Do you think AMD Zen will put AMD back in the game?

Guess I'm just being optimistic. We as PC users need AMD to do good. It will drive the industry forward. I hate the fact that my 4 year old 2600k is still very much relevant when compared to the coming Intel CPU's.

 

I'm more a realist, 9 years worth of sub par junk in a row on the cpu side. AMD seeming being happy just trying to match nvidia. They need to have a huge influx of r and d money and competent people in all areas, starting with their marketing team imo.

 

I think Zen will compete with all but the highest end i7's. I expect it to beat the i5's.

 

i5 of what cpu socket? There are many.

 

I would say yes, it will rival Broadwell and Skylake.. Even E series, if its really 8 core 16thread cpu.


btw Zen is 40% faster then Excavator, not Bulldozer.
And Excavator is what ~10-20% faster then Bulldozer /Vishera by raw core thread performance.


http://www.anandtech.com/show/9231/amds-20162017-x86-roadmap-zen-is-in

http://www.pcworld.com/article/2919601/amd-pivots-back-to-high-performance-computing.html

 

I'm hoping they bring good performance that'll finally make a worthy upgrade from my 2600k however I wouldn't buy in to the hype that they'll be monster cpu's that'll leave intel in the dust. I mean wasn't bulldozer supposed to knock intel off the throne? Yet they barely made a dent to even the early sandy bridge CPU's, now haswell CPU's are something like 40% more effective while having under half the TDP as AMD. I mean even if they have improved the IPC by 40% since their latest offerings, it would only mean the performance would be comparable to haswell when skylake is a year old and cannonlake (10 nm) is on the horizon.

 

With a reported 45% increase in performance over the FX8350 I can see Zen being very competitive indeed.

For instance,

- They are using actual cores now not integer
- They are targeting power consumption
- They will have a new chipset FINALLY
- They will have native PCIE3, USB3, SATA3, and finally DDR4 support at which point DDR4 will be cheaper and probably have higher speeds than it does now
- They also will be moving to 22nm process or if rumors are to be believed 16nm with a potential buyout/merger from Microsoft or Samsung.

If they can nail the price down, get the stock cooler right, release it with decent solder between the IHS and the die then they could have a really good CPU platform on their hands here.

 

Don't for get that there is already Zen+ mentioned for a year after Zen release, which is 10 percent faster (at least) above Zen again. So, if Zen is 45 percent faster than Excavator, which is 20 percent faster than Bulldozer, you have Zen being 74 percent faster than Bulldozer (not Excavator), and Zen+ being 91 percent faster than Bulldozer. Interestingly these are the lower estimations, seeing as it is not taking into account the utilisation of various instructions like AVX2 etc.

Why 91 percent? Well, the percentage is compounded. So, Excavator is 20 percent faster than Bulldozer, so 120 percent Bulldozer speed, Zen is 45 percent faster (45 percent on 120 percent, not 100 percent), then 10 percent on that result.

So, Zen+ could well be almost twice as fast, if not more, than Bulldozer, without the help of various instructions. Benchmarks of processors usually make use of these, so in reality due to the extra instructions available, and the improved performance of these instructions (such as AMD resolving the AVX slowdowns etc), you're possibly looking at much more impressive figures than the 91 percent would suggest.

Keep in mind that Zen FX is a true 8 core processor, so possibly around at least 3.5 times as fast (if not more)...

 

If AMD screws up like they usually do, Zen won't do anything for them. They have to release on time, without over-hyping. 2 things AMD has consistently failed at. They've never actually launched a product on time and they always over-hype their products to the point of disappointing nearly everyone. These 2 things are doing more damage to AMD than their meager budget is.

 

That's my take on the matter, hope I am wrong. I would get rid of that marketing team. They said nvidia gpus didn't have a chance against the fury x. Not to forget the bulldozer hype that was a total letdown.

 

This is sadly the truth. You would think that 12 year olds run the PR department at AMD.

 

Marketing and execution are critical for a company in AMD's position. It's generally better to stay quiet about a product's performance and specifications than it is to over-hype said product.

We all known Zen is an entirely new architecture. AMD is moving in a completely different direction now. That's great for AMD. However, if Zen is launched late, AMD is in trouble. Either they need to stop setting launch dates, or they need to set realistic dates and stick to them. If a product is expected to launch in June, people will prepare for a June launch. When it doesn't launch, people start looking to the competition. Every delayed launch turns into lost sales. When you're already bleeding funds, lost sales are deadly.

Over-hyping products results in customers losing faith in the company. Lost faith results in lost sales. AMD's marketing team needs to learn to be honest about the products they are marketing. If they're not "game changing" products, they shouldn't be marketed as such.

I love AMD, but I won't buy any product that's been over-hyped. I plan my upgrades around launch time-frames. If a product launch gets delayed, I generally won't be buying the product. This goes for Intel as well. Since I don't spend much time on my computer these days, launch schedules and marketing are becoming even more important for getting my money. I can hold out with my 2600K and 970 for a few more years rather easily. Even my Athlon 5350 is sufficient for all my needs these days. For AMD to get any further money out of me, Zen will have to be the amazing product they're already hyping it up as. I was planning to buy a 390 or 390X, but they failed to live up to the hype.

 

Maybe of interest to some.....

AMD confirms completing 2 14/16nm FinFET designs - http://*************/amd-confirms-taping-finfet-chips/

or

http://www.kitguru.net/components/g...-we-have-taped-out-our-first-finfet-products/

 

This is looking good:

Quote " Tape-out is the final stage of the design cycle of an integrated circuit, the point at which the artwork of the IC is sent to a maker of photomasks. Once the set of photolithographic masks is ready and verified, it is sent to a contract manufacturer of the chip, which produces the first working samples of the chip. It may take up to several weeks to prepare a mask-set. Production cycle of a complex FinFET processor is around 90 days from wafer start to chip delivery. As a result, if AMD taped out its first FinFET chips in June, then the company will get the first samples of its products in September.
Mass production of chips nowadays starts between nine and twelve months after the initial tape-out. Therefore, if AMD managed to tape-out its chips last month, then it is on-track to start their high-volume production next June or a bit earlier and release its first products made using a FinFET process technology in late Q3 or early Q4 2016."

So this september we will see some real tests with Zen? This is not bad.

I might have to wait for these numbers before I build my 5820K setup.If it's looking good then AMD for me next year.

 

yep

 

The first Zen's will be very interesting with a 4-core 8 hyperthread APU and a true 8 core CPU, 16 thread CPU. Keep in mind there is already a Zen+ for the following year with at least another 10 percent performance boost on top of Zen. Cumulative that would be 54 percent IPC boost.

 

Problem is, AMD seems to fail at delivering on promises....or on time. It doesn't matter what AMD promises, if the final product falls short and launches late. It could be the fastest processor on the planet based on it's expected launch time, but if it comes months or years late....it doesn't matter.

 

Simple answer...

no.

 

I think we'll just have to wait and see. But If I were to look at the last 9 years and then at AMD's PR team I would say no. Look at the Fury X launch, they marketed that card as a Titan X killer, yet it barely competes with the 980 Ti. If AMD works on perfecting their products rather than concentrating on the childish PR they seem to conjure up as of recent, then they just might be able to pull it off and bring some competition. As much as I would like to upgrade my system for the improvements the new Intel chipsets offer over the 7 series, I may just hold off until Zen arrives and we see some reviews. As of right now there is no reason for me to even think about purchasing skylake over my current CPU.

 

Intel seem to be struggling with the current tech process. Cannonlake has been pushed back to at least the end of 2017, if not 2018. Cannonlake is/was going to be the first processor on the 10 nm node. originally Intel planned to take the current technology through to 5 nm, but it seems that they have hit a bit of a brick wall. Currently the limitation is expected to be 7 nm, which may be the case, however it may only be suitable for low power, low end devices.

Basically, as it stands, Intel is effectively having troubles going to anything less than 14 nm. Instead of Cannonlake on 10 nm end of next year, we will have the new Kaby Lake on 14 nm. Kaby lake may not be a huge boost over Skylake, seeing as it is a new addition and they are unlikely to incorporate the initial design of Cannonlake for it, except at 14 nm (if that makes sense). Kaby Lake may brink new instructions though.

AMD Zen therefore is probably a worthy competitor to Intel. It too being at 14 nm it will presumably be able to compete on power, and AMD seem to be incorporating the extended AVX-512 instructions as well (not just the Foundation). So instruction set wise, Intel and AMD will probably be quite equal.

Cannonlake is now expected to be a direct 10 nm die shrink of Kaby Lake, however by then Zen+ is likely to be out by then. In any case, I do believe Zen and Zen+ will be worthy competitors.

Even if the AMD Zen is significantly faster, they probably won't grab a huge amount of market share from Intel. This is going on history with the original socket 478 Pentium IV, and the original Celerons without L2 cache, compared to cheaper and faster AMD solutons in the Athlon and Athlon 64 at the time.

Intel could bring out an atrociously poor CPU like they did with the original socket 478 Pentium IV (even the Socket 775 ones weren't overly flash), and AMD bring out an excellent CPU in the Zen, and more than likely people will still by the Intel...