct. Sony tested it with their set of tests, and found it not worth the cost. Very likely they just have a certain workflow they test for with memory setups that are designed for that. (Might Sony make more money from selling main memory, then turbo memory) MS will have tested for different circumstances, perhaps those that favor their cacheing system and with the knowledge that MS does NOT sell main memory?
So what does this mean to you? Make sure you check that any review of technology like this resembles what YOU do with your computer. Always run the same apps that stay active, handfull of large apps and can afford/have enough main memory, then don't bother. Are you someone who runs countless little apps, constantly closing them and reopening them and just don't have enough or can't enough main memory, then it might work for you. IF Vista properly regonizes what you are doing and can use the cache as it is intended.
So no Sony OR MS bashing needed here. Simply different views of how users us their computer.
According to several articles regarding this subject, the questionable utility of Turbo Memory is not the fault of MS alone:
http://www.tgdaily.com/content/view/31976/135/TG Daily reports that Intel's showcasing of Turbo Memory included benchmarks that's anything but real-world applicable: "The benchmark appeared to slam several pictures at lightning speed into Photoshop, something that would play to the strengths of flash memory because the pictures would already be stored in flash for fast opening by Photoshop. Realistically though, we think the average user wouldn't capture dozens of pictures and then open them all in Photoshop in one fell swoop."
Which leads to an Anandtech article showing that in many cases, performance suffered as a result of Turbo Memory implementation - particularly with boot and hibernation times. Now these are cases where users are MOST likely to notice performance differences.
Finally, in the cases where Turbo Memory would seem useful, it appears that HP discovered that using far more versatile, ubitquitous flash solutions such as SD and USB drives (not to mention just adding regular system memory (what a concept!)) yielded similar and more economically sensible results: http://news.zdnet.com/2100-9584_22-6188522.html
Maybe if Vista didn't need such obscene amounts of memory, this wouldn't be an issue; but I digress.
It's interesting that Intel themselves calls it an "entirely new system innovation for Windows Vista PCs..." and says that it "Works on Windows Vista only."
Perhaps you can point to the specification which would allow it to be used by other operating systems. If I have a dual-boot system, does the specification allow it to keep info for each? If so, how is it determined which OS gets use of how much of this memory?
"National Security is the chief cause of national insecurity." - Celine's First Law
The death of "Turbo" was when the Turbo Button stopped appearing on computer cases :X Now wasn't that cool or what, you had TURBO and the speed of your procesor was indicated on a led display.
Sounds to me it would be optimally used as a primary swap partition, with the normal hd-based one as backup once the turbo-space is full ?
The world always seems brighter when you've just made something that wasn't there before. - Neil Gaiman
Or maybe Afterburner Powered... that would explain why your laptop gets so hot.
When presidential candidates are debating about evolution you really must wonder how enlightened the masses really are.
Engineering is the art of compromise.
Don't forget exponentially more expensive than naturally aspirated cars. Turbo lag, funny noises, sleepless nights caused by "coking" of various bits of hardware that cost more than your kid's braces, and only marginally enhanced reliability.
As to people who think that "turbo" is simply another word for "fast," I submit that the English language is dying in bits and pieces. You're saying "Yeah, let's just use 'turbo' to mean 'fast,' because they're the same thing." I'm saying "But turbo implies a specific kind of automotive forced induction, and has nothing to do with computers anyway." So the answer: if you must describe a computer component as fast, then by all means, use the applicable term: "Fast." If the marketing types insist on quantifying how fast it is with respect to other blazingly fast technologies, fine: "This is our latest memory offering, capable of data retrieval at blazingly fast speeds. We call it 3GB/nanosecond Memory."
Check out this site for a few examples.
However, if you insist on a superhet design, though, you still can get by with only three...
P.S. I'm not actually old enought to remember this stuff, but hacking topics of any generation are is still pretty fascinating to read about...
The case of the AT box under my table which I use as a router-cum-fileserver has a "TURBO" button on the front display.
The box usually runs OpenBSD, so I tried starting a Vista installation to see what all the fuss was about.
Unfortunately, it appears 64MB of "tradtional" RAM is not enough for Vista.
I have such vitriolic hatred for both companies that I don't know which side to pull for. Guess I'll pull for MS - Sony is the root(kit) of all evil. By the way, am I the only one around here with a fetish for veiny women?
...Sony have a fecking clue about software.
The day I trust Sony's views on what makes good software is the day I call up Satan for his advice on which Snow-Plough model gets you to work fastest.
s/dying/evolving
Actually, French is dying, specifically because there is Official Standard French and the French government won't allow people to use the language they way they want to. English, on the other hand, can be abused and mutated any way you like, and because of that flexibility it has become one of the world's most popular languages.
Not that this has anything to do with memory, "turbo" or otherwise.
I don't care if it's 90,000 hectares. That lake was not my doing.
It's nice that there still is someone with a chance to push Microsoft around (a little), while we see countless other articles of large companies caving in to them. Good luck proving the 20-200% improvements that Intel is claiming. Using NAND for a disk cache is interesting, especially if it's uses an intelligent and configurable algorithm. The value is questionnable at the existing 512k-1M sizes, but for notebook computer, what it's made for, I could see a definite "maybe."
However isn't this the same technology used in the new diskless disk drives that, by the specs I've seen, are slower than disk based drives?
Oh crap... I don't know what happened but everything on my web server is gone... And it didn't read my mail :(.
That sounds so, so...eighties.
Or is it memory that can only be used by TurboPascal?
-- Alastair
How can you say sony has a better design than apple with all those defects? I've never had my iBook mysteriously power on, and I've accidentally dropped it before without issue. My iBook on a new battery gets 4-5 hours with wifi enabled. I have about a year old battery getting 3 hours with wifi. I would argue the appearance is better with most Apple products as well. Design is more than the outward appearance of a product. Its also the engineering behind it.
I would guess that both Sony and Microsoft are correct. Microsoft implemented a feature in vista and Sony tried to use it as it was described. We all know that Windows doesn't work as advertised. It never has regardless of service packs.
MidnightBSD: The BSD for Everyone
That did a benchmark of the USB flavor of "ready whatever it is". Their tests showed almost NO benefit of the "feature". I would, as much as I hate to admit it, have to agree with Sony here.
My wife doesn't listen to me either...
Yes, in Linux you could use flash memory in this way ever since flash memory has been around. With Linux there is nothing special about a swap partition; it's a partition like any other, so you could set this up easily enough. Let's say that your USB flash drive is called
Of course buying more RAM would be a better solution, but if you just wanted a faster swap device I would love to see some benchmarks to see if this would do the job.
It is a miracle that curiosity survives formal education. - Einstein
...before any other poor bastard has their mind corrupted with such an horrific image.
I need to purge this filth from my brain. Someone post a goatse link, quick.
Vista currently supports file caching not only on motherboard-loaded Turbo Ram (Intel's solution) but also on USB memory devices.
Sony makes the Memory Stick, which is just about the most expensive and least widely-usable flash-memory solution out there. But, if you plugged it into a USB port on a Vista machine, it could give you a caching boost.
Switching that process to the motherboard is one less advertising opportunity for Sony to foist their proprietary, expensive solutions off on people. Not to mention it lets them stick a thumb in Microsoft's eye for a time just as revenge for bumping them down to #3 in the video game console market.
Stupid idea. Flash has a limited number of write-cycles. Swap would burn through them in no-time flat, even if only in isolated spots.
Unless your flash memory could sustain writes of 20MB/s or more (unlikely).
Linux is very particular about what to swap out. Since it likes to mmap things, it'd much rather kill text/data and file cache first (none of which uses swap). That your combined bss/heap/stack or anonymously mmap pages would use up all your RAM doesn't happen all that often. And then its going to write out the entire process image all at once. It'll demand-page it back in.
Write-outs would be so slow to flash that it'd outweigh the HD seekiness of page-in on resume.
Your linux system really shouldn't be swapping. Ever. That's a last resort data-preservation technique when you run too many things at once or a cron job goes nuts. You shouldn't worry unnecessarily about it and the performance aspects of swap media X or swap media Y.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
Microserf: "Yeah, we almost ran out of Black Suites and brown paper bags that week for all the extra delivery goons we had to deploy"
So, if they aren't in fact aware of any issues, then how would you have them say that?
Advanced users are users too!
Yes..., number three, which is why the PS2 is STILL outselling the 360.
x^1.01?
So there's contention because Sony says a Microsoft feature doesn't work but Microsoft says Sony corporation just can't figure out _how_ to make it work?
Now there's an advertisement for hap-hap-happy Windows user friendliness. Wanna pay $400 to set me up with some Vista _real_ soon with news like that.
Probably because everyone stuck with massive PS libraries and not enough money to afford a PS3 are replacing their broken shoddy POS PS2s. Sony seems to be the master at building products that last until about a month out of warranty then suddenly drop dead.
"well.... it doesn't NOT support it!"
If you mod me down, I will become more powerful than you can imagine....
The bigger question here is:
When is the Linux kernel going to make use of Turbo Memory? It would be funny if, after MS touting this tech for so long, we can say "but you need Linux before it will work properly".
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
What a dilemna: In a fight between Sony and MS, which side do Slashdot users cheer for?
Microsoft said "There are no issues which we are aware of" as opposed to saying "There are NO issues." Henceforth, Microsoft has plausible deniability.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
put the freaking OS on it's own solid-state drive. Half of the time shit loads slow because it has to load subsystem crap from a moving platter. Once it's been stored to RAM, things load faster. OS+subsystems on it's own direct to FSB connected solid-state chip would fix half of this mess as it is.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
MS Win XP already has similar tech to turbo memory. It's the boot and file load optimizer that runs in background every few days to move + coalesce disk reads to all come from the "same area", "sequentially", on disk. Already they measure what areas of disk are used the most during boot and program launch. It would be a simple matter to write the most frequently read areas to "Turbo-memory", _backed_up_ by real disk. The Turbo memory does what it is supposed to do -- speeds up "reads" of frequently used data. Writes may, be done in background or synchronously -- but if they believe the turbo memory is reliable enough, they could write to it synchronous, then do the actual disk write in background, asynchronously.
If the memory "fails" with material on it, no prob. It's also been written to disk, so recovery is seemless though perhaps slightly slower.
Anything MS would put in as functionality over that, would be gravy -- but MS has to do very little to support turbo memory -- either what area of disk to "Turbo-boost" (gets shadowed, for Read purposes, by flash) or
Tells the disk what areas need to be "boosted" and let the disk take care of the caching.
In this type of incarnation, it might be nearly equivalent to current disks' cache, but the new Flash-cache will likely have Gigabytes of space vs. todays volatile RAM having megabytes.
I think Sony is likely to lose this one, as it is "theoretically" able to be implemented totally on the disk. Why wouldn't the disk, itself, monitor which sectors are used most often and cache them for reads? It's probably configurable whether or not to turn on write-caching, which might allow writing to Flash and not wait for the copy to reach disk, or force a wait for reliability.
As for failing FLASH -- that could be detected on write -- and could be detected by the disk drive.
Could someone perhaps explain to me why Sony is weighing in on what seems to me to be a private matter between microsoft and intel? This just seems to me to be analogous to mac suing sony because the ps3 doesn't support all hdtv's [or didn't, i don't know fixed that problem or not]
Coming to you live from another dimension.
I know this is supposed to be a joke, but...
a) Who reads mail as root?
b) Who cd's out of their home directory before running a mail command?
c) Who has a webserver with a mail client on it?
d) Who has successfully gotten apache working without knowing basic UNIX commands, like rm?
Ok, I'm laughing, hahaha, very, very funny...
Sam
Speak for yourself, I love my PS2 thank you very much. Years in and they still support it well, focusing on it instead of the PS3. Why? Because it still makes money.
In the years I've owned mine the only thing I had to do was clean off the laser, it got dusty and wouldn't play.
It's not the end all, but it's not a POS in my opinion.