Slashdot Mirror
AMD Outlines Plans For Zen-Based Processors, First Due In 2016
crookedvulture writes: AMD laid out its plans for processors based on its all-new Zen microarchitecture today, promising 40% higher performance-per-clock from from the x86 CPU core. Zen will use simultaneous multithreading to execute two threads per core, and it will be built using "3D" FinFETs. The first chips are due to hit high-end desktops and servers next year. In 2017, Zen will combine with integrated graphics in smaller APUs designed for desktops and notebooks. AMD also plans to produce a high-performance server APU with a "transformational memory architecture" likely similar to the on-package DRAM being developed for the company's discrete graphics processors. This chip could give AMD a credible challenger in the HPC and supercomputing markets—and it could also make its way into laptops and desktops.
I've been hobbling along with my FX-8350 and AM3+ for awhile now and have been wanting to upgrade. If it lives up to the hype, unlike dozer, and piledriver, then I'll definitely get one. Now if only the process actually works....
"An Ye Harm None, Do What Ye Will" Wicca Rede
14nm tech may be the end of the line for CMOS. The 10 nm node that follows may not even be possible
Featuring GGL (Gateless Gate Logic).
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
>40% higher performance-per-clock from from the x86 CPU core.
That could very slightly close the gap between AMD and Intel!
A latent existence
Anyone care to extrapolate from current benchmarks as to how this new processor will compare to Intel's desktop offerings? I would like to see Intel have some competition there.
I go for AMD if I can.
Why?
...gis sdrawkcab (usually not responding to ACs; don't bother posting as AC)
Finally, a processor that will meditate on existential paradoxes and encourage us to push to deeper self-understanding as it half-assedly pretends to work on solving a task.
Arsehole Puckering Up! :P :P :P
I do not fail; I succeed at finding out what does not work.
Note that, in comparison to ARM CPUs, x86 SoCs are *crazy* overpriced. There are superb ARM SoCs for just 20 USD. WTF are you doing selling similar consumer-grade chips for 100 USD??
Intel develops technology which doesn't doesn't make it into their plant for 5 to 10 years. Also they don't put things on their roadmap until they've proven possible.
http://www.xbitlabs.com/news/c...
Intel's 2012 roadmap shows 4nm process in 2022. Which means they have a process that has been tested to work, they are just tweaking it to reduce errors and working on the best way to outfit a plant for it. Also costs billions and time to refit a plant.
Microsoft, Apple, Google, Amazon what's the difference? All steal money from devs and control with walled gardens.
Good bang for your buck and ethical business practices. Hell, just the unethical business practices alone should make most informed people stay away from Intel.
"The tree of liberty must be refreshed from time to time with the blood of patriots and tyrants." ~Thomas Jefferson
TSMC has already produced test wafers on 10nm and plan to enter volume production in 2016
http://www.fudzilla.com/news/p...
A lot of people don't really understand that the CPUs are already "fast enough" and that they can include other important issues in their buying decisions.
I mostly buy and recommend AMD because:
1) better price/performance ratio
2) code I write is designed to scale horizontally, which loops back to #1 even when it is a high load service
3) better power/performance ratio on desktops and servers
4) the motherboards are cheaper for the same components, and I hate over-paying even if the motherboard cost is too small a percent of the total system cost to matter very much
Intel does mostly win on laptops due to lack of availability of alternatives.
What is the use case for a high end integrated APU? The current use case is for systems that need modern GPU capabilities, but not high graphics performance. Lots of use cases require something like OpenGL but would never have it maxed out. They are full-featured, they just don't have the number of cores that you want for playing fancy games.
If you want both to be high end, it seems more logical to have discrete graphics. But if you need OpenGL for CAD, WebGL, or some other use, then an APU is indeed "full featured" not "crippled" at all. They support the latest APIs and technologies, just not the latest games at highest settings.
Because it is not like there is another Xen around in the world of servers...
AMD is doing great? In what alternate universe? Their net income for the most recent quarter was -$180 million. That is not what a company that is "doing great" has happen.
So I can get a 64 core machine with 512GB or memory for $8k instead of $80k.
At the low end good enough and dirt cheap pushes towards AMD. In the middle there are niches where braindead developers still don't have a fucking clue in 2015 how to write multi-threaded code so a fast i7 with hardly any cores is the best tool for the job, but that's a diminishing niche as developers start to learn what they should have in 1995.
So you're an Intel troll now, too? Goody for you.
Il n'y a pas de Planet B.
"What is the use case for a high end integrated APU? The current use case is for systems that need modern GPU capabilities, but not high graphics performance."
The reality of the situation is that right now, even at current tech, if AMD used a bit more actual die space, and embedded a couple gigs of eDRAM for the CPU/GPU to share apart from main system memory, the performance increase would be quite huge.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Sorry, but without ECC RAM support these chips are useless.
AMD... please note, the Intel E3-1276v3 has ECC interface on the die (and also has GPU on die)... you need to do this, otherwise the server world is moving permanently to Intel.
*laughter*
Il n'y a pas de Planet B.
Cheaper. AMD always was and is cheaper than comparable Intel parts. Ditto AMD vs nVidia.
For the lower price you lose some performance, especially in the edge cases. But for most tasks, the difference is almost negligible.
P.S. Though the case is different in the gaming, where pretty much the whole market optimizes exclusively for the Intel/nVidia.
All hope abandon ye who enter here.
You lose more than "some performance". They also run hotter and use more power.
Hotter, but not so hot as to cause problems.
N.B. Historically, I actually had more termal problems related to the Intel's than to AMD's CPUs.
But it is true that in the power efficiency department, the Intel i-series are better than the AMDs.
Though it is not a clear cut, if you take the whole system power consumption. Intel CPU + nVidia GPU would draw about the same (or more) power as AMD's comparable integrated CPU/GPU.
All hope abandon ye who enter here.
"Fast Enough" for you maybe. That usually depends on what you are using your computer for. Gaming is a driver. Most common applications I'll concede are not. However I have noticed in certain professional applications that used to be dependent on primarily graphic cards, are more dependent on CPU and system memory than anything else now. All that aside...
I agree, while I have always bought Intel, I have always kept AMD in mind, One of the things that Intel does that is annoying, but probably linked to the fact that their chips are more advanced (and not some conspiracy to bilk more money out of me, adjusts tinfoil hat), is that the damn chipset and pinout keep changing each generation. While upgrading has become harder and less practical over the years, this pretty much drives the nail in the coffin. If EVERY release requires a new MB, you might as well solder the damn thing on. AMD has been at least better in that regard that they tend to support less chipsets (AM2/AM3).
Lastly, I would like to see some actual direction from AMD, rather that try to be everyman. One thing Intel has done the last several releases is to focus on the laptop market, as that is where they see the growth. All their CPU while powerful, take power usage as their primary concern now it seems, and they all have throttling. Before it was that you made a CPU, and then you tried to make a variant of that work in a laptop. Now it seems, you make a laptop cpu and try to make it work in a desktop. As much as people keep tolling about the demise of the desktop, it hasn't happened, nor is it likely to. In addition, the other markets, tablets, phones, etc... have been eating into the laptop market, slowing that growth. I'd like to see AMD actually try some damn strategy and focus on desktop and server chips, rather than try to be baby Intel. The only way their current strategy is going to work is if they come across some crazy technology to beat Intel with, however they are trying to do this while Intel has probably several times the R&D budget they have... so the odds are not in their favor. Anyway they need to do something, as they've been in decline (as far as I have been concerned) for the last decade.
They could even balance their portfolio out by using their graphics division to focus on laptops in integrated video, though the gains have been pretty small over the years.
A 64-bit memory bus is like the slowest available memory bus for GPUs. Minimum 256-bit wide memory bus path is recommended. Many GPUs are gimped specifically because of the low bit-width bus.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
"Zen will use simultaneous multithreading to execute two threads per core"
That is the exact polar opposite of what everyone wants. We want one thread executed on multiple cores. The single core performance of AMD 6 and 8 core CPUs is PATHETIC. Most software still runs on just one core.
I see lots of announcements - not just this one - shouting about their new microarchitectures, how cool they are, the amazing benefits, and so on. But documentation of exactly what the new microarchitecture is, exactly what it does, seems thin-to-non-existent. Maybe I'm not looking in the right place.
All "big" processors nowadays have fancy pipelines, out-of-order execution, branch prediction, multiple cores, and so on. Fine. But how is Zen different from past microarchitectures? What makes it revolutionary?
Details, please.
...laura
I'm currently bidding out a pretty hefty workstation (128GB RAM, RAID 5 disk array, highly parallel workload).
From what I'm seeing, AMD is pretty competitive on price/performance. Our work load is integer heavy, and I can get a dual 16-core 2.8ghz AMD machine for $2500 cheaper than a dual 10 core 3.2ghz Intel machine. Even if you assume the AMD is 20-30% slower per core, it stacks up quite nicely.
The MBs are cheaper, the processors are cheaper, and the registered DDR3 DIMMs are waaaaay cheaper than the DDR4 RDIMMs that Haswells seem to love.
Sam
The APU supports unified access to system memory which should be a big advantage in heterogeneous computing applications but I have not seen it pan out yet.
The old sideport implementations maxed out at DDR3-1333 speeds (10.6 GB/s) so they might as well add another channel of memory. Integrated GDDR5 would only be 40 GB/s and DDR4 system memory about half that so it does not really make sense at this point.
Their recent APUs look pretty neat but are there any applications which take advantage of their heterogeneous computing capability?
Do you think you even said anything? When you say it doesn't "pan out," you mean that my APU systems actually access their memory? Or, you did you just mean, the feature is just an implementation detail and it doesn't include any unicorns farting rainbows into your applications?
Or is it just pure, "not my favorite team" type of fan-commenting?
Is there a specific claim that they make that the chip can't achieve? That is what it would mean if what they did actually "didn't pan out." But that isn't the case, it did pan out; it does do the things they say it does. If you just mean, you haven't personally seen whatever it is people are doing or not doing, well gee, that isn't notable.
When I say that I have not seen it pan out, I mean that I use engineering applications which should be able to take good advantage of heterogeneous computing with a unified address space as implemented in AMD's recent APUs but none of them do and AMD's existing APUs come with a significant cost in single thread performance. If I just wanted video acceleration and processing, then the existing PCIe interface GPUs can do that.