>AMD Zen will launch in 2017
How the fuck are AMD going to survive 2016 when locked Skylake parts can be overclocked now?
Is there any reason to buy FX CPUs?
>Is there any reason to buy FX CPUs?
plenty of reasons, I'm sure you can figure one or two reasons out yourself if you think really hard.
>>51820343
>mfw amd dual graphics mode works with fx cpu's and amd gpu's better then nividia optimus
>>51820377
It's unproven
>>51820343
>caring about amemed
>>51820426
he posted anime, he must know what he is talking about
>>51820442
Shit, thanks for warning me anon, forgot about that
>>51820343
It will launch at 2016
>>51820343
dumb frogposter
>>51820462
it's a trap, bridget is /v/ gaymen material
Is there any reason to buy any new CPU?
It has been a long time since I was hyped for any new CPU.
Still using an old Phenom2
>>51820503
So they raided our board? Greentexting OP sentences? Shieet, they're no match for /b/ or even /pol/
>>51820490
Zen hasn't been taped out yet.
>>51820343
>Is there any reason to buy FX CPUs?
intel doesn't receive money
>>51820516
Improvements to the platform.
>Native USB3/3.1
>SATA express and M.2 + NVMe
>PCIe 3
do FX6300 and FX8350 lower Mhz to 1.0Ghz or something like that to lower power consumption on idle?
>>51820516
>Still using an old Phenom2
>>51820565
it's fucking nothing
>>51820516
>being this poooooor
>>51820588
>>51820597
>throwing away a perfectly good CPU just because it isn't the latest shit
kids these days
>>51820613
Well if you only watch your weebshit and shitpost here it might be fine :^)
>>51820516
No. Nehalem was the last big jump in CPU performance, you really don't need anything newer unless you're running SLI/Crossfire. Current CPUs are only like 50% faster clock-for-clock and that's 7 years on.
>>51820625
and actual work
>>51820557
wouldn't surprise me if they have a patent licencing agreement so they both get royalties from each other
>>51820626
>Nehalem was the last big jump in CPU performance
http://www.anandtech.com/show/2658/16
wow what a huge performance improvement over c2q
>>51820516
Do you know what Good CPU means?
Good CPU = Good Central Processing Unit
Good= Crysis 3
So Good CPU means that Good cpu is cpu which can run crysis 3 on ultra
Is your able to? No? then change it
>>51820651
#rekt
>>51820651
Is Crysis even that reliant on the CPU?
Dwarf Fortress should be the CPU benchmark
>>51820664
Yes because CPU's kids are put into GPUs
So if CPU has bad genetics, GPU will be bad too
>>51820516
>>51820626
Skylake has around double performance of Phenom II clock per clock. Upgrading would definitely worth it
http://www.anandtech.com/bench/product/186?vs=1544
>>51820681
The GPU just needs a good Mother-board to succeed.
>>51820343
Intel is botnet, and like only few people /b/ has it.
Because how can PC wake up when you call cortana(or how that bitch is called)? They're sending all info to intel's HQ, and indians go through all that
>>51820707
Well, mother-board is not black, so it needs a good dad too.
If dad is bad, mother and kid won't have good performance
>>51820588
The 1090T and 1100T are two very fucking solid CPUs.
One of the last good CPUs that AMD made.
>>51820650
>>51820705
The 0th generation Core i7 also brought several other improvements over Core 2 Quad, such as HyperThreading and QPI, not to mention a much higher overclock potential. Clock it to 4 GHz and you're laughing. Yes, Skylake is faster (and more expensive) but not necessary.
>>51820546
i keep reading that. what exactly does that mean?
>>51820516
>phenom ii here too
atleast im upgrading this month
feels good knowing already
>>51820343
Nvidia can't have a monopoly, I guess AMD will be on artificial oxygen for a year
>>51820625
I'm still running an i7 950 from around the same time and it still does the job more than well enough (programming, GPGPU computing, writing papers, web browsing, playing vidya, etc.).
>>51820721
I think your confusing Intel and my Microsoft senpai
>>51820557
>giving amd money
Why are we doing this we should just let them die
>>51822206
Probably because then Intel wouldn't have any competition and they would obviously move to screw us over by slowing product development to a crawl and jacking up the prices. They're already doing product development in the area of performance at a snail's pace when AMD can't compete in the performance market. Imagine them doing product development in all segments at the same pace.
>>51820343
AM4 is coming in 2016
>>51821759
>HyperThreading
That was introduced with the Pentium 4. Yeah, it was missing on core 2 but it was not new to the first generation core I series.
>>51821265
They get beaten handily by the athlon 860k in terms of single threaded. It's much closer in multithreaded.
>>51822467
When you have 6 cores sharing mostly the same hardware resources (memory bandwidth, same MMU, etc.) and additional thermal constraints due to additional cores you're not going to get the same per-thread performance as the same chip with 4 cores.
>>51820705
>recommending memelake when there is x99
>>51821878
Means they haven't even finalized the design to be manufactured yet, and won't until 2017.
>>51820587
With cool and quiet it sits at 1.4 ghz idle i think
>>51822186
Intel added the feature to the CPU to be able to listen all the time, thanks to Intel Cortana can wake up whenever you call her
>>51822509
>When you have 6 cores sharing mostly the same hardware resources (memory bandwidth, same MMU, etc.) and additional thermal constraints due to additional cores you're not going to get the same per-thread performance as the same chip with 4 cores.
But it isn't the same chip, the 860k is based on excavator and is much, much newer. It also overclocks better, uses less power and supports newer technologies like PCIe 3.
>>51822671
>Intel added the feature to the CPU to be able to listen all the time, thanks to Intel Cortana can wake up whenever you call her
You're fucking retarded. That's all I have to say to you.
>>51822899
Your retarded
>>51822433
amd is going to release excavator-based(!) cpus using am4 in 2016.
>>51823093
Not a bad upgrade for me actually if it's cheap, USB 3.1 access would be great
>>51823093
>amd is going to release excavator-based(!) cpus using am4 in 2016.
No, they're going to release excavator based APUs in 2016. There's a notable difference.
It looks like they'll be socket compatible with the Zen processors coming out while sporting the improved memory compression used in Carrizo and the r9 380. In fact, I'd go so far as to say it looks like it's just Carrizo retuned for higher TDPs and DDR4, which will be very welcome especially if they launch an Athlon variant for peanuts.
>>51820613
Have fun rendering anything. Shit had a hard time with Solid works sometimes.
>>51820343
Vulkan
>>51823996
>No, they're going to release excavator based APUs in 2016. There's a notable difference.
Source? Weren't desktop APU's stuck on Steamroller until Zen?
>>51823093
AMD Excavator Family 15h is a microarchitecture under development by AMD to succeed Steamroller Family 15h. On October 12, 2011, AMD revealed Excavator to be the code name for the fourth-generation Bulldozer-derived core.
The Excavator-based APU for mainstream applications will be called Carrizo and is scheduled for a 2015 release. [2] [3] The Carrizo APU is designed to be HSA 1.0 compliant. [4] An Excavator-based APU and CPU variant named Toronto for server and enterprise markets will also be available. [5]
Excavator has been confirmed to be AMD's final revision of the 'Bulldozer' family, with two new microarchitectures replacing Excavator a year later. [6][7] The next generation sister architectures will be the x86-64 Zen and AArch64 K12 architectures.[8][9]
so I guess they're just going to release desktop versions of carrizo
>>51827703
I've got no single source on that, just go look up Bristol Ridge. It looks like all the tech sites agree on that bit, at least.
>>51828211
But wouldn't that hold it back because carrizo is a laptop processor?
Any chance of a new fx processor with cmt on am4?
>>51828864
It's meant to be power efficient, but it'll probably be faster overall compared to the current mobile carrizo which has power delivery issues in some laptops
>>51828889
Faster because it's not held back by the shitty 15w tdp set by the OEMs?
>>51828935
and improvements in other places, even on the 35W variants the performance isn't really there because it's not made with performance in mind
>>51828864
It'll probably get a tune up for the higher TDP in the same go as it gets a DDR4 memory controller and the change in pins to fit the AM4 socket.
>>51828935
According to AMDs own benchmarks there's only a 40% difference in performance despite the 35 watt chips have 233% of the TDP to work with. I really don't see a reason to doubt this claim.
>>51828992
I've never been fussed with tdp as long as the stock cooler is enough, I'm waiting for a new APU like the a8 7600, but with ddr4.
>>51828992
You should realize that they have to do more than just raising the tdp to get decent performance, I'm hoping for some changes to the architecture itself to function more like a desktop processor.
>>51820343
What? I thought Zen was Q3/Q4 2016?
What the fuck went wrong?
>>51821265
Just don't try to use a 290 with it. I had to get a new motherboard because the 890 chipset and 290 causes spontaneous system freezing and there's no fix.
Other than that I still liked my 1100T when I replaced it and I think I'll put it in some other build later on.
>>51829142
amd haven't taped out and validated zen yet
will take at least 10+ months
expect a very late 2016 launch
>>51829138
>You should realize that they have to do more than just raising the tdp to get decent performance,
Which is why I intentionally used the ambiguous words "tune up"
>I'm hoping for some changes to the architecture itself to function more like a desktop processor.
Erm, what kind of changes would they be?
why doesn't AMD just borrow 100 billion dollars from Intel and make a new chip with 5x the IPC of Skylake and 1/5 the power draw?
>>51829286
Because Intel spends three billion dollars QUARTERLY on research and development. That's twelve billion per year.
It would take AMD quite a bit of money to get the facilities for such a research team set up. It is also highly unlikely that anyone is willing to give a company that is already in debt a loan that big.
>>51829201
>what kind of changes would they be
A tune up like you said. Increase the performance and slap a bigger gpu on there.
If AMD assumes zero performance gains from ddr4 then this will be a good processor, but if they're relying on ddr4 then it's fucked.
>>51829460
>Increase the performance and slap a bigger gpu on there.
If you look at Carrizo you'll see that there is a large performance boost already. If you look at the Carrizo GPU you'll see that it has the same number of compute units but that they perform drastically better than the ones found in Kaveri.
There are also a fair few new technologies associated with Carrizo, like the colour compression. Even so, this is still a budget part so don't expect insane performance. I expect the a8 variant to offer unrivalled performance for the price but that the a10 will be a waste of money, just like it was with Kaveri. Please note that I say that as a proud owner of a 7850k which I bought for the sole reason of overclocking. It was a stunning success and great fun.
>>51829576
But carrizo is a laptop cpu, I meant in comparison to existing desktop APUs.
If it's just the athlon 5350 all over again then fuck that, I'll wait for zen.
>>51829636
>But carrizo is a laptop cpu, I meant in comparison to existing desktop APUs.
Yes but Carrizo shows huge architectural improvements over Kaveri and Bristol Ridge looks like a modified version of Carrizo for desktops.
>If it's just the athlon 5350 all over again then fuck that, I'll wait for zen.
It'll have to be changed far more than the 5350 was, there's talk of DDR4 support among other things.
>>51820343
>How the fuck are AMD going to survive 2016 when locked Skylake parts can be overclocked now?
How is this an issue?
They sell gfx cards and cpus. The other half of the business will keep them ticking over in 2016.
>>51829768
their gpu business can't shore up the rest of the company though
>>51829707
>carrizo shows huge architectural improvements.
I have no doubts about that, but again, it's currently a 2015 APU so I'm hoping they can improve it even more since am4 is 2016.
And like I said before, if they focus their efforts on improving the APU even more and not just relying on ddr4 to increase performance, then I have no doubts it's going to be a great product.
I'm still running my i7 870, so I'm looking forward to an upgrade to zen. Especially an APU because of the dx12 multiadapter
>>51823001
>Your
>>51829860
>because of the dx12 multiadapter
I wouldn't count on that, from what we've seen so far that looks like it's game specific.
Overall I think AMD are going to keep doing what they're doing; APUs for budget and FX for dedicated. While Kaveri was great for the price (especially if you won the silicon lottery big like I did) and Bristol Ridge will no doubt be more of the same, it'll probably be roflstomped by things twice its price, like any i5 or the zen equivalent.
>>51830100
It might be game specific, but it'll be pretty worth it from what I've seen.
Plus I doubt I'd get a budget APU without the gpu here in aus anyways. No 860k sold here.
