Re: Why are PC sales declining ? (Skybuck thoughts on it too)

ndered

Great, Thanks for all that Martin.

George H.

reliable,

under

Oracle

In the '80s Goodwill Industries was getting bids by steel scrap haulers on donated mainframe racks... hoards of 'em. Literally tons of big computer grade caps and fans and other stuff could have been salvaged.

That was way back then.

Muy estupido.

cache? not.

I used "speeddisk" back when software based caching actually had benefit (286/386 days). It does not with today's drives and I/O buses.

I used a whole, considered huge at the time 16MB hardware based caching hard drive controller back when *that* had benefit.

It no longer does or there would be cards still being made.

Even big cloud racks full of hard drives use the hard drives themselves to get max throughput and there is no caching.

My cheap ACER laptop with a Seagate SATA 6GB/s 750GB HD+ 8GB SSD boots in 15 seconds flat to Win7.

Oh and what THIRD PARTY application are you using to 'create' a 'RAM_Disk' with?

As always, wrong!

Disk cache is built into modern OS kernels. The OS keeps track of what you've accessed from, or recent written to the disk in the otherwise unused memory.

--- news://freenews.netfront.net/ - complaints: snipped-for-privacy@netfront.net ---

You are an idiot. you proved it yourself. Time and time again.

You are an idiot... proven with your retarded posts... Time and time again.

You didn't say anything, dumbass.

And it is the last thing a dumbass like you needs to be trying to get installed. and working.

Perhaps? Perhaps you are a clueless twit who should not be attempting to guide anyone's PC build endeavors.

And we need you to deliver these common sense bits of info? No. We can see perfectly fine when we choose our video cards. We can weigh all the factors just fine. We do not need an utter idiot, who can't even keep a single machine running, trying to tell folks about computer parts. One repost of one of your stupid rants where you blame things on stupid reasons which were not even true and folks would see that your brain is no more than eleven years of age in term of maturity, and your knowledge is not much more advanced than that.

And that never happens, right?

Really? can't just buy a bunch of parts that will not work together and mash them together anyway?

Do you ever say anything that isn't already an obvious requisite?

Only for dorks like you, where Mommy is signing the check.

And locking it in a cage so YOU can't lay your hands on it and break it, like all the others you touched.

There are plenty of games that fully tax the PC.

Before or after you touched it?

Before or after you touch them?

So, do you act like you are guessing your way through life, or are you really actually guessing your way through life?

You're an idiot. A complete and utter stooge.

The only thing you have ever said that even comes close to being right.

For ALL computer use, a graphics card is essential, idiot. For gaming a "fairly good" graphics card is needed.

Not with you and all your bent perceptions at the helm.

You are an idiot. you proved it yourself. Time and time again.

C'mon, Decadent. Don't hide your feelings. Tell us what you really think of Skybuck.

I recently bought a new laptop and promptly UPgraded it from Win8 to Win7.

Did you pay the Micro$oft tax twice !?!?

In another thread someone was complaining that MS doesn't do enough to promote safety and stability in Windows. This is one of the things they are doing and no one likes it. I know since XP they have told developers how to separate programs and data, but no one bothered to deal with it. Since Vista the default settings have been to enforce these restrictions and not let programs access data in program directories... resulting in failed installations. Even when this is pointed out to many program developers, they choose to blame the OS rather than recognizing the true issue and going with the flow by changing the way their programs work.

Using that rule of thumb you won't buy a new machine for the next 20 years! lol I don't think PCs are actually getting much faster these days are they? The clock speeds stopped increasing some 12 years ago and it seems like the only thing they can do now is increase the amount of on-chip memory.

I wouldn't say there is no need for speed, they just can't deliver any more. I do need to replace my 4 year old laptop though because it only has 3 GB and that is just not enough anymore. I'm hoping the price of flash disks comes down a bit more and I will extend the life of this box another couple of years by compensating for the lack of memory with faster disk access.

I saw a mention of some sort of combined disk. I expect it was a rotating disk with a flash component. That would be very interesting if it fits a 2.5 inch drive size.

Hybrid hard drives - Such as:

It is about every six years now to get a threefold increase. I generally buy for the best price performance at the time. Clock speed used to be a simple metric from the original 4.7MHz right up to some now on offer brutally overclocked to 4.8GHz that is a 1000x increase since 1981 or roughly speaking a 25% average improvement annually. Recent improvements have been largely in CPU utilisation, pipelining and speculative execution rather than raw clockspeed.

My old Q6600 benchmarks at 2962 and uses ~350W with graphics card whereas the new i7-3770K benchmarks at 9461 and about ~120W all in.

Actual benchmark speeds are still increasing provided that you have the right software and can use fully multicore and multithreaded code. The problem is that after about 6 CPUs the law of diminishing returns sets in and the code spends an increasing amount of its time sharing the load between threads or worse still doing work in parallel that will later be scrapped when the independent thread results are combined.

These days the biggest performance increase can be had by putting frequently used files onto an SSD with essentially zero seek time and a transfer speed that maxes out SATA3. The Samsung 830 & now 840 drives are very impressive - beware that some benchmarks give artificially high performance figures of merit on highly compressible data.

The SSD upgrade is capable of giving old kit a new lease of life.

You can get the 256GB Samsung 840 for around £180 and the cheaper consumer grade 250G one for ~£130. Provided that you keep it backed up just in case the storage medium fails then what are you waiting for? The prices may fall a bit more but the technology is available and the price is not outrageous for the performance boost it offers.

Be sure to disable MS lame indexing and never defrag it!

My desktop has the Intel SSD cache installed. It does make the thing a lot more nippy. It also has a dedicated Samsung 830 SSD for bulk high speed data which makes a big difference for random access lookups.

Indeed, also in Win7 they withdrew certain legacy 16bit components that a well known brand of software installer still relied upon.

The problem is if you have old and expensive software that you like then you don't have a lot of choice if their development team decided to stuff some config files in the program directory. The registry isn't much better and always reminds me of Zork whenever I have to traverse its maze of little passages all alike with my flint axe looking for whatever installation screwup has just occurred.

I can think of several examples where an old version of the software has the right features and works correctly and a new improved modern bloatware version is seriously defective - but looks a lot more sexy.

Same way you find some programs that insist on having God like privileges to work at all. My biggest pet hate at the moment is the number of major manufacturers issuing driver updates and installers that are not digitally signed so that users get used to pressing the ignore security concerns about this update button routinely.

And I am talking about PC and peripheral major players here! (My sword has begun to glow blue)

I can't say I've run any tests, but I don't see how they are getting more processing other than adding to the cache sizes. Pipelining and speculative execution should have been mature some 10 years ago. What exactly is left to improve on?

That assumption is a big one! The study I read said the turn was less than 4 CPUs. Many apps just won't see much improvement with even two processors. The observed increase in performance is because the OS needs elbow room, so a second processor helps get it out of the way of the user app.

I'm taking a look at the combined drives now. I'm not going to pay an arm and a leg for one. I can get a SSD for under $200 that is bigger than what I have now. A combined drive should be close to $100 I am thinking.

I'm not sure I want to go that route. I need to weigh my options. Also, I have some work to do for the next month or so and can't have any down time. After that I'll look at this harder.

I get that, but developers have had since XP and possible Win2000 to learn how to do it *right*. MS has been telling them what to do, they just won't listen.

So who ya gonna call? Well, I don't think Ghostbusters will be of much help, but no point in blaming MS for the entire industry's problems.

then

Worse, the fundamental structure of an 8086 promoted proper data and code separation back in the DOS 3.0 days. For that matter most of the compilers did as well.

that

Win7 represents an advance on this front. An ordinary user can install programs for personal use to their own directory tree. Not many programs are ready for this but some are (e.g. Amarok), and am i glad.

Win7.

No only one and a half times. Not much better, but better.

?-)

generally

improvements

Yes and no. It has taken the past 20 years to move it from multi board level to chip level.

the

load

be

Twenty years ago it was hard into the performance limit curve. But that was multi board level processors. The delay limits are relatively smaller on chip and 8 way multi core is not as far around the diminishing returns knee. Still memory bandwidth (even with today's huge multilevel caches) is the fundamental limiting factor.

high

up

The

price

Sounds like a good plan. Also a reason i have multiple machines, i can slog stuff between them mostly. Not everything can be moved though.

?-)

I had to think about that since I largely agree. Although register colouring and other tricks are comparatively recent as a refinements that keep the execution pipeline from stalling so easily.

Taking CPUs I have most experience of tormenting:

Back of the envelope calculations suggest that the most recent benchmark improvements have come from on demand turbo boost - that is the difference between 3770 & 2700 and 3570 & 2500 can be largely explained by the 10% faster clock when asked to work really hard.

And even my old Q6600 with a toasty 105W TDP if it could be scaled to the same hardware spec as the new CPU would improve by the ratio of the clock speeds and a factor of two for hyperthreading (optimistic).

Suspiciously close agreement! So arguably they are gaming the standard benchmarks now to make new chips look more attractive. Underlying performance is rather similar except when heavily loaded by optimally designed parallel algorithms designed to use all the cores at once.

It depends on the application. Some things benefit whereas others don't. Hyatt did a lot of work on optimising chess search algorithms on multiple processors (and that is a tricky algorithm to parallelise).

N_CPUS 1 2 4 8 16 Naive 1 1.8 3 4.1 4.6 EVP 1 1.9 3.4 5.4 6 DTS 1 2 3.7 6.6 11.1

Taken from his paper

Naive is typical of what happens if you try and parallelise without thinking very carefully about the bottlenecks and DTS is more typical of what you get with a streamlined optimised multiprocessor algorithm.

Most decent multicore code is somewhere between these two extremes ~ EVP where you do OK on up to 4 cores.

I found the Samsung drives performed better on incompressible data which was important to me.