Posts by Bronek Kozicki 2859 publicly visible posts • joined 6 Sep 2007
not because I would expect new architecture to perform badly, but rather because it usually takes a while to iron out programming with new paradigms and in new ways. If this is as revolutionary as claimed, it will be a while before driver writers learn to use the new instruction set properly.
One other thing I'd like to know - how will this compete with CUDA.
does Safari implement CSS shaders or WebGL? If it does not and if there is indeed no kernel level software installed with Safari, then indeed it would be Windows bug. Hope to be fixed soon.
However, if WebGL or CSS shaders are involved, that would mean graphic card drivers is involved as well.
Lets wait until more of this is known.
... put few dozens of crapware programs on top of Windows.
One laptop I would consider buying would have to be free of crap and come with great LCD (IPS or similar). SSD and memory I can put myself, since markups commanded by vendors for these small user-replaceable bits are ridiculous anyway.
I'm none of the above and I (almost) love my BlackBerry. And I'm looking forward to sensible alternative tablet OS based on strong realtime OS. Brand name? Of course anyone can call it BlackBerry 10 but I'm sure commonly used shortcut will be B followed by B followed by 10 - or perhaps roman X.
Silly mistake indeed, but I sincerly hope that's only because they are focused on other things - like delivering great OS.
Simples, really. Agency model harms overall competition - between agents (no incentive for a shop to compete against Amazon; no incentive for a publisher to sign more agents than minimum necessary), between titles (you buy what publishers want to sell you, not what market priced as a good title at good price) and between publishers (if they set the price, who is to say they don't agree on it first?).
I lived in centrally planned economy and I know from experience it does not work. As you browse through Kindle section on Amazon it should dawn on you, too.
and as such you actually don't see it often, outside of sales channel and spec sheets.
However the train I'm commuting on everyday is all plastered with ads of Toshiba Z830. The point is that vendors matter, models matter, specific selling points that customers are willing to pay for ("thin" "light" "fast") matter.
Some may say "meh" since remote access might imply long latencies, but this is InfiniBand. Its latencies are only little more than 1 microsecond, although I'm curious about other overheads in such system employing rCUDA.
For comparison, single context switch in Linux kernel on modern Intel x64 is in the range of 2-3 microseconds, so we are indeed talking low latency access to remote computing resource here. This makes CUDA practical for whole range of new purposes.
I know this is silly question, but isn't IA64 architecture more sane and (potentially) more efficient than IA32?
If so it should be possible to manufacture "consumer" version of IA64 with smaller cache and fewer cores (and smaller power envelope). Well this is big "if" and presumably not before Paulson is out of the door. I'd like see Ellison's face then ;)
CFLT turns spikes of UV light into spikes of light in visible range. There is attempt to spread the visible output but that's it - the result is still very uneven. You might be right but if there are no UV spikes to begin with, the output could be more even; although I would not expect this to be as good as halogen with appropriate coating applied (e.g. Solux)
you are correct on this. You want SSD with good garbage collection, I think majority have it (certainly Crucial and Intel do).
Anyway if you marry SSD with RAID, your contention is going to be RAID bandwidth anyway (unless you set the limit to few drives only). Modern SSDs are really much faster than HDDs and RAID cards can "feel" it. Also only few RAID cards support SSD, e.g. newer LSI.
since around the time support for TRIM started appearing in major OSes, SSD manufactures started implementing alternative solution in firmware, i.e. garbage collection. As of today it's mature and reasonably well working technology on systems where TRIM is still lacking (ie. older OSes). Garbage collection is also implicitly used by new systems when "something" stands in a way between actual SSD and an OS, e.g. RAID , and it works well.
In principle you are right, but the problem is timing. Yields tend to go up in minutes, days maybe. And then they stay up for weeks or months with no force to push them down, if there is no lender of last resort. There is positive feedback in one direction only. In practice this turns "little" problem of illiquidity into big problem of insolvency, and the latter tends to go viral.
I think you would agree that this is a big problem. Yes nations have to work hard, pay taxes etc. BTW one cannot fail to notice that Slovaks where *average* wages are *lower* than minimum wages in Greece, all in same currency, are also contributing to bail Greeks out. The trouble with this is that systematic changes are needed to increase tax income, cut public sector, cut dole etc. . This takes years (months at best), which is just enough time to push not just one nation, but everyone into insolvency.
All due to lack of balancing force (lender of last resort) on the bond market. I really think it's time to put ideology aside, no matter how righteous it is.
El Reg please, focus on something worthwhile. Making news out of the fact that Nokia apps can be hacked is like telling everyone "DRM does not work". We all know it!
And do you really think this is going to make a dent in Nokias revenues? Is your reasoning along the lines "oiks aren't going to buy these new ugly Nokias anymore because they can have ugly Samsung or HTC instead" ? From what I've heard these new Nokias are apparently best looking Windows phones around and I very much doubt general population has the skills and determination to install hacked apps on 3rd party handsets. So no, this is not going to cost Nokia a penny.
More likely add to bottom line, thanks to nice free advertising. Apparently someone likes Nokias apps enough to hack them!
... since we just started to learn what breach of trust can do to PKI, I think I will wait until good patterns for handling system failure emerge. Give it 15 years, maybe 50 (or more ?)
on the other hand, if identity systems NOT linked with private protected information were to gain gov. support , that /could/ be nice. Something to help me to link Google+ , facebook , various forums etc. together, but /away/ from my bank.
reminded me why I chose "long in the tooth" blackberry, although one with physical keyboard. I'm email addict. BB integration with email is indeed very good - I'm using both gmail and generic IMAP and they work flawlessly, although initial setup to filter only stuff I'm interested in can be tricky. There is also additional "push" (delivery time measured in seconds) account with phone contract.
here : "In China, India and other emerging economies, patents don't hold back progress"
it's only our western habit of navel gazing which prevents us from seeing the obvious. Few more years of this patent nonsense coupled with weakened retail markets (did anyone say "recession"?) to offset litigation costs, and I can see plethora of less-known (in the West) companies doing some serious export here.
apples and oranges. You are comparing price of generic HDD against Fusion IO which is highly specialized and expensive piece of kit. Look down to Kingston V100 and likes, you may find something which is actually affordable if not as main storage, then at least as a fast boot drive.
well it's not exactly load of bollocks, you made maybe one good point (that Android and Windows were not so mature as iOS - it is arguable whether they still are). You also made more than few bad points, let me comment on just this one:
"Yes, I am willing to spend money to be more individual of course"
well of course you want to be more individual, meaning you buy same thing 75% of others do. Or did you forget that Apple has 75% of table market?
... but that would risk canibalizing CPU revenues.
Still, smart thing would be do it early. Otherwise if they join late, it would mean they are desperate - that could be interpreted as admission to failure of their own lines of CPU. That would not come as shock to many, since x86 architecture long outlived its "best before" date.
"Why haven't SSDs been flogged with enough cap on board to flush ram to the drive?"
they have , e.g. http://www.anandtech.com/show/4244/intel-ssd-320-review or more detailed analysis http://www2.electronicproducts.com/Supercapacitors_for_SSD_backup_power-article-farr_capxx_mar2009-html.aspx
This is, however, fundamentally different from solution presented in the article, because RAM being flushed in SSD is just its own internal cache. There is (was) currently no way to flush RAM used by CPU to SSD in the even of power failure [*], since DRAMs need too much power and the operation would have to facilitated by the CPU which needs even more power. Since the power has been just cut, there is no way to do it. *) Unless you have some kind of high-capacity power backup (i.e. battery - see above how this compares to capacitor) in which case you have classical hibernation while the computer is sustained by UPS.
Or you could just bypass CPU and marry flash with DRAM which is what these guys have done. In other words the memory module is in CPU address space (thus usable to programs directly, like any RAM) , there is 1:1 size relationship to flash and no battery is required.
This is innovative - although I suspect if price was announced too, one might possibly say "not so much" .
RAID controllers use external battery to sustain write cache in RAM long enough so that cached data can be written to the attached storage (which must have its own power, too) in case of power failure. So basically what you are talking about is backup power for RAM cache for nonvolatile storage.
This is different, since flash capacity is same size as RAM and there is no external battery involved. You can see it as either of:
* regular DRAM backed by nonvolatile storage (awesome)
* nonvolatile storage with full size DRAM cache (probably very expensive per GB)
The most interesting thing is blurring line between RAM and nonvolatile storage, which could facilitate new system programming paradigms. For example, why read the data from a filesystem, when you can simply map it to your process, *without* any IO or cache on the way? What do you want your filesystem to look like when IO is removed from the equation? Why do you need filesystem in the first place if your data can live in memory indefinitely? (well you do, but consider carefully how are you going to use it!).
The challenges are almost as exciting as those posed by memristors and we will have to answer these questions, assuming there is more than grain of truth in this http://www.theregister.co.uk/2011/10/10/memristor_in_18_months/
Needless to say, open source is best positioned to set the course :)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
