Problems? See this thread at archive.org.
??
Anything anyone know of that can compete with the specs AND the real function AND the popularity/satisfaction of the Toshiba CB35-B3340?
13.3" 1920x1080 IPS Display [which everyone raves about]
N2840 or better?
4GB or better?
SSD? Yes its small but big enough.
Fast OS. Boot times on the order of 12 seconds?
Light
Fanless and does not overheat... or even get hot
USEFUL touchpad that works! [remember the good 'ole days when touchpads functioned?]
$299 list. $250 on sale.
Chomebook is quietly taking over the Education space in USA.... Future buyers being trained
-
-
-
-
To be perfectly honest, the "light but functional" ideal for a ...not a laptop, but a pad with a keyboard, or a lightweight sliver where the screen is the heaviest bit, is a lost cause. It was essentially lost a long time ago when the ION implementation in practice turned out to be a discrete gpu + cpu package (on the pain of Intel lawsuit -http://arstechnica.com/business/2009/11/intel-and-amd-bury-the-hatchet-under-125-billion-in-cash/ ), rather than a SoC design with integrated graphics units and the bus-construction (i.e., the ISA, the industry standard architecture) spread out on several boards. Note that the infamous "settlement" in the link above here involves that AMD cannot integrate these circuits in their SoC designs. And neither will Intel move ahead on that front; it's completely calculated, and intended to avoid the situation where you would get these small one-chip platforms that run x86 instructions.
But it was done, it was demonstrated that it works, it was a perfectly good work-computer, and it could run video in HD at 3-4W. It buried the EeePC designs at the time, and the entire computation unit was one chip with an integrated bus. There are no real disadvantages to doing this from a technical perspective in relation to performance. In fact, it's the opposite case - improvements on bus-technology and ram interfaces that we knew frankly in 2001 would have to happen, and which they did, would (and obviously still will) make an integrated bus the best solution. But it's not going to happen on an Intel platform or an x86 platform to the extent that it makes the device competitive on performance.
The only option for us - and I'm talking about the technical part here - left out to AMD was to pursue one of it's lost causes from back in 2010, or the integrated cpu and gpu cores with a separate bus-interface between the two "computation devices" on the same chip. Better known as Llano, along with the disaster that was for AMD commercially. But it's an "improvement" on that design that turns up with the carrizo-l platform eventually now, where they offer that "full desktop feature" platform in a chip, just with the graphics bus and external component array placed on separate cards outside the die. Where the same mainboard and chip construction can be specified to anything from 12-35w depending on the laptop's cooling scheme, size, market positioning, etc.
If laptop-manufacturers find a reason to pounce on that design - for whatever reason there would be over differentiating the laptop market in "work laptop", "gaming laptop", "sub-notebook", etc. That, as we all know, allows laptop-manufacturers to reuse their designs and sell them as new every year at a very high rate, at quite high prices, far over what the actual hardware itself is worth over time. I would imagine that "consumer demand" would be such a thing, if people are more articulate about what they would like to see and use, and we don't end up with exclusive content marketing on Anandtech that dominates the market's focus. Although that seems one excessively high order considering how things have been working over the last 15 years. But if that happens, what we would end up seeing would be smaller laptops with very decent battery life that would work as fairly powerful net-tops for work, travel, entertainment. That would make fairly large parts of the "cheap laptop market" obsolete. And where laptop-makers would suddenly want to compete on screen, design, quality, keyboard and chassis designs. And, in my imagination, also on ability to upgrade and switch out components, add various features from screens, external graphics cards, network components, cpu, etc., via the pci-e reboot (that frankly takes up back to scsi more than anything
, with high-level programmable interfaces controlling device response and so on).
But it's going to be dependent on choosing that platform and spiting the biggest market, and gambling on that people will think about what they're buying for a while, and then pick something "untested" and unpopular that might have issues of all kinds. It also means burying the Intel/Microsoft deals that laptop-makers essentially live by. So this isn't going to be a complete turnaround in another 10 years even if laptop-makers actually wanted to offer these types of products that compete in, like mentioned, screen, battery, watt-drain, performance as high as possible on passive cooling schemas, etc.
And it's from that background that you get things like the Chromebook, that essentially is a mobile phone that looks like a laptop. Where they choose ARM designs that in truth are more than decent (the Tegra 3 design being my favorite - still have and use often a Transformer Prime based on the tegra 3 design). On the condition that they are not to directly compete against any laptop devices the same manufacturer sells. That was the entry point for these devices, but try to understand that these chips - even if they're quite capable, and you can actually run a completely decent linux distro on them, with flat and standard recompiled programs for the platform without any issues whatsoever - will never compete with the overall laptop segment from necessity and marketing concerns the laptop-makers themselves have. The Chromebook for example is this outlier google came up with, that they could justify by marketing their own brand (and locking you into all kinds of fantastic software, etc.), and laptop-makers can't do that.
In fact, it's the same problem that laptop makers who ship linux distros run into - they have to compete on hardware, and there's no money in that. By consumer choice. People are not willing to pay for a more expensive laptop with better screens as a rule, and the market essentially is oriented around the old and tried schema where you offer the deluxe version to the enthusiasts, and the cut-down version to the business-partners and the larger market globally, the budget market. You then essentially run into the issue that you're sitting on a Rolls-Royce you want to sell to the general public, that performs the same as everything else, but looks and drives very comfortably - but you have to offer it at a higher price. And even if the price isn't massively higher, you can't justify it over selling cheaper plastic shells and so on.
So you say, but hey, look at Apple - they did it? Sold the same device and competed on OS and features. Certainly there's a market for that? And the truth is that there isn't a market for that unless you can insure that your customers stay on your platform and buy fairly expensive software that only runs on your platform. The entire idea of selling a device that is open, performs in the right segments, and encourages people to use free software over commercial software, is simply not tenable in the market as it exists. Because, as explained, you don't have a market where manufacturers compete on hardware features such as screens, possibly optional software bundles, upgradeable cpus, brilliant plexiglass designs that would outlast all the components in the laptop, etc. That simply does not exist, and I would estimate that it won't happen in the next 20-30 years, simply because entering the market as an actor who wants to sell such a product is not possible with the current hardware availability - even if it is technically possible to go for it, and has been for at least 10 years.
It's a bit like wanting to go into a business that does actual journalism rather than "content marketing", or where you sell integrity and critical review as a feature rather than a quirk that select customers are entertained by in small doses. You can't do it, because you won't get partners in the segments you're covering in on it to sponsor you(obviously - they can just go to any amount of **** blogs and often don't even have to pay them to get advertisement retold in "casual language", or what appears to be proper reporting). And you also don't get your readers in on it, because they prefer to get their news "straight from the source", as one spiteful guy told me outright, when they seriously insisted that they would rather trust a corporation's PR department for news than a news-outlet, or so I assume he thought, whose own agenda would dilute the truth of things.
Of course, this is something we've brought on ourselves. Like I said, the battle was lost many years ago. And it's lost by the fact that content portals, indeed forums, newspapers, etc., notice pretty much instantly that original content has to be of such high quality that it stands out as literary works of art before it's ever shared or read by external sources. While embedded blogging and sponsored content gets hits and generates activity by virtue of being assisted by said PR departments with a good SEO strategy. Or simply through common tagging in google ads. It's simply an extremely difficult proposal to suggest that you're going to start a subscription-driven service. And if you do so anyway, and dedicate time to it, you end up with a small isolated segment that has a fraction of the traffic that potential subscribers will even see in the first place.
And even if it did succeed, what influence do you have? You're orienting your entire business-model around demonstrating and telling the implicit story about how all major outlets who get press-contacts and early news-content are essentially corrupt, and that they participate willingly in shaping market-demand at the request of any industry contact. And you're doing this while saying, straight out, that you never expect to get early news, you never expect to get exclusive interviews: you only offer your critical voice. And it has a price that is so low it can be bought by mere subscription fees.
So what you see more and more of now is that content marketing becomes an operative standard, that easily outcompetes original content, by direct consumer request. Two examples: PCGamer, who always did their own thing back in the day, and had a cordial relationship with advertisers, as long as they stayed off the editorial pages and article content - discover that there is a real demand for a brand-specific review, for example. People don't search for general terms and function-oriented reviews, they search for finding the product they want from a specific brand. Obviously more general content exists, even at PCGamer, but they do get traffic in on the brand-specific hall of fames much more often than the other kind. People in fact swear to brands and want to read content about that brand. That's what "little brother" and SEO research undeniably insists. That your average customer wants to see types of content that looks at brands separately, without confusing them with analyses that create difficult choices. And this doesn't come in a folder from that brand's PR department, you can read it right out of your own content.
Kotaku is another good example. They used to have these pink articles that were headed off with a huge watermark that said "sponsored content". And they discovered, I'm sure not to a great surprise, that this sponsored content often was better coverage of a specific subject than their original content. And sometimes, very often when the content would simply be regurgitations of PR events anyway, the whitepapers with the summaries basically were more readable than the oral style some of their writers would, ignorant otherwise, attempt to cushion the corporate PR tour de force in. On many occasions you also could not see the difference between their sponsored content and the original content, something which obviously the PR departments would see as a great success - that they were able to mimic the style to such a degree that readers would not be able to tell the difference between targeted advertisement and the actual article written by allegedly independent writers. Which obviously hangs together with how, as mentioned, there really is an outspoken demand for getting "truth" "straight from the source". Rather than having something coloured by the biases of the writers.
Now - this is a bankruptcy statement for actual journalism, that we are unable to provide or produce content that actually give consumers value over that which the PR departments offer on their own, with studied awareness of market demands and what people tend to fall for when it comes to content they like to read. And it's also the seed of what ends up being a cordial relationship between publications and PR departments, where publications actually have people hired as writers to create content specifically, in the publication's style, that appeals to advertisers. You have to understand that this is a real thing, and that selling these articles to sponsors is the main income for most tech and entertainment publications, just as it is the biggest part of your average news-publication.
Where the goal, as explained, isn't to really inform buyers or customers about tech, or give them tools to make good choices. But to inform them of what the advertisers want. You even get, and this is common, the idea that it's actually a benefit to the publication to have this relationship, as it gives them more access to get information they can then write their niche-articles about. And it does - but like said, they don't compete with the masses of serial-produced content. And that's the consumer's choice screaming out very clearly - we prefer to read, by statistics, the content that confirms your own views, and the content that predictably reads pretty much the same every time you click it, at any time of day, in very frequent and predictable batches.
And in that market, with these types of customers, you don't front anything that challenges people's views. This is democracy, right? Where you turn for example my own ideal of an independent press, served to informed participants in a capitalistic society - into sheer activism. While the embedded content marketing and paid for content is simply "neutral", or even more honest. That's the result here, and it's the customer's choice to have that. You might argue it's not by choice, but it is: we do not ask questions, and the ones who do are not supported in a real way. We do not care enough collectively to change it, and perhaps more importantly, we don't collectively trust independent voices to actually be upfront about their biases, or to be discernible enough to actually provide critical analyses that have value over something that we know is more predictable and easily digested - and in the most extreme sense therefore more honest.
Back to AMD and Intel: what you're looking at is that your wish for a passively cooled intel platform or an x86 compatible like AMD, is not market-sustainable. It won't be inserted as a narrative into the sales-strategies, as it would suppress products that already are profitable. Actors that might want to challenge that, even with the relatively cheap costs hardware can be produced at now, are met with an insurmountable obstacle in that they don't actually offer a competing product, but a new product altogether that offers something incomparable.
And this brings us back to the embedded content again - how would these challengers actually get their version of the story out? Surely someone who covers their story without balance are simply super-enthusiasts for that brand, right? Who have, for invisible reasons to you, chosen to firebrand that product - likely from a content marketing standpoint. So you end up with having to insert yourself in the existing narratives - by paying for advertisement, in an amount that competes with their market share competitors. Otherwise, you are invisible. Or, more accurately, you're running propaganda.
And the point of the entire rant is that this affects the kinds of products these challengers would be able to or even want to offer on the market. You offer something that fits in the "realities of the market", or you can forget it altogether. Even if, like explained, you would very much like to do it anyway.
So remember that, the next time you click down the sponsored content, or have a paid for article dropping down in your mail you enjoy reading, or read a clique from the main newpaper you own with fantastic photographs about a current event that somehow seems utterly related to a particular product you recognize (Coca Cola are awesome at this): maybe you would have been willing to imagine a reality where you click weird-ass crazy blogger content from guys who stole a Press-pass and snuck into the PR event, rather than getting the sanitized version from a disinterested "neutral" journalist. And would be willing to finance said content by a no strings attached subscription fee, rather than through google adverts. That in the end materially is a way to rate content and assign monetary value to the content that is "most interesting" to the advertiser of the related products. To the point where that content will be favoured, not just by the advertisers, but also by the ones who pay for writers to write content on their portals.
Just keep it in mind. Because you get a lot of writers out there who see no difficulty with accepting that the most traffic-generating content by definition must be the best. And therefore adopt an attitude towards covering content that functionally is no different from advertisement. Frankly, my greatest surprise about this was to discover that prospective "journalists" often also adopt that style, because it's easier to sell to the publications, of course. But also because they genuinely think that that is what journalism looks like. Advertisers of course rejoice in that phenomenon, because they now have an inexhaustible supply of people who write advertisement for them, for free. To the point where they compete out the few publication paid journalists who actually have some work ethics - which is a serious problem in tech: you simply don't have to deal with critical questions, or answer something you haven't prepared for. So you don't.Mr.Koala and Starlight5 like this. -
..yeah, I'm not completely sure if it's just difficult to get someone on the record about it, or if it's difficult to explain it well.. or at all.. by actually referring to these patents specifically (stuff that I don't have to because I'm just writing bs on a forum - but then again, the actual patent language is completely opaque, and doesn't explain much at all).
But it has to do with that abandoned design AMD had to combine gpu and cpu cores into a single type of "core", or series of computation segments on the physical cores, and then just assign jobs to them as needed. Because.. a cpu is of course capable of performing gpu code. And with CISC generally being optimized on the microcode layer, where practically all improvements on performance has happened since Core2Duo. Then.. you just have to want something like this to happen. And it's not like you don't have offshoots that very specifically is geared into this paradigm at Nvidia either. The entire Tegra chipset is based on exactly this schema: get more instructions on the same cores, to get more potential microcode optimization. And to allow complex instruction set commands to be programmed on the high-level.
You don't need extreme amounts of imagination to see how that might cause problems for a lot of the industry, though. Graphics-cards, all things cpu, dram, bus construction, mainboards, peripheral units. Kind of making everything obsolete, and moving everything proprietary and expensive in either domain directly to the trash heap. So it's not incredibly curious that Intel would invoke the licensing patent for x86 compatible cores, to stop the instruction set from being embedded on chips like that.
I mean, it makes sense, right? They don't own x86, but they own the conditions for how the instruction set is to be used when put in hardware. So why not screw us as customers, and everyone else, really - as opposed to giving away their entire business. That's.. just common sense.
But it has a few implications for what exactly has to happen before we're going to see integrated SoC designs with many, many cores that are programmable with custom instruction sets capable of complex math. Where you would then be able to exploit parallelism in the same way, but to a much higher degree than what is done on a graphics card. While of course also being able to mix direct cpu-type direct order number-crunching jobs with other parallelizable tasks asynchronously.
Because.. I'm not sure, but it kind of sounds like the argument implies that as long as something can technically speaking be reduced mathematically to x86 instruction set commands. Then you're not allowed to put this on a computation core that does other types of operations as well, and then sell it afterwards without Intel accepting it. Including that it's the other way around - say you had a core that extends x86 instructions, for example - and you're in trouble. Or you have a computation device that among other things can execute x86 instructions, or be programmed to do it, and it's suddenly a problem legally.
I expect that that's not what the patent actually says, but the arguments I heard sure sound like that. -
-
-
-
-
I understand. Things can really get messy when you have too many independent systems to manage. I run into similar problems managing what software is installed on what system. "XXX: command not found" is no biggie in the office but really scary in the field.
Currently the best solution I've found is keeping all my production environments to one distro (which shouldn't be a problem under Windows) and use a script to do most of the initial setup when I get my hands on a device. As long as the script doesn't throw out errors I know I haven't forgot anything.alexhawker likes this.
What WIntel-thing can compete vs. CB35-B3340?
Discussion in 'Hardware Components and Aftermarket Upgrades' started by cognus, Oct 20, 2015.
