Earlier in the week came distressing news from many manufacturers of PC components and now Intel has made their financial state a little clearer. The Register has posted the numbers, predicted earnings for Q1 of this year have dropped from USD13.7 billion +/- $500 million, down to USD12.8bn +/- $300 million. Losing about a billion dollars in profit is going to hurt anyone, even the mighty Intel. The drop in the PC market comes from a variety of sources but two of the most likely candidates are the lack of cash in consumers pockets to upgrade and a lack of competition driving an urge to upgrade. Once many gamers would willing live on ramen noodles for a time so that they could afford the next GPU or CPU upgrade thanks to the impressive performance increases the next generation offered. Now new releases tend to offer a small incremental performance increase and occasionally new features which are impressive but nowhere near what an upgrade 10 years ago offered. Certainly part of the issue is the difficult of coaxing a bit more performance out of silicon and with the reduced competition it is less financially attractive to fund expensive and risky R&D projects than it is to work on small incremental increases in efficiency and performance.
Here's hoping for a change to this market in the coming years.
"Intel has lowered its revenue forecast for the first quarter of its fiscal 2015 by nearly a billion dollars, citing a weaker than expected PC market."
Here is some more Tech News from around the web:
- IBM Reported To Be Developing Blockchain-Based Currency Transaction System @ Slashdot
- Linksys EA9200 Tri-Band Router Review @ Hardware Canucks
- This isn't Net Neutrality. This is Net Google. This is Net Netflix – the FCC's new masters @ The Register
“The Register has posted the
“The Register has posted the numbers, predicted earnings for Q1 of this year have dropped from USD13.7 billion +/- $500 million, down to USD12.8bn +/- $300 million. Loosing about a billion dollars in profit is going to hurt anyone, even the mighty Intel.”
First, earnings in the quote from the register should be revenues, or sales, as earnings is profit or income. Second, profit in your statement, should again be revenues, or sales, and should read something to the effect of “$1 bill less revenues than originally expeted,” but you were relying on the garbage in, so logically, to no fault of your own, you had garbage out.
I think most will agree that
I think most will agree that the last few years have been interested but maybe not as much when compared to the past in which a generation upgrade in CPU or GPU tech would yield a much greater increase in performance.
These ~10% and (if you’re lucky) ~20% increases in performance and lack of competition are the problem.
In the past going from GeForce 6 6800 Ultra to 7 7800 GTX was about a doubling of performance, similar with Intel Pentium D to the Core 2 Duo’s.
Thankfully the GPU market still seems to offer pretty good performance increases when you jump to a new generation, but even there you are seeing GPU designs spanning multiple generations more and more, which is part due to the increasing complex designs of modern GPU’s.
I am hoping that the push for 4K and higher quality games (in part due to consoles), will lead to more performance and more competition in the future.
I don’t think lack of
I don’t think lack of competition is the problem. We haven’t had much gpu scaling for the last few years since they have been stuck at 28 nm. Also, we are not getting as much of a power reduction for process shrinks leading to power limitations. Graphics is still a massively parallel problem, so if you can throw twice as many transistors at it, you can get near twice the performance. We should see a bigger jump with the upcoming switch to 20 nm, but not as big as in the past due to power limitations and the fact that they were stuck at 28 nm for so long, they had to push the design more. We should get a boost from new memory tech, although this may not be as spectacular as expected since so much had to be done to reduce bandwidth usage.
With CPUs, throwing more transistors at single threaded performance just waste lots of power with little performance gain. That is just the nature of the problem. Anyway, we haven’t really needed much more single threaded performance. Most consumer applications which are used to test cpu performance are either increasingly irrelevant (web browsing is fast enough) or are being switched to run on gpus (video encode/decode, etc). DX12 should reduce cpu overhead and, hopefully, put some more of the cpu cores to work. I suspect if you have a quad core cpu (like an i7 920), you will be able to play most games with a good graphics card for a while yet.
The market for PC/laptop
The market for PC/laptop devices is mostly saturated, with little growth prospects. I have 4 laptops of various Intel core i series generations, all purchased within the last 5 years, year old models(purchased new but on sale). I have been able to save many dollars not buying Intel’s latest and greatest, and Intel’s overall trend is towards dual core i7 U SKUs, and the M SKUs. For those looking for the higher power quad core i7 series based regular form factor laptops, Intel is not improving these regular laptop SKUs much generation to generation, and those that really game are still getting laptops with discrete GPUs. I’ll stick with last years, or older laptop SKUs that are new, and remain on the bargain pricing shelves. There are many great deals for laptops with quad core i7 CPUs, paired discrete mobile GPUs, that will outperform most/All Ultrabooks, for less price than the Ultrabooks.
Nothing competes with Intel’s current crop of weak processor SKUs like Intel’s older model processor SKUs paired with discrete graphics! Intel’s, Apple Envy, Ultrabook obsession has flooded the market with unwanted U series core i series SKUs, that no serious laptop user/gamer wants.
The Sandybridge SKUs were the last meaningful single year’s overall improvement that justified my replacing a laptop with a leading edge offering. Intel still will not invest in much graphics improvements over the competing discrete mobile offerings, and for sure the Intel GPUs lack the SP/other execution resources that the discrete mobile offerings have. The discrete mobile GPU offerings’ higher SP/tessellation/other execution units counts make the discrete mobile SKUs better for other Non gaming graphics workloads, and not only just the gaming oriented.
Yes, this. Thank you.
Yes, this. Thank you.
Those 5% yearly IPC
Those 5% yearly IPC improvements at best are really selling them chips.
The fact that alot of us not
The fact that alot of us not using the dollar have seen huge price spikes for everything that is sold in dollars does not help. Where I am the dollar rate has increased by 35% since september. Price hikes like that are not good for peoples willingness to upgrade their tech.
I think its also value over
I think its also value over cost of upgrade the fact that were only getting 10% increase but there charging as if we are getting double.
Ive had my firstname.lastname@example.org for over
Ive had my email@example.com for over 3 years now. This is the longest ive kept a CPU, MOBO, Ram. Miss those exciting days of upgrading every year and seeing a difference. Maybe next year!
I was never a fan of Intel’s
I was never a fan of Intel’s tick tock strategy. Putting stuff out for the sake of putting it out is wasteful and rushed.
Skylake is gonna kick ass and I’m waiting til then to upgrade. Currently still happy with my 2500K as are a lot of people.
I’ll be waiting on skylake-E
I’ll be waiting on skylake-E as I’m sure it’ll be the one that introduces nvme natively least I’m hoping.
there’s only so much Intel
there’s only so much Intel can milk with its 5/10% performance every gen. And nVidia`s turn is next, as the sheep realize they’ve been had and stop upgrading for 5/10% (GTX 980) and same performance as last year GTX 970 = R9 290/X
It also shows how much AMD is
It also shows how much AMD is needed in the GPU and especially in the CPU world, no matter the fanboys say.