KitGuru KitGuru.net – Tech News | Hardware News | Hardware Reviews | IOS | Mobile | Gaming | Graphics Cards
Earlier this year, we got our first look at AMD's Summit Ridge/Zen 8-core desktop processor thanks to an engineering sample leak. However, the SKUs floating around at the time were still a bit off from being the final consumer spec. However, as we get closer to the Q1 2017 launch, new engineering samples have been sent out and this week, two new SKUs were apparently spotted.
The previous engineering sample leak detailed a quad-core and an octa-core SKU, with both running at around 2.8GHz with a 3.2GHz boost clock. The quad-core chip ran at a lower 65W TDP while the eight-core variant runs at a higher 95W TDP.
Two new samples have apparently been spotted, according to a curious poster over on the Anandtech forum. According to this post, AMD's new 8-core sample comes with AMD's own implementation of Hyper-Threading, along with a 3150MHz clock speed, which boosts up to 3300MHz.
The new quad-core sample still runs at its 65W TDP but the base clock has been bumped to 2.9GHz with a 3.1GHz boost clock. The final tidbit from the post claims that retail AM4 motherboards will have retail availability by February, hinting that we may see Zen launch sooner in the quarter, rather than later.
It is worth noting that the source of this information is unverified, so take it all with a grain of salt. This is deep rumour territory, so don't take it all as hard fact.
Discuss on our Facebook page, HERE.
KitGuru Says: It is always exciting to hear more about AMD's upcoming Zen chips, after all, it has been a while since we've had a major shake up in the CPU market. Are any of you planning on building a new system when AMD's Zen processors come out?
Unless they new Zen CPU’s are faster clock for clock than a Intel CPU I would hope they will be releasing faster clocked 8 cores and quads than what is leaked here. Yes I know MHz don’t mean very much so lets hope the new cpu’s can match the Intel units or beat them at the same clock speed. I want to upgrade my IVY i5 K CPU in the spring but if AMD can compete it should force Intel to lower their prices or better yet get a Zen at a lower price point but they are a unproven CPU at this point so going Intel is the safer bet for me and gaming on them.
If AMDs demo is to be believed then Zen is slightly faster than Intel i7 6900K.
step one never believe first party results step two find a impartial reviewer. i am excited for amd to make a come back so i hope zen is amazing or on par with intel. even if they are 5% behind some people will still buy them.
I’ll wait for final reviews of the chip but if the specs of engineering samples are correct and like it was mentioned on one another sitr one core boost is 3,6 than with lower TDP this chip clocks similar to 6900K which is great. But until this all is confirmed i’ll take this all with grain of salt
That’s why I said “to be believed.” I need a new CPU upgrade but I’m waiting for Zen and Kaby Lake to see which one is the better CPU before I decide.
It’s 3.6 for single core only and 3200 for multi.
Did you read my comment at all??
It doesn’t clock similarly to the 6900K. The 6900K can maintain the full 3.6 boost on all cores.
Ok. I thought it is 3.2. For single core i know its 3.7
But, 3.3 GHz on engineering sample is not far behind. With rated TDP of 95W. But, thruth to be told i expected 4 core sample to be clocked much higher than 3.1. We’ll see what is what when samples will be put through its paces
3,15 and 3,3 huh?
Well they should at least have 3,2 and 3,4 for the final version. But I really want even higher clock speeds. Boost clock should be 3,5 or higher.
Building a Zen system? …. Yup, when there is an APU version available
But see, they’re engineering samples, their only purpose is to test out and certify software and hardware combinations. Their clockspeed doesn’t matter so much apart from a real-world case scenario. I don’t know how anyone can use those specs to mean anything in retail.
Remember this:
http://wccftech.com/intel-haswell-engineering-sample-benchmarks-leaked/
Back in the day, someone got a hold of an Haswell ES quad-core, it had a clock of 2.8GHz. To compare against Ivy Bridge, a 3770K had to be downclocked for an IPC test. But the ES has no reference to retail parts that ended up debuting much faster, therefore you can’t see the low clock of the Zen ES and think it means anything towards retail capability.
“step two find a impartial reviewer.”
That’s impossible, bias is everywhere. True objectivity requires either not taking a side or not making it part of a conversation– which is difficult with hardware reviews since the website has no choice but to pay attention to the types of readership they tend to receive and honor thier status quo.
Meaning, if enough readership at a particular website expects Zen to fail, then is what the website will put out (i.e. CPU-intensive tests and those that have historically benefited Intel’s featured extentions); hardly impartial. AMD bias is pretty obvious, they go after GPU-intensive to show the CPU doesn’t have a big role anyway, so might as well choose a cheaper CPU, which I’ll bet AMD will go for value against Intel.
I don’t think AMD’s marketing department will maintain a wolf in sheep’s clothing and tone down their Zen capability against Broadwell-E during the recent demo while so-called independent reviews discover even greater gain that AMD hasn’t capitalized yet. Therefore, don’t kid yourself waiting on Zen since no doubt Kaby Lake is faster than Broadwell in IPC while debuting 7700K up at 4.2GHz w/4.5GHz boost. A Summit Ridge quad w/HT would have to debut faster just to match, and that isn’t happening in the first revision since it is way too high a climb from the ES.
If you’re about performance, your mind is already made up; but if you think in terms of per unit dollar or whatever is your currency, then wait for Zen.
You can’t judge what clock speeds AMD will actually ship, or how available they will be, from these engineering samples. But at this point the architecture should be pretty much set, so you can judge things like how much performance you will get at a given clock speed, and they look very good by that measure.
The benchmarks seem all over the place. Some released benches show it is equal to broadwell. Other benchmarks (such as leaked GeekBench, and bench comparisons with Ivy Bridge), show it is far behind Broadwell, and is slower than Sandy Bridge in some cases.
engineering samples are almost always brute-force low clocks and high voltages just so they run stable, they can also not have a configuration at all and require a manual entry in the BIOS.
Retail models should reflect the same clocks as the existing FX-43xx’s and FX-83xx’s, reference clock should be at least 3.5GHz on the high-ends.
Kaby lake will have better ipc but the zen cpu will probably be cheaper.
3.2GHz isn’t a whole lot of good really. Their old chips ran at up to 5GHz stock, which is more than 40% higher than 3.2GHz, so even with 40% higher IPC, we still won’t have got anywhere with performance…. right?
AMD will probably score terribly in any benchmark where the code was generated by an Intel compiler. Intel is known to stack the deck in their favor that way, and has been doing it for years. Here are some links on the subject:
http://www.extremetech.com/computing/193480-intel-finally-agrees-to-pay-15-to-pentium-4-owners-over-amd-athlon-benchmarking-shenanigans
https://www.reddit.com/r/pcmasterrace/comments/1xba7o/why_i_am_starting_to_doubt_cpu_benchmarks_yes_amd/
http://www.agner.org/optimize/blog/read.php?i=49
The AMD demonstration used Blender, an open source application that is normally built with GCC, an open source compiler that has code generation that is more equal in quality for both brands of CPU. (But they didn’t say which set of optimizations the code was built with…)
That would make Zen on par with Broadwell, yet there are plenty of folks that have the dumb hope of matching or beating Kaby Lake. That Blender demo was all the evidence we needed, but people are setting their hopes to unrealistic levels and set themselves up to hate Zen.
Zen was never going to catch up to Intel, that’s not the point; the business definition of competition involves taking market share, which has nothing to do with being the new champion. Intel’s dominance is from maintaining their OEM contracts since most sales isn’t coming from DIY enthusiast sales.
Are you saying Geekbench is code generated by Intel compiler? Because AMD’s Zen benchmark did terrible in Geekbench single threading benchmarks. A Zen cpu @ 1.44ghz gets 984 in this Geekbench single core. That’s pretty bad if it’s legitimate.
http://www.redgamingtech.com/amd-zen-based-naples-specs-benchmarks-leaks-32-cores-512mb-l3-cache/
Here is the other Geekbench 4 article. I’m hoping these scores aren’t true.
https://www.pcper.com/news/Processors/Leaked-Geekbench-Results-AMD-Zen-Performance-Extrapolated
I don’t know what compiler the Geekbench people are using. But it wouldn’t shock me if it’s from Intel. They almost certainly aren’t using GCC, which is probably the most brand-agnostic of the available compilers in terms of code quality.
AMD’s Zen in trouble? Lookit http://carbonite.co.za/f20/amd-zen-141214/index1.html#post972366 Tis one more credible, and explains many stuff seen and heard in the past few months like……
#1, Poor results from leaked benhes…..
#2, Early prototypin mobo complete with test points at ‘Zen demo’ event, bcoz still bein debugged….
#3, Low frequencies with Zen ES includin the one at AMD’s Zen ‘demo’ event…..
#4, Usin open sauce Blender bcoz they can apply software workaround instead of usin perf hamperin bug fix in hw….
#5, Current low end AM4 motherboards doesn’t support Zen yet….
#6, No signs of high end AM4 motherboards with X370 chipset so far….
#7, Digitimes citin ‘after February’ for Zen shipments. May as well be March or even much later….
#8, Su talks bout Zen competin against Core i5 also which isn’t part of HEDT lineup…..