8 year upgrade cycles future? Intel running on fumes.

[cbn]

The future of home computing I believe is thin clients. One powerful server PC with a bunch of cheap thin client laptops. There will be a large powerful home server (acting like an appliance) with multiple mobile devices(thin clients) connected to it for processing power. More and more companies are moving to Citrix where everything we do is done via a virtual desktop and I have no doubt one day our personal home computing will follow the same trend.

That makes sense to me.

And maybe in some cases the Linux console I discussed in post #10 becomes this home server with the addition of more cores and/or larger iGPU.

 

[postmortemIA]

At two of the places I work, they changed their mid-2000's XP machines in 2014 -- one before the XP support cutoff; one after. The companies involved changed the machines only because support was cut off for the XP machines.

We're on Win 7 now. MS has announced they will cut off support for Win 7 in what -- 2025? So that's an 11-year update cycle. Will those machines be obsolete? No, for the work we do now (mostly medical data entry). I suspect retailers are down with my companies.

2020 is end of extended (security fixes) support. So most companies will move in or by 2019.

The best upgrade right now, even in companies is to switch to SSD. It is pretty cheap upgrade that extends life of PC considerably

 

[frozentundra123456]

How much profits did Polaroid,Kodak,lehmans,gm did in the mid 1990?

Stop thinking one quarter at a time and think 10 years down the line.

Its not my job to think 10 years down the line, I dont work for Intel. There are also servers and emerging markets in addition to replacements.

Point is, I dont think anybody would argue that Intel doesnt face serious challenges down the line. But titling a thread that Intel is running on fumes is the height of hyperbole, and just seems like a pointless, negative thread, unless you have something constructive to say to suggest how they should solve the supposed crisis.

 

[ehume]

2020 is end of extended (security fixes) support. So most companies will move in or by 2019.

The best upgrade right now, even in companies is to switch to SSD. It is pretty cheap upgrade that extends life of PC considerably

I very much agree with you about SSD's. I replaced a 500GB spinning disk with a 500GB SSD so my wife could keep on using her 2008 machine. But the IT departments of the agencies that service the places where I work have no time for that. For the cost of replacing a HD they can get a whole new machine.

So it looks like we can expect new machines in 2020. Nice to know. Thanks.

 

[Charmonium]

The market will find a solution regardless of what Intel or AMD do. Intel displaced IBM in the hardwar business and I expect someone else will displace Intel if they're unable to make the breakthroughs necessary to advance computing.

 

[videogames101]

People want faster, smaller, more battery life. Web scale computing requirements are only getting higher. Our computing needs are not decreasing. Intel and others are going to push device and scaling technology until their stock price hits $0. And once that happens, someone else will keep pushing it. In my opinion

 

[soccerballtux]

I upgrade based on RAM requirements. I would upgrade to a DDR3 system right now with 32-64GB RAM and sit tight until 7nm at which point I would upgrade again.

 

[Madpacket]

At two of the places I work, they changed their mid-2000's XP machines in 2014 -- one before the XP support cutoff; one after. The companies involved changed the machines only because support was cut off for the XP machines.

We're on Win 7 now. MS has announced they will cut off support for Win 7 in what -- 2025? So that's an 11-year update cycle. Will those machines be obsolete? No, for the work we do now (mostly medical data entry). I suspect retailers are down with my companies.

Windows 7 is already done. Mainstream support ended in January. Extended support until 2020.

http://windows.microsoft.com/en-ca/windows/lifecycle

 

[senseamp]

The new battlefield is in the cloud, and the application of the future is machine learning. It's an inflection point and those are dangerous to incumbents who have the most to lose and are often slow to react due to complacency or for fear of alienating their installed base.

 

[escrow4]

The new battlefield is in the cloud, and the application of the future is machine learning. It's an inflection point and those are dangerous to incumbents who have the most to lose and are often slow to react due to complacency or for fear of alienating their installed base.

Not here. Everything important is offline at a minimum or hard copy. When its not needed its physically destroyed. All of this small and light rubbish is just designed to last 6 months then get chucked and voila! time for a new tracking device.

I've found that the more tech has infected life the more I've reduced it. I'm now down to 2 fat and chunky coolermaster desktops that look like IS bombed them repeatedly and a fat (though modern IVB) Lenovo laptop that if dropped would likely crack the floor never mind the laptop.

An old Windows tablet and other bits are shelved or chucked. Meh. All this interconnected always online I'm watching you you are in every database NOW NOW NOW NOW is just meh.

 

[biostud]

with an SSD an a Sandy bridge or newer cpu, you will have a computer that run +95% of the software used by mainstream users. So intel's worst competitor is themself. As already mentioned performance/watt is the new king in town, both for mobile and data centers. The only thing desktops are ahead of laptops in, is dGPUs. (And the tiny percentage that actually need a powerful +4 core workstation)

 

[NTMBK]

The new desktop is called "the console" and I would like to see Intel move into that area rather than reserving large iGPUs purely for workstations (including mobile ones).

So...

quad core GT4e = workstation SKUs

dual core GT4e (made from quad core GT4e dies) = console replacement SKUs

P.S. Going a step further, I would like to see the console SKU performance optimized for Linux (maybe even centered around a BGA 5" x 5" reference board). This, in part, so Intel can get another software ecosystem (outside of Android and Windows) into critical mass phase.

I think you posted this in the wrong thread, the 2 core + GT4e one is that way ->

 

[cbn]

I think you posted this in the wrong thread, the 2 core + GT4e one is that way ->

If Shintai is claiming the desktop is dead, and Intel has these new larger iGPUs available it only makes sense for them to expand into this new non-mobile area (Steam Box, etc.)

 

[ShintaiDK]

It's all about growth. The only real growth right now is in the portables area and AFAIK, Intel hasn't made any significant inroads in that area.

There's nothing wrong with being in a mature industry. You can still make money. But you want growth because that's what supports your stock price. Look at the p/e multiples for a company like Amazon or Netflix. They get multiples hundreds of times their earnings per share because of growth. Once that stops, down goes your share price. Even look at Apple. You'd think that would be a huge growth company but it's not. P/E is higher than normal but still only in the 20's - and they're in a growth market. Intel isn't. They're left being dependent upon the upgrade cycle which is gradually getting longer and longer.

Only smartphones got growth that is losing steam fast, and they will soon be a shrinking market.

 

[ShintaiDK]

People have been saying that for well over a decade.

Where else would gamers slap their big GPUs in ?

You mean the discrete GPU market that shrinks with ~40% a year currently?

 

[StrangerGuy]

Except the wall Intel is approaching is opaque, it's unknown if there's anything on the other side. With supersonic air travel the wall is transparent, we know we can go faster it's just an issue of cost and politics. We don't know yet if current processor technology is sustainable, coming up to the wall companies like Intel and AMD have to determine if it's worth trying to knock down or better to find a new path.

It doesn't matter how much Intel want to push fab tech when there is simply no consumer killer app to justify new PCs or Intel CPUs for the vast majority of people. We are now in a situation where even ARM has too much excess general CPU power for this market, computation-intensive H264/HEVC decoding can be handled by the GPU, and game devs on Android/iOS/x86 are not pushing the graphical envelope further due to diminishing returns and increasing costs.

 

[bigboxes]

Windows 7 is already done. Mainstream support ended in January. Extended support until 2020.

http://windows.microsoft.com/en-ca/windows/lifecycle

You mean that they've worked the bugs out of Win7. They keep on supporting Win7 until 2020 for security reasons. Even then, they'll probably push back that date out of pressure from their OEM customers.

 

[ShintaiDK]

You mean that they've worked the bugs out of Win7. They keep on supporting Win7 until 2020 for security reasons. Even then, they'll probably push back that date out of pressure from their OEM customers.

Windows 7 is still full of bugs. And no new features will be added either.

Support wise its really hard to justify Windows 7.

We can then argue about what OS version we like and we dont. But there is also a reality to take into account.

 

[BUnit1701]

We have a 5 year cycle. Can't say which company I worked (it is one of the largest employer in the us)for but a few years ago we went for 3 years to 5 years tech refresh. Even now, majority of the engineer refused the upgrade because the negligible gain for day to day email, web, Microsoft stuff ran just fine on the old core duo. It was more trouble moving all our stuff onto a new computer. Finally SSHD tempted more people to upgrade, but that had nothing to do with the cpu

Its generally not a question of performance, but of cost. It is expensive both in parts and labor to maintain systems that are no longer in warranty. The places I have worked, IT has maintained a policy refusing to support hardware that is out of warranty.

 

[Compman55]

Can you guy imagine if thehardware industry freezes right now and forced software developers to actually write efficient coding? Look at a core 2 duo runing 2GB ram on windows vista, vs that same platform running windows 8.1 or 10. The difference is mind blowing.

Software needs to catch up to hardware.

 

[kimmel]

Can you guy imagine if thehardware industry freezes right now and forced software developers to actually write efficient coding? Look at a core 2 duo runing 2GB ram on windows vista, vs that same platform running windows 8.1 or 10. The difference is mind blowing.

Software needs to catch up to hardware.

I'm not sure what you mean by your comment as a core 2 duo running win 8.1 or 10 is much nicer than vista from a performance standpoint. The focus has been on low power, less performance optimization ever since Apple released the iphone.

 

[TheELF]

Can you guy imagine if thehardware industry freezes right now and forced software developers to actually write efficient coding?

When did you ever see an ad for a program or a game that said
Now 10% faster
(with latest skylake processors only,your mileage may vary)

They just add more features (and a prettier interface) and sell it to you for more.

 

[Cerb]

Fleets of PCs get more problem-prone over time. Typical warranties are 3 years. Past about 5 years, it gets to be a better choice to replace rather than fix, due to the excessive human time spent dealing with minor issues.

Home users are already on that kind of upgrade cycle, if they take care of their PCs, and don't too anything too stressful.

Intel has plenty of time, and plenty of money. Every other semi company that mainly deals with processors is facing all the same challenges, but they are all in worse financial shape than Intel, with fewer paths to try, and greater risks of whole-company failure when (not if!) they screw up.

 

[Cerb]

Can you guy imagine if thehardware industry freezes right now and forced software developers to actually write efficient coding? Look at a core 2 duo runing 2GB ram on windows vista, vs that same platform running windows 8.1 or 10. The difference is mind blowing.

Software needs to catch up to hardware.

You mean like how 7 and 8 blow away Vista on low-end machines, with installation defaults? Probably 10, too, but I haven't used it much, and not on anything too old.

Do you want to give up features? If not, then you don't want much more efficient software. It's all a balance between development time and cost, support costs, and requirements.

Modern software is far more reliable than it used to be, and those costs cannot be gotten around. They can be moved, but removed. Past that, however, efficiency has been increasing, not decreasing. FI, Web apps can do all they can today because the heaviest code is being compiled to machine code. What used to be written in interpreter engines is now written for JIT systems. Popular interpreters have JIT or AOT options available, if you'd benefit from them. It may not be the efficiency you so want, but it is the efficiency that the rest of us want, so that our software can remain high quality, while not getting slower as we use more energy-efficient hardware.

 

[Cerb]

That makes sense to me.

And maybe in some cases the Linux console I discussed in post #10 becomes this home server with the addition of more cores and/or larger iGPU.

No. That's been "the future" for 40 years. The reality is that fat clients keep getting better, and servers are inefficient for small-scale tasks (too much reliance on a high-latency network infrastructure, which just shifts the load).

Once Atom-like and ARM CPUs exceed Core 2 IPC (in real world use), which is probably going to be in another 5 years or so, SFF PCs will be financially viable for any desktop user, and convertible tablets that are worth using for everything will be cheap.

You can build an ARM desktop for $70-100 that you can do software development on (RPi 2, fully encased and powered, with a sizable and fast SD card). For $150, you can get some that leave the RPi in the dust.

The new battlefield is in the cloud, and the application of the future is machine learning. It's an inflection point and those are dangerous to incumbents who have the most to lose and are often slow to react due to complacency or for fear of alienating their installed base.

Except for businesses that have learned from the passing fads in the past, and know better.

Offline hard copies are always valuable. Anything gained by the machine that can not have its value, including how the conclusions were reached, analyzed by multiple people, is ultimately not worth even storing; yet, making it to that point requires a great deal of human effort. Lots of fads and hand-waiving, but little math and experience. Cloud infrastructure outages not only happen, but can be very costly; it can also be costly for known ongoing processing needs.

The reality is that the cheap small junk is growing, server needs are growing (in part due to being more affordable than ever, including, but not limited to, IaaS), and everything else is stable in use, thus shrinking in sales. Within 10 years or so, with no disruptive changes, it'll be like automobiles, with only artificial growth markets, that still boost and sag with the economy at large.