The other part is that a 'normal' connection can be 100 gigabit, and most of even the more aggressive networking environments don't usually go beyond 10gbit a part, and the industry standard switch chips are readily available to do those at full line rate with large numbers of ports.
With that, the things that used to be cisco's bread and butter (proprietary switch 'stacking', chassis switches, etc) become less relevant. In fact, in many product areas they allowed themselves to be leapfrogged in many performance areas (cisco was one of the last switch vendors I could remember trying to sell gigabit switches that were internally oversubscribed, for example).
Nobody's going to demand Cisco "just cut back profits" as they lose business"
Well, if they don't cut back profits, they *will* lose business. That's the problem with the leaders of these businesses, they demand growth, but cut back. Of *course* you can't sustain or grow your business if you are defeatist and ditch your people. A good business will accept lower profitability for the sake of investing in some way that delivers growth next go around.
Of course, if you are consigned to a reality that you *won't* grow, go ahead and do the layoffs to increase margin to milk your brand strength while the milking is still somewhat good. Don't expect that to last very long though, as you piss away the value that build that brand strength in the first place.
It would be one thing if they were losing money and simply could not afford to invest in having good talent, but when you have the profit margin to spend, better spend it on your technical vitality rather than tossing it all at the shareholders just because times get a bit lean.
Incidentally, I think that curve is generous. It's a way for Gartner to say 'just because the hype died down, our insight on that fad is still valuable because it will come back'. I would say more often than not, the 'disillusionment phase' is a bit more persistent, and 'enlightenment' is not something that frequently improves the fate of a 'dead' hype.
The problem being that 'smart' is not exactly open and shut, and if we are serious with ourselves, our current 'neural networks' may be fundamentally different from the way human intelligence functions. It's just the most convenient model we've figured out to do some fancy tricks.
At this point, ML can do some useful tricks, but it takes a gigantic amount of resource to train something so that it could almost kind of sort of compete with toddlers at very specific recognition tasks.
The problem is what people are hyped about and how it doesn't connect to the proclaimed conclusions.
ML Training and inference do not have any plausible path to replacing 'programming', even in theory. So this specific claim is ridiculous, since this is pretty much the *only* 'AI' branch people are talking about, because we've traversed a moderately interesting inflection point in terms of hardware and research for training.
Not all machine vision is ML. Given the timeline, most check reading is not inference after ML training. Ditto for a lot of face recognition.
That's one of the peculiar things, ML Training for image recognition of new things is a popular demonstration, but most of the day to day image recognition was done using techniques prior to that being practical. New we can much more easily create new recognition, but demos still like to use faces because it's approachable. Counterproductively though, a high end machine vision demo in a conference doesn't *look* all that different from what an attendee has seen their own phone do.
Additionally, contrary to what Google and nVidia will tell you, there aren't *that* many people with a good idea of a useful goal even when they understand the principles.
Similar problem has afflicted 'big data', lots of people who know the principles and can do useless examples, not that many people who have an idea what to do with those techniques.
Especially the 'machine learning' focus area that nVidia is in love with is not remotely ready to deal with it.
As it stands, there's a relatively narrow field of problems that the techniques currently called 'AI' can be applied to that make even the vaguest hint of sense. Even then, there are a lot more people hyping it up and excited about *how* it works than there are people with ideas of *what* to do with that capability.
Well, at this phase, it won't even physically go into anything apart from server designs built specifically around this specific card.
Here there's an issue of volumes. While the enthusiast gaming market is small, the number of units to move of this sort of accelerator makes it look gigantic by comparison. It's interesting, since nVidia began coming to prominence when people started figuring out how to use off the shelf GPU to accelearte HPC workload, because the accelerator market couldn't deliver a viable product for an acceptable price.
Now, it seems they've come full circle and are able to sell exhorbitantly expensive accelerators, with very little in common with their GPUs anymore..
Two, the comparison would be 7.5 Tflops, since V100 is 7.5 DP64, and top500 focuses exclusively on FP64 performance
Three, we are comparing Rpeak to Rmax (and Rpeak is increasingly not sensible).
Of course, all that said it's still an impressive acheivement, and their big headline about 120 'tensor tflops' is what they seem particularly focused on, though I have no sense of how impressed I should or shouldn't be, since I don't know tensor performance so much.
Analysts are sadly useless for knowing about what you *should* do or what the *future* is. They can do a serviceable job of past trends and current data, but their predictions about what could and will succeed are about as accurate as flipping coins.
To do IPO they brought in analysts, who made recommendations.
I see a rough road for IPO at this phase. They've been a fixture for over 10 years and their repeated attempts to succeed as a business have been widely observed and have failed. While undoubtedly popular, it is painfully obvious because they are the most straightforward free option. They have not shown any hint of being able to parlay their status to significant revenue. Instead they have to keep hand waving less useful metrics about users of their software than any business relationships, and intentionally fuzzing things up by swapping the word 'customer' and 'user' as it makes sense ('user' to have big numbers and share, then pivot to referencing customers, to suggest the users==customers, rather than the reality that the vast majority of users of the platform will never become a revenue source).
If they had IPOed 10 years ago, things would have been new enough for the investors to be enamored with the visions of what *could* be, but the passage has time has dashed pretty much all of the hopes.
That's true, but also, a job that can find 20% opportunity to work remote still means you are working 80% there. So even if you can't overcome the misconception that you can just boil down everything to a math problem of headcount and one human is the same as any other human, you can point out that the vast majority of time is local.
If your goal is employee retention and attracting talent, then flexibility on remote work is a valuable perk. This is pretty much the entirety of what is being said, that the *employees* like being able to work remote (which of *course* we do, we like more options)
If your goal is assuring your team is the most effective, and you have little to no concern about labor shortage or retention, then an employee who can work in the office at least *some* of the time is invaluable.
Doing meetings remote can work ok, but it still pales in comparison in having everyone sitting in the same room. Even if you aren't *explicitly* marking some time as collaborative, the ability to turn your head and speak is great.
Being videoconferenced in during the whole day is just too bizarre and seemingly obtrusive. A text chat is suitably unobtrusive, but can drag on. Also if you see someone, you get a sense whether they are really busy or maybe they have some time to help out.
The truth is a balance is best, some home time for a comfortable working environment and at least for *some*, fewer distractions than at work. Some time at work for collaboration, and for *some*, fewer distractions than at home.
Incidentally, I don't argue that Surface is good and iPad is bad, my argument is that Tablets are not the PC market extinction event that almost every outlet predicted. Surface tablets are a bigger letdown than iPad, but that doesn't mean iPad isn't a letdown. My argument is that MS giving up on Surface tablets would make sense, just as Apple putting iPad far in the background is a sound decision on their part, at least a more sound decision than how much Apple is neglecting their desktop/laptop product line
They made 36% more DOLLARS on products (Macs) that generally cost 3 to 6 times that of a typical iPad.
While units moved of iPads could have been higher, from a business case the dollars are what matters. If you are trying to say the iPad is a gateway for app store revenue, what yo uhave to be counting are App store accounts, since the device count doesn't correlate. There may be a story there, but data would be required..
the dollars don't lie.
Yes, the dollars say the tablet form factor is not going to be driving growth any time soon, regardless of vendor. Certainly the concept of a 'premium' tablet has been dead on arrival for the general mass market (Surface Pro, iPad Pro, etc).
Same across the board for most 'digital' with physical counterparts. If I 'buy' a digital video, it'd be more expensive (and more limited) than a blu-ray copy. Steam is about the only venue I've seen the downloaded copies sell for cheaper than boxed copies.
Actually it is front lit, and a very nice frontlit design at that. If you are in a bright room, the light can be off, as the display is working *with* the environment lighting rather than against it.
I would *hope* so, but there's frankly a limit on how many people would purchase a work. I'm not so sure that a 50 shades series book would have much opportunity to sell to more people than they already did. I don't think I've heard many people say "oh, I would read that book, but I can't afford it".
Of the people I know roughly their buying habits, their time budget for reading limits them far more than their money budget.
MS, like almost every company on earth, would kill to have a "fad" like that...
The article you link in fact state "In contrast to the iPhone and Mac, the iPad continues to struggle". They made a lot more even on Mac computers. Their competitors make more (revenue) on laptops/desktops than Apple does on Macs. While no company would turn down an extra $5 billion in revenue certainly, the players in the industry don't have much reason to be *exceedingly* envious of that particular product.
iPad fever had the world on fire as it went from $2 billion a quarter to 5 and then 11 billion, with people assuming that trend would continue. $11 billion was respectable in its own right and would outpace most companies PCs sales if sustained, but people were *mostly* focused on the presumed future. Since then iPad sales half fallen to half of that, without a sign of that trend reversing.
I don't know if it sells at a loss or not, but 'revenue' does not mean it didn't sell at a less. If in that quarter, they spent $1.5 billion on the product development and manfucature, then it would be a loss. If they spent $1 billion, thene it would be a $300 million profit.
Slashdot Mirror
The other part is that a 'normal' connection can be 100 gigabit, and most of even the more aggressive networking environments don't usually go beyond 10gbit a part, and the industry standard switch chips are readily available to do those at full line rate with large numbers of ports.
With that, the things that used to be cisco's bread and butter (proprietary switch 'stacking', chassis switches, etc) become less relevant. In fact, in many product areas they allowed themselves to be leapfrogged in many performance areas (cisco was one of the last switch vendors I could remember trying to sell gigabit switches that were internally oversubscribed, for example).
It's early in the thread, but I'll bite:
Nobody's going to demand Cisco "just cut back profits" as they lose business"
Well, if they don't cut back profits, they *will* lose business. That's the problem with the leaders of these businesses, they demand growth, but cut back. Of *course* you can't sustain or grow your business if you are defeatist and ditch your people. A good business will accept lower profitability for the sake of investing in some way that delivers growth next go around.
Of course, if you are consigned to a reality that you *won't* grow, go ahead and do the layoffs to increase margin to milk your brand strength while the milking is still somewhat good. Don't expect that to last very long though, as you piss away the value that build that brand strength in the first place.
It would be one thing if they were losing money and simply could not afford to invest in having good talent, but when you have the profit margin to spend, better spend it on your technical vitality rather than tossing it all at the shareholders just because times get a bit lean.
Incidentally, I think that curve is generous. It's a way for Gartner to say 'just because the hype died down, our insight on that fad is still valuable because it will come back'. I would say more often than not, the 'disillusionment phase' is a bit more persistent, and 'enlightenment' is not something that frequently improves the fate of a 'dead' hype.
The problem being that 'smart' is not exactly open and shut, and if we are serious with ourselves, our current 'neural networks' may be fundamentally different from the way human intelligence functions. It's just the most convenient model we've figured out to do some fancy tricks.
Or even the worst humans....
At this point, ML can do some useful tricks, but it takes a gigantic amount of resource to train something so that it could almost kind of sort of compete with toddlers at very specific recognition tasks.
Except you likely would design it to be obsessively interested in programming to spec...
Actually this is how a lot of crappy software happens. The 'spec' is generally lacking in vision/understanding.
The problem is what people are hyped about and how it doesn't connect to the proclaimed conclusions.
ML Training and inference do not have any plausible path to replacing 'programming', even in theory. So this specific claim is ridiculous, since this is pretty much the *only* 'AI' branch people are talking about, because we've traversed a moderately interesting inflection point in terms of hardware and research for training.
Not all machine vision is ML. Given the timeline, most check reading is not inference after ML training. Ditto for a lot of face recognition.
That's one of the peculiar things, ML Training for image recognition of new things is a popular demonstration, but most of the day to day image recognition was done using techniques prior to that being practical. New we can much more easily create new recognition, but demos still like to use faces because it's approachable. Counterproductively though, a high end machine vision demo in a conference doesn't *look* all that different from what an attendee has seen their own phone do.
Additionally, contrary to what Google and nVidia will tell you, there aren't *that* many people with a good idea of a useful goal even when they understand the principles.
Similar problem has afflicted 'big data', lots of people who know the principles and can do useless examples, not that many people who have an idea what to do with those techniques.
Especially the 'machine learning' focus area that nVidia is in love with is not remotely ready to deal with it.
As it stands, there's a relatively narrow field of problems that the techniques currently called 'AI' can be applied to that make even the vaguest hint of sense. Even then, there are a lot more people hyping it up and excited about *how* it works than there are people with ideas of *what* to do with that capability.
Well, at this phase, it won't even physically go into anything apart from server designs built specifically around this specific card.
Here there's an issue of volumes. While the enthusiast gaming market is small, the number of units to move of this sort of accelerator makes it look gigantic by comparison. It's interesting, since nVidia began coming to prominence when people started figuring out how to use off the shelf GPU to accelearte HPC workload, because the accelerator market couldn't deliver a viable product for an acceptable price.
Now, it seems they've come full circle and are able to sell exhorbitantly expensive accelerators, with very little in common with their GPUs anymore..
Well, one, that was 1997.
Two, the comparison would be 7.5 Tflops, since V100 is 7.5 DP64, and top500 focuses exclusively on FP64 performance
Three, we are comparing Rpeak to Rmax (and Rpeak is increasingly not sensible).
Of course, all that said it's still an impressive acheivement, and their big headline about 120 'tensor tflops' is what they seem particularly focused on, though I have no sense of how impressed I should or shouldn't be, since I don't know tensor performance so much.
So if there are're many professions in OSS, why there is no Linux distro or other free OS of comparable quality to Windows 7 or even XP?
Why should the distros *lower* their quality? ;)
Analysts are sadly useless for knowing about what you *should* do or what the *future* is. They can do a serviceable job of past trends and current data, but their predictions about what could and will succeed are about as accurate as flipping coins.
To do IPO they brought in analysts, who made recommendations.
I see a rough road for IPO at this phase. They've been a fixture for over 10 years and their repeated attempts to succeed as a business have been widely observed and have failed. While undoubtedly popular, it is painfully obvious because they are the most straightforward free option. They have not shown any hint of being able to parlay their status to significant revenue. Instead they have to keep hand waving less useful metrics about users of their software than any business relationships, and intentionally fuzzing things up by swapping the word 'customer' and 'user' as it makes sense ('user' to have big numbers and share, then pivot to referencing customers, to suggest the users==customers, rather than the reality that the vast majority of users of the platform will never become a revenue source).
If they had IPOed 10 years ago, things would have been new enough for the investors to be enamored with the visions of what *could* be, but the passage has time has dashed pretty much all of the hopes.
That's true, but also, a job that can find 20% opportunity to work remote still means you are working 80% there. So even if you can't overcome the misconception that you can just boil down everything to a math problem of headcount and one human is the same as any other human, you can point out that the vast majority of time is local.
If your goal is employee retention and attracting talent, then flexibility on remote work is a valuable perk. This is pretty much the entirety of what is being said, that the *employees* like being able to work remote (which of *course* we do, we like more options)
If your goal is assuring your team is the most effective, and you have little to no concern about labor shortage or retention, then an employee who can work in the office at least *some* of the time is invaluable.
Doing meetings remote can work ok, but it still pales in comparison in having everyone sitting in the same room. Even if you aren't *explicitly* marking some time as collaborative, the ability to turn your head and speak is great.
Being videoconferenced in during the whole day is just too bizarre and seemingly obtrusive. A text chat is suitably unobtrusive, but can drag on. Also if you see someone, you get a sense whether they are really busy or maybe they have some time to help out.
The truth is a balance is best, some home time for a comfortable working environment and at least for *some*, fewer distractions than at work. Some time at work for collaboration, and for *some*, fewer distractions than at home.
Incidentally, I don't argue that Surface is good and iPad is bad, my argument is that Tablets are not the PC market extinction event that almost every outlet predicted. Surface tablets are a bigger letdown than iPad, but that doesn't mean iPad isn't a letdown. My argument is that MS giving up on Surface tablets would make sense, just as Apple putting iPad far in the background is a sound decision on their part, at least a more sound decision than how much Apple is neglecting their desktop/laptop product line
They made 36% more DOLLARS on products (Macs) that generally cost 3 to 6 times that of a typical iPad.
While units moved of iPads could have been higher, from a business case the dollars are what matters. If you are trying to say the iPad is a gateway for app store revenue, what yo uhave to be counting are App store accounts, since the device count doesn't correlate. There may be a story there, but data would be required. .
the dollars don't lie.
Yes, the dollars say the tablet form factor is not going to be driving growth any time soon, regardless of vendor. Certainly the concept of a 'premium' tablet has been dead on arrival for the general mass market (Surface Pro, iPad Pro, etc).
Or else they learned that for their market, volumes didn't increase enough with price cuts to offset the per-unit revenue loss.
Same across the board for most 'digital' with physical counterparts. If I 'buy' a digital video, it'd be more expensive (and more limited) than a blu-ray copy. Steam is about the only venue I've seen the downloaded copies sell for cheaper than boxed copies.
Actually it is front lit, and a very nice frontlit design at that. If you are in a bright room, the light can be off, as the display is working *with* the environment lighting rather than against it.
I would *hope* so, but there's frankly a limit on how many people would purchase a work. I'm not so sure that a 50 shades series book would have much opportunity to sell to more people than they already did. I don't think I've heard many people say "oh, I would read that book, but I can't afford it".
Of the people I know roughly their buying habits, their time budget for reading limits them far more than their money budget.
MS, like almost every company on earth, would kill to have a "fad" like that...
The article you link in fact state "In contrast to the iPhone and Mac, the iPad continues to struggle". They made a lot more even on Mac computers. Their competitors make more (revenue) on laptops/desktops than Apple does on Macs. While no company would turn down an extra $5 billion in revenue certainly, the players in the industry don't have much reason to be *exceedingly* envious of that particular product.
iPad fever had the world on fire as it went from $2 billion a quarter to 5 and then 11 billion, with people assuming that trend would continue. $11 billion was respectable in its own right and would outpace most companies PCs sales if sustained, but people were *mostly* focused on the presumed future. Since then iPad sales half fallen to half of that, without a sign of that trend reversing.
Anyone can sell products at a loss but no one will judge that as a success.
Obviously you haven't been paying attention to the myriad of unicorn tech startups that never make money ;)
I don't know if it sells at a loss or not, but 'revenue' does not mean it didn't sell at a less. If in that quarter, they spent $1.5 billion on the product development and manfucature, then it would be a loss. If they spent $1 billion, thene it would be a $300 million profit.