ZEN ES Benchmark from french hardware Magazine

[Vesku]

The fact it only performs 11% behind the 6900K in games with a gimped clock paints a good picture. Hopefully the 4C/8T version can OC to 4.3GHz or so on air, it will probably do very well against the 6700K / 4790K.

Yes, roughly 11% behind in that game bench and 6% lower in clock. Closer to Intel than I was expecting. Now I'm a bit worried we won't see a SR7 SKU below $400.

 

[DrMrLordX]

issue is, multi-gpu support has really gone down the tubes as of late so having 24 or 42 or whatever amount of lanes means nothing if 2 gpus are as good as 1.

I see where you are coming from, it's just that there are other devices that can chew up that bandwidth. I have not seen any kind of in-depth reporting on how PCIe bandwidth limitations would affect NVMe storage performance, so maybe it is a non-issue just as it is with multiGPU setups.

 

[esquared]

Eight of the past 15 reported posts from today have been about this thread.

It is getting perilously close to being locked and/or infractions/vacation given to many of the
people in this thread.

Lighten up everyone.


esquared
Anandtech Forum Director

 

[escrow4]

Quad channel memory makes those boards very expensive, as it requires more layers, which adds a lot of cost. On a side note, I was glad to see amd didn't go that route.

The cpus, aside from the base 6 core, is the real ripoff. There is a good $500 intel monopoly tax on the 6900k.

What expense. The Asrock X99 Extreme 4 (which I have):

http://www.newegg.com/Product/Product.aspx?Item=N82E16813157543

is around $200 not counting sales. High end Z97 boards cost more. And the only reason that Intel sells $1K CPUs is AMD blows for the past decade.

 

[DerpEinstein]

What expense. The Asrock X99 Extreme 4 (which I have):

http://www.newegg.com/Product/Product.aspx?Item=N82E16813157543

is around $200 not counting sales. High end Z97 boards cost more. And the only reason that Intel sells $1K CPUs is AMD blows for the past decade.

Good z170 boards are $100.

 

[Dresdenboy]

This one has the exact same ES OPN as the Cinebench CPU we saw a week ago...

Maybe it wasn't a Chinese leak?

I heard of an OPN (no rumour, but a trace) which fits the CB pixels even better than my guess. The "5" is one of the 5 differences.

 

[krumme]

Maybe it wasn't a Chinese leak?

I heard of an OPN (no rumour, but a trace) which fits the CB pixels even better than my guess. The "5" is one of the 5 differences.

My french is very bad. But as i can read from the start of the review you need to have a specific bios to unlock all the performance of zen. They had such a bios. I asume they only got it recently. Thats my understanding.
But we need some with better french to read all pages and chime in.

For better quality go here
https://forums.anandtech.com/index.php?posts/38648267

 

[krumme]

My tranlation from page 1 is their sample have an recent AGESA ( firmware that is integrated into the bios).
https://en.m.wikipedia.org/wiki/AGESA

 

[BeepBeep2]

Maybe it wasn't a Chinese leak?

I heard of an OPN (no rumour, but a trace) which fits the CB pixels even better than my guess. The "5" is one of the 5 differences.

Interesting...

 

[faseman]

Another thread where Sweepr posts garbage info and bleedingly obvious bias.

Safe to add him to the ignore list fellas!


Insulting other members is not allowed
Markfw
Anandtech Moderator

 

[BeepBeep2]

Another thread where Sweepr posts garbage info and bleedingly obvious bias.

Safe to add him to the ignore list fellas!

Did you read a single word in this thread?

Good info from CanardPC, can't get much more of a reliable source than that. Low clocks of the sample and normalization of scores to 6600K make it harder to extrapolate what launch performance is really going to be however.

 

[ultima_trev]

I'll go the conservative route and assume final production clock at 3.4 GHz and all core turbo at 3.6, dual to few core at 3.7 and single core at 3.8.

Assuming the chip isn't too thermal or power constrained to reach turbo clocks, that's about an 8% increase in clock speed over the engineering sample. Seeing the 4790K and 6700K scores ahead of everything in the gaming aspect, it seems clocks trump cores/threads thus play the majority role. Even if the IPC to clock scaling isn't as good as Intel's, I'm pretty certain in gaming loads the Ryzen should be 1-2% ahead of the i5 6600K and probably only 5-6% behind the i7 6900K.

Since this Canard PC is supposedly a legit source, it seems it really all comes down to final production clocks at this point. Of course, I'm sure increased clocks will hamper the performance/watt a bit, pushing it (albeit slightly) in Intel's favor.

 

[LTC8K6]

No, I am not. Intel won't drop it's prices, because their monopoly controls the market share and the mind share.

We have already done this dance! AMD has had superior products, and even had them priced lower than intel's components, yet the ignorant consumers still bought the intel products.

Intel will make more money by keeping their prices the same.

If Intel's prices are higher than they need to be, as you claim, then Intel can indeed lower them at any time. If Intel does not lower their prices, then they can't be too worried about Ryzen. If Intel does indeed get worried about Ryzen, then as you claim, they can simply lower prices to combat Ryzen, and still profit on each chip.

You create a scenario where it's impossible for AMD to gain on Intel, and it's impossible for Intel to lose.

I contend that your scenario is likely wrong.

 

[krumme]

I'll go the conservative route and assume final production clock at 3.4 GHz and all core turbo at 3.6, dual to few core at 3.7 and single core at 3.8.

Assuming the chip isn't too thermal or power constrained to reach turbo clocks, that's about an 8% increase in clock speed over the engineering sample. Seeing the 4790K and 6700K scores ahead of everything in the gaming aspect, it seems clocks trump cores/threads thus play the majority role. Even if the IPC to clock scaling isn't as good as Intel's, I'm pretty certain in gaming loads the Ryzen should be 1-2% ahead of the i5 6600K and probably only 5-6% behind the i7 6900K.

Since this Canard PC is supposedly a legit source, it seems it really all comes down to final production clocks at this point. Of course, I'm sure increased clocks will hamper the performance/watt a bit, pushing it (albeit slightly) in Intel's favor.

I actually hope the fastest bin is 5% below bwe 6900 viewed over a broad load and not say 5% faster.
I think the price elasticity of amd cpu will be more sensitive than a far more established brand as Intel. And as the fastest cpu will always command an irrational crazy premium what we dont want is amd to be fastest.
Its a large symbolic effect if its faster or slower by 5% even if it from other perspectives is of minor importance. We dont know eg die size or efficiency.
And then i hope gf have tons of capacity. Because what we dont want as consumers is competing with professional servermarket for the the same ccx. We can hope we get the low efficiency ones and that there is plenty of those.

 

[Vortex6700]

You create a scenario where it's impossible for AMD to gain on Intel, and it's impossible for Intel to lose.

I contend that your scenario is likely wrong.

Intel created that scenario during A64 by offering billions in rebates to companies that refused to do any business at all with AMD.

The scenario was very wrong, and very illegal, but it did happen.

 

[lolfail9001]

Intel created that scenario during A64 by offering billions in rebates to companies that refused to do any business at all with AMD.

The scenario was very wrong, and very illegal, but it did happen.

Just saying, AMD still gained during that period, so it [AMD not gaining on Intel] did not happen even in that period.

 

[Dresdenboy]

Just saying, AMD still gained during that period, so it [AMD not gaining on Intel] did not happen even in that period.

So despite these headwinds they gained something? How would that work out without strong headwinds?

 

[Conroe]

When core 2 came out a ~$200 e6400 blew away the ~$600 x2 4800. I hope AMD can return the favor.
I'm interested in the 4c/8t zen. If it can also run with the i5's and is priced with the i3's it could be a money maker. To counter it all Intel needs to do is lower every skew one price bracket and perhaps add a new halo product.

 

[Makaveli]

The x2 4800+ was never $600.

u are confusing that for the FX60.

http://www.tomshardware.com/reviews/price-performance-charts-oct-14-2006,1866.html

 

[Conroe]

http://www.pcmag.com/article2/0,2817,1989161,00.asp

Intel's 2.4-GHz E6600 will ship for about $316, while the 2.13-GHz E6400 and 1.86-GHz E6300 will be priced at $224 and $183, respectively. Meanwhile, AMD's 2.4-GHz Athlon 64 X2 4800+ is priced at $645, while the 2.2-GHz 4200+ and 2.0-GHz 3800+ are priced at $365 and $303, respectively.

As I recall they were just over $500 on newegg at the time. Never were worth the MSRP. I had a Opteron165 that was better than most 4800's.

 

[badb0y]

This looks damn good to me, I don't know why some people here are trying to downplay it. An ES that clocks around 3.15-3.3 Ghz and it swinging right at it's 6/8-core competitors from Intel? Sign me up fam.

 

[Makaveli]

http://www.pcmag.com/article2/0,2817,1989161,00.asp


As I recall they were just over $500 on newegg at the time. Never were worth the MSRP. I had a Opteron165 that was better than most 4800's.

I never saw the 4800 at that price in Canada.

But then again the Opteron was the better deal anyways I still have an Opteron 170 doing HTPC duty

 

[Shlong]

I really hope AMD's cpu is good. My last AMD cpu purchase was an Athlon64 X2 over 10 years ago, I've bought nothing but intel since. I want a nice 8 core cpu so I really hope this AMD chip lives up to the hype and is released at a great price ($500 - $600).

 

[escrow4]

Good z170 boards are $100.

The frisbee ones sure, the ones that can sustain a decent overclock and have a stack of extra ports ain't $100. $100 is H170 pricing, Z170 would be heavily cut back and probably wouldn't even be full ATX but cut down one side.

 

[Conroe]

Your linked article is two weeks prior to launch and mine is from October a few months after launch. I think the pricing in the latter will be a little more accurate.

I never saw the 4800 at that price in canada.

What core 2 did was historic. Sandy is the closest thing since. Ryzen could be another great moment in cpu history, but it's not like Conroe. It is not faster than anything else. It looks good but it's pricing and overclockablity is still in question.