• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** AMD "Zen" thread (inc AM4/APU discussion) ***

Q1 2017 with a slim chance of Dec '16 release it seems. I really do hope AMD comes thru this time.

Urgh. We all do. :( I costed up a Broadwell-E system earlier. £160 difference between the 6800K and the 6850K for the sake of Intel switching on the extra PCI-E lanes. Want to get to eight from six cores? You jump from £410 to £950. My FX-8350 has eight cores. Sure, they don't have a FPU per core, but then who cares? As software finally catches up with utilizing higher core counts (too late for poor Piledriver / Excavator, but there you go) there are scenarios were the AMD chips are hitting parity with the Intel ones despite the latter's higher IPC. I hope like crazy that Zen is a big success. Don't get me wrong - I'm not naïve enough to think that AMD will sell it cheaply, but they'll at least undercut Intel a small amount which should lead to some competition and less cheap tricks like the 28/40 PCI lanes price jump.

And I hope we finally see a bit more variety in motherboards, too. Why does every OEM see the need to jam four PCI-Ex16 slots on a board? Give me one with just a couple and then a barrage of M.2 sockets and SATA? I'm hoping AM4 boards show greater variation than we have today. I really want Zen to shake things up a bit.
 
Urgh. We all do. :( I costed up a Broadwell-E system earlier. £160 difference between the 6800K and the 6850K for the sake of Intel switching on the extra PCI-E lanes. Want to get to eight from six cores? You jump from £410 to £950. My FX-8350 has eight cores. Sure, they don't have a FPU per core, but then who cares? As software finally catches up with utilizing higher core counts (too late for poor Piledriver / Excavator, but there you go) there are scenarios were the AMD chips are hitting parity with the Intel ones despite the latter's higher IPC. I hope like crazy that Zen is a big success. Don't get me wrong - I'm not naïve enough to think that AMD will sell it cheaply, but they'll at least undercut Intel a small amount which should lead to some competition and less cheap tricks like the 28/40 PCI lanes price jump.

And I hope we finally see a bit more variety in motherboards, too. Why does every OEM see the need to jam four PCI-Ex16 slots on a board? Give me one with just a couple and then a barrage of M.2 sockets and SATA? I'm hoping AM4 boards show greater variation than we have today. I really want Zen to shake things up a bit.

It keeps em honest, which is definitely what's needed. Intel's separation of mainstream and HEDT platforms has never sat well with me. Especially when you consider the mainstream has a more advanced architecture and chipset features than the HEDT platform. It's just odd. But in they're massively dominant position they can easy get away with it. And yeah, i was nearly crying having to shell out £500+ on the 6850K but that's life.
 
There's a good article about Zen here:
https://scalibq.wordpress.com/2016/08/20/amd-zen-a-bit-of-a-deja-vu/

I'll admit the guy is a little biased against AMD but he's been proven to be largely correct over the years unlike the AMD fans who usually attack him.

At any rate, Zen looks more promising than Bulldozer ever did. I think AMD made a wise choice in going back to ‘follow the leader’-mode. Not necessarily because Intel’s architecture is the right one, but because Intel’s architecture is the most widespread one. I have said the same thing about Pentium 4 in the past: the architecture itself was not necessarily as bad as people think. Its biggest disadvantage was that it did not handle code optimized for the P6-architecture very well, and most applications had been developed for P6. If all applications would be recompiled with Pentium 4 optimizations, it would already have made quite a different impression. Let alone if developers actually optimized their code specifically for Pentium 4’s strengths (something we mainly saw with video encoding/decoding and 3D rendering).

Bulldozer was facing a similar problem: it required a different type of software. If Intel couldn’t pull off a big change in software optimization with the Pentium 4, then a smaller player like AMD certainly wouldn’t either. That is the main reason why I never understood Bulldozer.
 
Last edited:
gvzdfg.jpg
 
I'm still not sure what the point of Bristol Ridge is. Is it basically so they can stagger the roll out a new socket and chipset and a brand new CPU architecture rather than doing it all at once? i.e. old architecture + new socket for mainstream, new architecture + new socket for enthusiast, then new architecture + new socket for mainstream.
 
I'm still not sure what the point of Bristol Ridge is. Is it basically so they can stagger the roll out a new socket and chipset and a brand new CPU architecture rather than doing it all at once? i.e. old architecture + new socket for mainstream, new architecture + new socket for enthusiast, then new architecture + new socket for mainstream.

It's a relatively quick and cheap respin of the existing Carrizo APU that reduces power consumption whilst also allowing the platform to move forward and on to DDR4 (that Intel have used for a while).

It gives a chance for motherboard Manufacturers to get some AM4 boards out in preparation for Zen based chips, and allows consumers to have a AM4 board + DDR4 ready, in preparation for a drop in upgrade.
 
It gives a chance for motherboard Manufacturers to get some AM4 boards out in preparation for Zen based chips, and allows consumers to have a AM4 board + DDR4 ready, in preparation for a drop in upgrade.

IIRC, they did/said this back in the Phenom II -> Bulldozer days... Afraid I no longer have the faith to buy a motherboard that can be 'upgraded' with an AMD cpu at an uncertain future date :/

My personal 2p is that it would be a mug's game to buy a new board and an old CPU architecture before Zen is in reviewers' desktops and putting out real benchmarks. I will be beyond delighted if it's competitive with Intel, but I wouldn't put any money into it without knowing what I'm buying and knowing that it exists.
 
I'm still not sure what the point of Bristol Ridge is. Is it basically so they can stagger the roll out a new socket and chipset and a brand new CPU architecture rather than doing it all at once? i.e. old architecture + new socket for mainstream, new architecture + new socket for enthusiast, then new architecture + new socket for mainstream.

What's the point of Skylake if Kabylake is coming out a year later?

Bristol ridge will seemingly be available soon on desktop and with AM4 boards, it will be almost a year till Raven Ridge APUs are available. Why wouldn't AMD have something available for that time if they can do so. Bristol Ridge is surprisingly good for the power and price and it's noticeably more competitive than the apu it's replacing.
 
What's the point of Skylake if Kabylake is coming out a year later?
Well yes, that's a (roughly) year cycle for new releases. In isolation even that seems a bit pointless due to the minor improvements each time but of course OEMs love it and it makes Intel money so whatever.

This is a bit different though because Zen is so close and it's a far bigger jump than any of Intel's new CPUs is these days. It seems as though Bristol Ridge is basically a tinkering platform for AMD to try some stuff on a mature architecture before jumping straight to Raven Bridge. That seems very sensible but at first I couldn't see why anyone would buy it. I guess the fact that it's AM4 means there's a nice upgrade path over the next few years, plus they're only going to OEMs (for now) who can of course make use of the lower power requirements, so they'll shift some units.
 
Back
Top Bottom