Sad news again from AMD as roughly 710 employees from across the globe will be getting severance packages for Christmas. The cuts are likely to come from the Computing and Graphics division as they saw a 16% year-on-year decline in income. The Enterprise, Embedded, and Semi-Custom division saw a 21% increase and were the reason AMD's total income only dropped 2% when compared to this quarter last year. The news for the future is also not good, with The Inquirer reporting that AMD expects its revenues to slide another 10-16% per cent in the next quarter. Perhaps that is part of the reason Lisa Su will take home a salary that is $150K less than what Rory Read was earning.
"Following a grim earnings report on Thursday, AMD has announced a restructuring plan that includes axing seven per cent of its workforce by the end of the year.
The plan will see AMD issuing layoff notices to about 710 employees worldwide, and is expected to cost the chipmaker $57m in severance payment."
Here is some more Tech News from around the web:
- TSMC to start production of 16nm process products in 2Q15 @ DigiTimes
- The New TrueCrypt – VeraCrypt Or CipherShed @ TechARP
- IBM announces Internet of Things cloud services @ The Inquirer
- iPad Air 2 and iPad Mini 3 pre-orders go live at the Apple Store @ The Inquirer
- Devolo dLAN 500 WiFi Network Kit Review @ NikKTech
- Linksys SE4008 @ HardwareHeaven
- Introducing the F*Watch, a Fully Open Electronic Watch @ Hack a Day
This makes me sad. Unless
This makes me sad. Unless they have some great product they are secretly working on, and I hope they are, some new magic core, some game changing technology, something they are just stalling for so they can get it to market, this might be the end. Normally when a company just keeps cutting staff over and over it’s only a matter of time till it dissolves into non-relevance.
Anytime you switch to
Anytime you switch to targeting a different market, you are going to have some staff turn-over. I have to wonder if they are essentially giving up on the x86 market to focus more on ARM based products. I don’t know if ARM ISA is much of an improvement over x86 (really AMD64) ISA. It may have been preferable to have PowerPC ISA everywhere rather than ARM, but Sony and IBM messed that up. At least we aren’t being forced to move to IA-64…
the ARM move just seems like
the ARM move just seems like a weird one to me. It’s already a pretty saturated market and making a socket that can handle 2 different architectures just seems…… weird. Like I don’t understand the point. There probably is some good thinking behind it that is just beyond me but it seems weird. In my mind the biggest push could be to once and for all design the powerhouse apu that everyone would go mad for. It could be huge, power hungry and need the sickest water-cool/hybrid cooler possible and a crazy motherboard and we would still all get wet for a i5 competitor with r9-280m graphics on a single chip. Of course while I have my magic cap on I’d like a pony too, and a hover board, and a hover board for the pony. I (being a classic AMD fan-boy, and a “little guy” fan-boy) just don’t like the way AMD is going from a business survival point of view. It NEEDS to step up or the whole “Do I buy a free-sync or a G-sync monitor” discussion will go to “Do I buy a non gaming monitor or one with G-sync” and a monopoly is never good for consumers, a duopoly is bad enough.
who would even buy AMD at
who would even buy AMD at this point? Apple?
They have even less presence in the phone/tablet segment than INTEL for god’s sake, so it’s not like they are bringing much of anything to the table. The only thing they’ve done is throw insane TDP’s at Intel and Nvidia in a desperate attempt to compete. They should have seen the writing on the wall years ago and focused on mobile, mobile, mobile. Now they’re done.
AMD does not have the money
AMD does not have the money to compete, and to stay alive they have to pick their markets carefully. The Low cost GPU market is one, along with some of the enterprise and custom chip business, where AMD has made some headway. If you expect AMD to outright compete against Intel and Nvidia, perhaps you were not listening to what AMD said a few years ago! AMD’s primary job is to stay alive, and innovate where it can, and plan for the long term, while staying alive to bring the long term goals to fruition. AMD’s APU products will have to do, until they can get their new x86, and custom ARMv8 based SKUs to market in 2016, and until then
do not expect anything but small incremental improvements across AMD’s product lines. Providing Apple with a low cost solution to GPUs is a big plus, as we all know with Apple, top dog performance has never been on Apple’s Radar, but every sale helps. AMD will be trying to give users something to counter Nvidia’s latest, but do not expect AMD to catch up to Nvidia in the power usage category, and AMD will offer a competitive, in GPU performance, SKU while using more power to do so, but the price will also be a bit more affordable. AMD is making some headway in the power usage metric, but without the big money to throw at the problem, the improvements will be incremental.
AMD has some very innovative products, more so on the server side, and the enterprise side with some custom SOC, and server(ARM based) wins. Consumers will have to wait for AMD’s new x86 microarchitecture, as well as AMD’s custom ARMv8 APUs, before AMD will be able attempt to compete, starting in 2016. There will be some product improvements, but they will not be dramatic, more like small updates as they can be provided, and it’s not worth the effort for AMD until the new microarchitecture projects are completed. AMD has a stable customer base, across the enterprise market, but the consumer market has been more difficult. AMD has a server business, and a custom CPU/APU non consumer business, to go along with the consumer side, and AMD will have to make money where they can, in the markets that AMD currently has something to offer, so do not count AMD out, as so many have been doing every year, for so many years now! AMD appears to be very good at barely surviving to try to compete another day.
I don’t really care that much
I don’t really care that much about load power consumption, since I don’t game 24/7 and I don’t mine crypto-currency. AMD still has the better price/performance, so I have no problem going with their video cards on the desktop. I would go for a 980M if buy a laptop though.
Nvidia has more market share and more money than AMD, but this still isn’t much compared to what Intel makes. Once Intel starts putting stacked memory chips on the cpu/gpu package, dedicated graphics will become an even more niche market than it is now. This is a big threat to both AMD and Nvidia. I like AMD’s APU strategy, but this isn’t really much different than Nvidia’s strategy with their Tegra K1; main difference is Nvidia is using a custom ARM core rather than x86. I don’t know if we will see AMD going a similar route with ARM based offerings targeted at tablets. The Tablet and phone markets are quickly becoming saturated, so AMD’s push into network server/appliance devices may be a really smart move. There is still a large market for building out networks and upgrading to lower power devices.
I would prefer to have healthy competition, but there hasn’t been much in the cpu market due to the process tech issues. The gpu market is going a similar way, but this may be remedied soon with other foundries (hopefully) catching up with Intel. Intel has spent a lot of money at trying to break into the mobile market, but it seems like Nvidia has had more success really. The desktop may get some spin-offs from the mobile market, but otherwise it may be stagnant for a while. Maxwell exist, essentially, due to Nvidia’s work on mobile parts; the desktop is not their main focus anymore.
Stacked memory really is not
Stacked memory really is not a big threat to Nvidia, or AMD, as both have their own versions of stacked memory on the way. AMD has its HBM, and Nvidia has stacked memory, and its mezzanine module Pascal project. Nvidia is also developing a mezzanine module product to attach its GPUs to IBM’s in house power8 CPUs for the server/HPC market. Intel’s Graphics leave much do be desired, at too high of a cost, for such few SKUs that even get Intel’s attempt at high performance integrated GPUs. AMD does have a custom ARMv8 ISA based microarchitecture that is expected to be released at around the same time as Keller’s reworked AMD x86 microarchitecture in 2016. AMD is behind both Apple, and Nvidia, in getting a custom microarchitecture that can run the ARMv8 ISA, Apple’s A8/A8X cores can, if the Cyclone 2 microarchitecture has not been enhanced more, execute 6 IPC(at least), per core, while Nvidia’s Denver was presented at the Hot Chips Symposium, as being capable of 7+ IPC, per core. Nvidia’s Denver presentation at the Hot Chips Symposium is worth a read. Nvidia and Apple are currently going head to head, but the jury is still out on the A8X’s GPU, I am hoping the A8X would at least get the PowerVR wizard GPU, just to begin a race to implement Ray Tracing hardware on the GPU among GPU makers, and free graphics application users from any dependency on the very expensive server/workstation CPU SKUs, that are currently needed for Ray interaction calculations.
Do not discount future competition from the Licensed power8 CPUs giving ARM, and x86, based products more competition in the future, both the ARM, and Power8, cores are RISC designs, but the Power8 cores have something that ARM based designs lack, and that is SMT(Simultaneous multithreading) built into the microarchitecture, expect many makers like Samsung, Nvidia, and others to license the base Power8 core design, and adapt it to a wide Varity of PC/Laptop SKUs, beginning in 2015, and because of Power8’s open licensing, even AMD could take the base core Power8 microarchitecture and enter a new market, along with Custom ARM and x86.
I called it back when AMD
I called it back when AMD slides leaked around the time the Radeon HD 7000 should have had a refresh (year after release), where they had no intention of dropping a refresh because the Never Settle Bundle was making them so much money.
They lost market dominance in the discrete market (significantly) and it caused them to play catch up to Nvidia that dropped 2 cards (600/700) after the HD 7000. Whatever intention, or however much money they were making, it just looks bad when you sit on your ass and your competition drops two visible product lines after your last one.
I said it then and I say it now, it’s going to bite them in the ass. I foresee a buyout in AMD’s future sooner rather than later. Only problem is who really wants a CPU/GPU company these days when everyone is trying to get out of it?
this is going to keep
this is going to keep happening as long until 2016 when they finally plan to release their new x86 chips and unless those chips will be made out of magic and be able to compete with intel’s offerings that far surpass amd that may very well be the last x86 architecture they ever release, at this point amd is only part of low end market, they don’t have a competitive product for mid range anymore, and ‘its been years since they had anything in high end market, don’t get me wrong i had a fx8350 machine balls to the wall overclocked @5,2ghz and it just couldn’t keep up in anything that wasn’t heavily multithreaded, the ipc of those chips is terrible, i hope they can get back in game, but in my opinion they are a bit to late for it, i mean two years until it comes out, leaving intel to gain more and more of the market, they should have abandoned bulldozer module design a year a go and started to work on next core architecture that doesn’t use modular designs, atleast then they could give us something sooner and regain more of the lost market share
I still don’t have much hope
I still don’t have much hope for there next-gen x86 “Zen” architecture that like Fusion and Bulldozer, is being hyped up shrouded in secrecy. We’ve heard AMD do this many times in the past, all in a row, that didn’t pan out as they intended.
I hope whatever they drop in 2016 blows things out of the water. Athlon style. However, I just don’t see how they’re going to do it.
Give AMD a few billion
Give AMD a few billion dollars, and they may listen to what you have to say, or just buy that much of AMD, and get a seat on the board.
Keller’s 2016 x86 microarchitecture is supposed to come with SMT, and better single core performance, but do not expect it to outright rival Intel’s product, as the money is not there for that. AMD’s best will have to have better financing, but do expect AMD’s product to be improved enough to have a better showing, and maybe the gaming engine designers will go more into multicore, as the console’s gaming engines have to make use of 8 weak cores, just to make some titles playable on the consoles, and maybe the entire gaming industry, that competes in the gaming console market, will take the experience of multithreading weak CPU cores to get the maximum gaming performance over to the PC gaming engine market, and single threaded performance will not be such a factor going forward, it takes a while for the software to catch up, but once it goes all in with multicore the problem of single threaded performance will not be as severe as it once was. AMD’s x86 will most likely offer enough improvement that AMD will need less of a core count advantage to take on Intel in the midrange, and if the SMT is done right, AMD will have a more power efficient core to compete more in laptops. With a good enough SMT, and single threaded performance metric, 8 cores on a desktop SKU would mean at least 16 threads, and give AMD a shot at the non E series i7’s in performance metrics, it sure would interest me in rendering/ray tracing workloads, as long as the SIMD was there on AMD’s reworked x86 microarchitecture. Intel’s 8 core SKUs are too pricy, but once AMD gets the SMT, and processor threading that goes with SMT, it could attract more of the rendering market, and AMD’s product would be much more affordable.
well they said it years ago
well they said it years ago that they were focusing on apu’s, enterprise, custom and arm, most of that has come true.
unless someone shows us an updated roadmap with a new (not a retread) x86 chip then you will continue to see cuts.
but the last non apu desktop design was released when? 2011? and the last chipset a couple years before that.
the only thing they need to do is stop the money bleed- and readjust the profits lower to meet the market for low cost chips.
but amd is not a desktop maker for the enthusiast anymore
yep A
yep A