How to speed up Windows

[dude_56013]

also, XP uses 1.5GB of my small boot drive, Win7 and Vista use 15GB-30GB.

Keep in mind that the average size of a hard drive was much smaller when XP was "around" so-to-speak. You were talking 20-80GB IDE drives for the most part, some intermediate level drives with 120GB or more for some, but not often. Nowadays, the average hard drive size is much larger--'cept for SSDs which are a gimmick IMHO--so the O/S can take up much more space without people caring too much.

 

[Aastii]

also, XP uses 1.5GB of my small boot drive, Win7 and Vista use 15GB-30GB.

As I said earlier, and as dude summed out quite well, times change. You can get TBs of space for the same price as you could have 100GB back then, everything was smaller back in the day and everything was much slower

Keep in mind that the average size of a hard drive was much smaller when XP was "around" so-to-speak. You were talking 20-80GB IDE drives for the most part, some intermediate level drives with 120GB or more for some, but not often. Nowadays, the average hard drive size is much larger--'cept for SSDs which are a gimmick IMHO--so the O/S can take up much more space without people caring too much.

SSD's aren't a gimmick, they are easily the next big thing. HDD's are getting towards the end of their prime, they are good for storage, but for performance, SSD's get rid of the bottleneck, or rather make it much less. Secondary storage is the biggest bottleneck in every system and hard drives can't deal with that. As flash memory becomes cheaper to produce and lasts for more read/write cycles, SSD's will be the norm, and eventually hard drives will be seen as tape is now

 

[Remeniz]

SSD's aren't a gimmick, they are easily the next big thing. HDD's are getting towards the end of their prime, they are good for storage, but for performance, SSD's get rid of the bottleneck, or rather make it much less. Secondary storage is the biggest bottleneck in every system and hard drives can't deal with that. As flash memory becomes cheaper to produce and lasts for more read/write cycles, SSD's will be the norm, and eventually hard drives will be seen as tape is now

They probably will supersede plattered hard drives, eventually. But i'm not sure we'll see them in data critical servers any time soon...

 

[dude_56013]

As I said earlier, and as dude summed out quite well, times change. You can get TBs of space for the same price as you could have 100GB back then, everything was smaller back in the day and everything was much slower

And yet, at the time, that was the "norm". People didn't say GOSH DANG THIS IS SLOW! It was "ok". I guess it's kind of paradoxical in nature, as it's hard to imagine that from times now. Sure you can look back now and say that was ridiculously slow, but it's hard to predict how things will be in the future. If that makes sense.

SSD's aren't a gimmick, they are easily the next big thing.

With their current price/performance ratio, unless you can justify the cost from a professional standpoint, I'd call them a complete gimmick, just like every other new product to enter the market at a price point far above the average consumer's pocket book.

HDD's are getting towards the end of their prime, they are good for storage, but for performance, SSD's get rid of the bottleneck, or rather make it much less.

I disagree. HDDs are not at the end of their prime. In the next few years, they will most likely be used heavily as storage/slave disks. You just can't beat the price/performance ratio. For as cheap as they are in comparison, you could RAID0 some and come out just about equal to SSDs. In terms of enterprise solutions, you won't see SSDs take over for quite a long time, yet.

Secondary storage is the biggest bottleneck in every system and hard drives can't deal with that.

1) I don't understand what you are getting at by saying "secondary" storage. What is your primary that you are referring to?

2) I disagree to some extent. There will always be a bottleneck, and for many systems it is the RAM and/or GPU, not just the hard drive.

As flash memory becomes cheaper to produce and lasts for more read/write cycles, SSD's will be the norm, and eventually hard drives will be seen as tape is now

You know tape drives are still very prominent in some companies, right?

They probably will supersede plattered hard drives, eventually. But i'm not sure we'll see them in data critical servers any time soon...

Bingo.

 

[berry120]

If traditional hard drives are dead, why are we still seeing announcements like these?

Hitachi's New Hard Drive Paves the Way for 4TB and 5TB HDDs - X-bit labs

If the companies had confined them to the history books then there wouldn't be any research like this still going on. SSDs probably will catch up and take over as the popular choice eventually, but not yet.

 

[Remeniz]

+1 to the two previous posts.

The SSD is a brilliantly great break-through but seriously, the plattered hard drive will get even faster before the SSD catches up.

SSD cannot beat platterd HD reliability and continuously sustained throughout.

 

[Aastii]

With their current price/performance ratio, unless you can justify the cost from a professional standpoint, I'd call them a complete gimmick, just like every other new product to enter the market at a price point far above the average consumer's pocket book.


I disagree. HDDs are not at the end of their prime. In the next few years, they will most likely be used heavily as storage/slave disks. You just can't beat the price/performance ratio. For as cheap as they are in comparison, you could RAID0 some and come out just about equal to SSDs. In terms of enterprise solutions, you won't see SSDs take over for quite a long time, yet.


1) I don't understand what you are getting at by saying "secondary" storage. What is your primary that you are referring to?

2) I disagree to some extent. There will always be a bottleneck, and for many systems it is the RAM and/or GPU, not just the hard drive.


You know tape drives are still very prominent in some companies, right?


Bingo.

First point: You are absolutely correct that at the moment they are above the price point, but as another example, look at prices of a GTX260 when they first came out. Now, look at the prices of a HD5770 when it first came out. Different manufacturers, yes, same performance, yes, different prices, yes. Technology is always more expensive when it first comes out, SSDs will, and are, lowering in price, we are getting very, very close to the $2/GB mark now, which compared to when they first came out is nothing at all. At the moment they are for a nieche market, in a couple years time, many more will be able to afford them, and will be able to afford the newer, faster models. Production of NAND memory will become easier with new developments in construction, and they will become more reliabble, lowering prices, and making them a more viable option.

Second point: Allow me to rephrase: HDD's are coming towards the end of their prime as main drives. With in the next couple of years, SSD's will be the main with hard drives just as storage, because you are right, they are unrivalled at the moment for price/storage ratio, and a few years after that, storage on SSD's will be more than common in systems

Third Point: 1. Secondary storage is hard drives, SSD's, memory sticks, CD's/DVD's etc. Primary storage is system memory, RAM, cache etc. It is storage that isn't accessible directly by the CPU

2. I didn't say hard drives are the only bottleneck, I said they are the largest bottleneck. If you have SSD's, your system boots faster, everything loads faster, everything saves faster because the SSD, although still slower than RAM and the CPU, is a hell of a lot quicker than a hard drive. Although they are expensive, the performance gain that you see with SSD's supersedes that of spending the money on a new CPU or more memory (generally, obviously if you are on an ancient system, money would be better spent elsewhere, but on a relatively new system, Core 2 for instance, SSD is a better upgrade than going to i5/i7 CPU).

Final Point: People are still using Commodore 64's, but in the big scheme of things, it is a tiny number. The vast majority are using hard drives because of the advantages over tape. When SSD's, or whatever the new thing in storage is, a small number of people will still be using hard drives, but then you will be saying when I say HDD's are now dead "you are aware some companies still use hard drives, right?" So what, most use solid state storage

+1 to the two previous posts.

The SSD is a brilliantly great break-through but seriously, the plattered hard drive will get even faster before the SSD catches up.

SSD cannot beat platterd HD reliability and continuously sustained throughout.

And that makes SSD's not the next big thing and not a gimmick how? You've pretty much just said that hard drives aren't as good as SSD's, but they won't catch on just yet because hard drives are still the "better" option and still have a little bit to improve. I agree that they are going to improve, but without increasing rotation speed, you aren't going to get much more out of them, and when you increase rotation speed, you increase power usage (making them useless for the mobile market), you reduce reliability and you increase heat/sound output.

In that final point is another reason that SSD's will take off - power consumption. They use less power than hard drives do making them a much better option for mobile devices. More and more people are choosing smart phones, laptops, netbooks and hand held devices over desktops, and because power consumption is a massive factor in the mobile market, those that produce mobile devices will be eager to get bigger and better solid state drives to improve their products and use less power.

 

[berry120]

First point: You are absolutely correct that at the moment they are above the price point, but as another example, look at prices of a GTX260 when they first came out. Now, look at the prices of a HD5770 when it first came out. Different manufacturers, yes, same performance, yes, different prices, yes. Technology is always more expensive when it first comes out, SSDs will, and are, lowering in price, we are getting very, very close to the $2/GB mark now, which compared to when they first came out is nothing at all. At the moment they are for a nieche market, in a couple years time, many more will be able to afford them, and will be able to afford the newer, faster models. Production of NAND memory will become easier with new developments in construction, and they will become more reliabble, lowering prices, and making them a more viable option.

I don't think people are arguing with you that they will eventually replace traditional hard drives as the norm. It's the timescale we're talking about, I personally think it's going to be longer than 2 years before that happens. But hey, I could be wrong - no-one really knows!

Second point: Allow me to rephrase: HDD's are coming towards the end of their prime as main drives. With in the next couple of years, SSD's will be the main with hard drives just as storage, because you are right, they are unrivalled at the moment for price/storage ratio, and a few years after that, storage on SSD's will be more than common in systems

Two points here - very few consumer based systems have two hard drives in at present, and I doubt that will change. When SSDs make the mainstream, they're going to have to replace conventional hard drives, not work alongside them. Secondly, I just can't see them rivalling conventional hard drives on price / storage ratio any time within the next 5 years or so, there's just too big a leap for a technology that's already been out for quite a whlie (if SSDs were starting off at these prices I'd say differently, but they've already come down somewhat.) With a conventional hard drive I can get a decent 1TB drive for <£50 without any difficulty at all. Even the lowest capacity, slowest SSDs cost upwards of that and they're 32GB. That's over 30 times more expensive - if we were talking about double the price, again, I'd agree that in the next couple of years SSDs would pinch it. If I could get a 1TB SSD for £100 then that's a whole different board game!

2. I didn't say hard drives are the only bottleneck, I said they are the largest bottleneck. If you have SSD's, your system boots faster, everything loads faster, everything saves faster because the SSD, although still slower than RAM and the CPU, is a hell of a lot quicker than a hard drive. Although they are expensive, the performance gain that you see with SSD's supersedes that of spending the money on a new CPU or more memory (generally, obviously if you are on an ancient system, money would be better spent elsewhere, but on a relatively new system, Core 2 for instance, SSD is a better upgrade than going to i5/i7 CPU).

I agree that for some users they are the bottleneck, but you can't really accurately say it's the largest bottleneck - it just depends what you're doing. If I'm shifting stuff from a fileserver over gigabit ethernet with a decently powerful machine then yes, they probably are the bottleneck. If I'm ripping / encoding video streams on the fly then almost definitely not.

I agree that they are going to improve, but without increasing rotation speed, you aren't going to get much more out of them

Sorry, but this one just isn't correct at all - there's a variety of ways that manufacturers can (and have) increased hard drive speeds and capacities dramatically over the years without increasing the rotation speed.

In that final point is another reason that SSD's will take off - power consumption. They use less power than hard drives do making them a much better option for mobile devices. More and more people are choosing smart phones, laptops, netbooks and hand held devices over desktops, and because power consumption is a massive factor in the mobile market, those that produce mobile devices will be eager to get bigger and better solid state drives to improve their products and use less power.

In the mobile market I actually agree with you for this reason - for netbooks and the like they make perfect sense. But where I disagree is that this will directly translate to the desktops. People who own desktops these days generally have a reason for doing so, and this reason is often at least partly for storage capacity. A 40GB drive might be fine in a netbook, but on a desktop I wouldn't put up with anything less than 1TB these days. I'm also not sure their success in the mobile market will drive hugely larger capacities, simply because larger capacities aren't really called for in the mobile market (honestly, if you're given a 1TB drive in a netbook or your phone what on earth would you put on there?)

 

[Aastii]

I don't think people are arguing with you that they will eventually replace traditional hard drives as the norm. It's the timescale we're talking about, I personally think it's going to be longer than 2 years before that happens. But hey, I could be wrong - no-one really knows!

If you had the choice between a 250GB SSD and a 500GB HDD for the same price, which would you take? For the extra performance, I would sooner take the SSD. It is half the capacity, but speed wise, it is worth losing it for in my opinion. You are losing storage, but gaining a heck of a lot extra performance, and much more stable performance. SSD's don't have to be the same price as hard drives before they become the main, they can be double the price and people will get them, even if it is the slower ones, because OEM companies will say "It has an SSD, much faster than hard drive!!" and computer illiterate people will lap it up.

Two points here - very few consumer based systems have two hard drives in at present, and I doubt that will change. When SSDs make the mainstream, they're going to have to replace conventional hard drives, not work alongside them. Secondly, I just can't see them rivalling conventional hard drives on price / storage ratio any time within the next 5 years or so, there's just too big a leap for a technology that's already been out for quite a whlie (if SSDs were starting off at these prices I'd say differently, but they've already come down somewhat.) With a conventional hard drive I can get a decent 1TB drive for <£50 without any difficulty at all. Even the lowest capacity, slowest SSDs cost upwards of that and they're 32GB. That's over 30 times more expensive - if we were talking about double the price, again, I'd agree that in the next couple of years SSDs would pinch it. If I could get a 1TB SSD for £100 then that's a whole different board game!

Very few consumers have 2 or more internal hard drives, but a lot of consumers have an internal drive and an external storage drive in one form or another, be it an external hard drive or memory stick. It doesn't have to be internal for it to be a conventional hard drive. As I say, a lot of people already have those alternative storage media, and so moving it over from one computer to another isn't an issue, they would carry on using the storage drive as they always have done and just see the computer as "faster, even if it does have less storage".

I agree that for some users they are the bottleneck, but you can't really accurately say it's the largest bottleneck - it just depends what you're doing. If I'm shifting stuff from a fileserver over gigabit ethernet with a decently powerful machine then yes, they probably are the bottleneck. If I'm ripping / encoding video streams on the fly then almost definitely not.

Hard drives are the biggest bottleneck, regardless of the use. In actual use once things are loaded, it may be that your video card or your CPU is pinned at 100% and so is the bottleneck, but the largest bottleneck is secondary storage. To explain:

Assume you have a fairly average system:

Athlon II x2
4GB DDR3 1333
WD Caviar Blue 250GB (WD2500AAKS)

Athlon II has 33.1GB/s bandwidth
The memory has 21.2GB/s
Your hard drive has 300Mb/s

These are all theoretical numbers, but say they are only 12.5% accurate (1 8th of the number according to those figures), you are still talking about a hard drive being over a million times slower than primary storage. Samsung put 24 SSD's in RAID0 to get 2GB/s throughput and the difference is ridiculous. It is a retarded video to show it, but the tests show what happens when you unlock more potential from the memory and CPU:

Sorry, but this one just isn't correct at all - there's a variety of ways that manufacturers can (and have) increased hard drive speeds and capacities dramatically over the years without increasing the rotation speed.

I understand you can change other parts, tweak parts, add parts etc, however the largest performance gain you see is by making the platters spin faster. Look at the biggest performance gain seen in the last few years from hard drives - the Seagate hybrid drives (momentus XT) which added solid state memory along side the hard drive. the only drives that come close to it are the 10-15k drives. You can squeeze out extra performance, and one day maybe get SSD performance (modern day SSD) out of a 7200RPM or less drive, but by then solid state drives will have left hard drives in the dust. There is nothing, at least in the forseeable future that will improve the performance of hard drives to the point that they will hold off solid state storage for any length of time.

In the mobile market I actually agree with you for this reason - for netbooks and the like they make perfect sense. But where I disagree is that this will directly translate to the desktops. People who own desktops these days generally have a reason for doing so, and this reason is often at least partly for storage capacity. A 40GB drive might be fine in a netbook, but on a desktop I wouldn't put up with anything less than 1TB these days. I'm also not sure their success in the mobile market will drive hugely larger capacities, simply because larger capacities aren't really called for in the mobile market (honestly, if you're given a 1TB drive in a netbook or your phone what on earth would you put on there?)

The potential in mobile devices is probably greater than in the desktop industry. It will indirectly affect the desktop sector. As innovations in design and production are made for the the mobile market, these will be implemented into the desktop sector. Also, because they will be used more, initially, in the mobile market, desktop solid state technology will get money for R&D from the mobile device sales.

Think of it like AMD and Nvidia get money from people buying their very low end graphics cards and integrated graphics which then goes into the R&D of their higher end cards, people with the other needs (in this case the need for less power consumption in mobile devices) fund the R&D of the niche high end market, improving performance and lowering the prices

 

[berry120]

If you had the choice between a 250GB SSD and a 500GB HDD for the same price, which would you take?

Yes, but you're talking about double the price / storage ratio which, as of yet, is miles off. As I said in my previous post:

With a conventional hard drive I can get a decent 1TB drive for <£50 without any difficulty at all. Even the lowest capacity, slowest SSDs cost upwards of that and they're 32GB. That's over 30 times more expensive - if we were talking about double the price, again, I'd agree that in the next couple of years SSDs would pinch it. If I could get a 1TB SSD for £100 then that's a whole different board game!

Very few consumers have 2 or more internal hard drives, but a lot of consumers have an internal drive and an external storage drive in one form or another, be it an external hard drive or memory stick. It doesn't have to be internal for it to be a conventional hard drive. As I say, a lot of people already have those alternative storage media, and so moving it over from one computer to another isn't an issue, they would carry on using the storage drive as they always have done and just see the computer as "faster, even if it does have less storage".

You can't equate external hard drives and memory sticks to internal drives - with memory sticks especially people use them as a convenient method of data transfer, not a larger backup in case the smaller drive goes down! Besides, do you really think it's a good idea to entrust all your files to an external hard drive that might be bashed around or a memory stick that might be lost or broken? And sorry, but there's a reason large drives are often provided with PCs - people need the storage capacity! True this isn't always the case, but if anything it's becoming more the case as fewer people that just need the computer for a few word documents and email are migrating to netbooks and laptops. Generally speaking, people that see the need to buy a desktop these days won't be able to put up with a 40GB drive. I know I sure wouldn't, I've just got way too much stuff to fit on one!

Hard drives are the biggest bottleneck, regardless of the use. In actual use once things are loaded, it may be that your video card or your CPU is pinned at 100% and so is the bottleneck, but the largest bottleneck is secondary storage. To explain:

Assume you have a fairly average system:

Athlon II x2
4GB DDR3 1333
WD Caviar Blue 250GB (WD2500AAKS)

Athlon II has 33.1GB/s bandwidth
The memory has 21.2GB/s
Your hard drive has 300Mb/s

These are all theoretical numbers, but say they are only 12.5% accurate (1 8th of the number according to those figures), you are still talking about a hard drive being over a million times slower than primary storage. Samsung put 24 SSD's in RAID0 to get 2GB/s throughput and the difference is ridiculous.

Sorry, but your understanding of a bottleneck is just plain wrong, it's not just the slowest component in the machine. Yes, in loading applications up and copying files around, as that video shows, the hard drive is the bottleneck. But how much time using your PC do you spend doing the above tasks? Comparatively not much at all. When you're playing games for instance, the hard drive is barely in use and definitely not the bottleneck, it'll most likely be either the graphics card or CPU. Otherwise we'd all be swapping out our hard drives with SSDs and seeing frame rates shoot up by a factor of 10!

Even by your own definition it doesn't stand that the hard drive is the bottleneck - it'd be the DVD drive which on average will read data at around 20MB/s. So should we all be investing in super fast DVD drives to make our computers run faster?

I understand you can change other parts, tweak parts, add parts etc, however the largest performance gain you see is by making the platters spin faster. Look at the biggest performance gain seen in the last few years from hard drives - the Seagate hybrid drives (momentus XT) which added solid state memory along side the hard drive. the only drives that come close to it are the 10-15k drives. You can squeeze out extra performance, and one day maybe get SSD performance (modern day SSD) out of a 7200RPM or less drive, but by then solid state drives will have left hard drives in the dust. There is nothing, at least in the forseeable future that will improve the performance of hard drives to the point that they will hold off solid state storage for any length of time.

You're arguing a bit of a moot point here, of course conventional hard drives will never have the speed of SSDs, I never stated they would. Your statement that they wouldn't get "much more out of them" without increasing the rotation speed was my point, no they'll never be lightning fast but they can still increase relative speed without ramping the rotation up to 50k rpm!
Oh and those hybrid SSDs are actually nowhere near as fast as they claim to be in real world use, a number of real world benchmarks have shown the conventional, faster spinning drives are the better option.

The potential in mobile devices is probably greater than in the desktop industry. It will indirectly affect the desktop sector. As innovations in design and production are made for the the mobile market, these will be implemented into the desktop sector. Also, because they will be used more, initially, in the mobile market, desktop solid state technology will get money for R&D from the mobile device sales.

It will fuel it somewhat, but perhaps not as much as you think. If the money's in the mobile device sector, then that's where the R&D will be focused (what's the point of taking money from that and focusing it on a less compelling market?) So it really all hinges on whether the advances made in the mobile sector are relevant to desktop use. If the advances are in capacity, sure. And capacity / price ratio will slowly and surely increase, I've no doubt about that. But I'm not sure it'll actually advance that quickly since I'd imagine more of the research would be put into performance increases, reduced power consumption and improved manufacturing efficiencies (these things being more critical to the mobile market.)