The AMD Zen/Ryzen reviews and benchmarks are hitting the web (Ars has a review and a look at the Zen architecture, Tom’s Hardware has a review, and there’s bound to be more), but as always, the one you want is AnandTech’s (they also have an interview with AMD’s CEO):
For over two years the collective AMD vs Intel personal computer battle has been sitting on the edge of its seat. Back in 2014 when AMD first announced it was pursuing an all-new microarchitecture, old hands recalled the days when the battle between AMD and Intel was fun to be a part of, and users were happy that the competition led to innovation: not soon after, the Core microarchitecture became the dominant force in modern personal computing today. Through the various press release cycles from AMD stemming from that original Zen announcement, the industry is in a whipped frenzy waiting to see if AMD, through rehiring guru Jim Killer and laying the foundations of a wide and deep processor team for the next decade, can hold the incumbent to account. With AMD’s first use of a 14nm FinFET node on CPUs, today is the day Zen hits the shelves and benchmark results can be published: Game On!
Gaming performance seems to lag behind Intel, while for workstation tasks, it has them beat. For me, an upgrade to Ryzen from my i5-4440 would amount to a total sum of about ^a‘not900 (processor, motherboard, RAM, and cooling), so I’m going to wait it out for now – especially since gaming is what my processor is most used for. That being said – give it a year, and Ryzen will be up there on all fronts with Intel’s best, but at a lower price point.
AMD is definitely back, and I’m very excited to see what competition will bring to the market.
Pure Gaming – a 7700k is a better deal, has better performance, and there won’t be any teething pains dealing with a new platform/chipset/etc.
Gaming/Streaming (with 1 PC) – Probably a good pick. Having more cores should definitely help a lot. If you already use a 2nd PC for broadcasting (which is all around better imo), I would stick with Intel.
Video Editing/Media Creation/Multi-threaded workloads – No contest, definitely the cheapest way to 8 core bliss. Intel is going to have to price cut soon to compete.
I think things will get more interesting when the 4 core versions release, I don’t think AMD will be able to take the performance crown, but they could definitely hit Intel in the margins and release some “good enough” options for way less money.
Anyway, lots of rumors about the possibility of Ryzen based iMacs this year or next year (just google “Ryzen iMac”)
Kinda hope those rumors pan out – an 8 core/16 thread iMac with an MXM based GTX 1070/1080 would be pretty killer – as an option at least.
Depends. TODAY the 7700 might be a bit better(but Ryzen is fast enough anyways), but a year from now more modern game engines will use more threads and the 7700 will be beaten.
Long term Ryzen is an excellent investment.
I like how all the reviews can’t recommend it for gaming because of the odd 1080p numbers, even though no one could tell the difference in real world use.
Most likely the 1080p numbers are off because of a windows driver issue.
They should benchmark it on linux and see how big of a difference there is on that platform.
I don’t think that is why they are not recommending it, it is just an interesting data point. If you are doing 1080p gaming there are many cheaper mid-range CPUs that perform well enough that it almost doesn’t matter.
I think the why is really simple. It is slower, clock-for-clock than Kaby Lake. Not by a lot (10%-20%), but pretty consistently across the board.
Kaby Lake also clocks faster (right now). So unless you need 8 cores, a 7700k is a better deal (faster and cheaper). If you want or need 8 cores though, AMD wins handily on pricing (that and currently there is no cheap 8 core Kaby Lake – performance vs Broadwell is closer to parity).
From the benchmarks I’ve seen so far, this thing is looking like a no-brainer for developers… while not quite as Intel for outright speed, having more parallelism for builds and tests is always nice.
Agreed. I mostly do web development – not a whole lot of stuff for me where heavy multi-threading really pays off, but if I had to do big compiles/builds I would be all over this.
http://www.phoronix.com/scan.php?page=article&item=amd-ryzen-gaming…
http://www.phoronix.com/scan.php?page=article&item=ryzen-1800x-linu…
There are linux benchmarks.
At this stage there are still known to be missing a few key drivers.
Like a new car, a new processor will have teething problems until the compilers catch up. I will be getting my system in the next few months. Already, the GCC and CLANG compilers have a parameter that tunes the compiler output to AMDs architecture. Currently those compilers optimise the code to INTEL. When optimized, AMD will run better and Intel, not so.
There is a middle road in optimizer output. Give the “Must have it now” buyers to get their systems, and then, when price reductions come, as they will, pickup your new system.
THAT is the misery of the Intellectual Property Landscape, right now.
Huh? Intellectual property protection funds billions of R&D for both Intel and AMD. Meanwhile, what happened to all those free OpenSPARC derivatives and to OpenGraphics? Oh yeah, they went nowhere.
Take your two-bit activism to Huffington Post please.
Edited 2017-03-03 09:33 UTC
http://www.pcworld.com/article/3111693/hardware/open-source-25-core…
kurkosdr open graphics was point but opensparc is still way too early to call. Particularly when there are 8000 chip units like this planned. Of course they are using older 32nm chip production not FPGA.
There is going to be a lot more 32nm stuff as universities and other places can afford to have fabrication producing that grade. 2001 is when the 32nm patents are from. Those expire 2021 and IBM has stated they are not going to enforce them against academic. Intel CPU got to 32nm in 2010. So 6-7 years behind in technology. Yet people are still running systems that old because they perform good enough.
So yes there was a very interesting wrinkle in 2016.
I had several reasons to Use 6 cores back at the good times. Now all them would imply IP breaching, Kurkosdr.
You’re only allowed to CONSUME at this Environment. That’s why there is no more fun at Power Computing.
… for a lot of people the question isn’t “Intel or AMD”, but “why upgrade at all?”.
Even as a software developer doing compiled stuff, not web stuff, my current machine is good enough, and neither the latest Ryzen nor the latest Core i7 would give me any speed improvement that i’d notice. And I bought it in 2012 (4C/8T Ivy Bridge).
For gaming you’re better served by a video card upgrade than a CPU upgrade 90% of the time as well.
It’s a bad time to be selling CPUs…
On the other hand, when we’ll need to upgrade for one reason or another, we’ll be happy to find them, they’re offers and at reasonable price, even more if there is competition.
From now on, sorry to Intel folks that do work well, I’ll stick to AMD when I have to buy another PC. Matter of taste mostly.
Sad it lacks haxm, or at least support hardware virtualization under windows regarding Android emulator acceleration.
Likewise. I’m on an FX9350 8 core which is still more than fine for building and testing code. No real urge or need to upgrade. But I’ll definitely bear it or a future revision in mind for when I do upgrade.
Don’t worry, Web browsers and Electron apps will get you to upgrade eventually.
This is the REAL problem. As a developer I suppose I need to be happy because we are going in a future were if you want a software do go faster a good developer is needed you cannot count anymore on waiting a faster cpu.