Thursday, August 11th 2016

AMD "Summit Ridge" ZEN CPU at 2.80 GHz Beats 3.40 GHz Core i5-4670K

According to performance numbers of an AMD "Summit Ridge" ZEN CPU engineering-sample put out by WCCFTech, AMD's claims of IPC gains are gaining credibility, and showing signs of the gaming PC processor market warming up again. An engineering sample featuring 8 cores and 16 threads (via SMT), beat Intel's Core i5-4670K processor. This sample featured clock speeds of 2.80 GHz, with 3.20 GHz boost.

The "Summit Ridge" sample provided 10 percent higher frame-rates than a Core i5-4670K, in the "Ashes of the Singularity" 1080p benchmark. The chip is still convincingly beaten by 12 percent, by a Core i7-4790 (non-K), running at 3.60 GHz, with 4.00 GHz boost. This shows that AMD could leverage the new 14 nm FinFET process to crank up clock-speeds, and produce SKUs competitive with current Intel "Skylake-D" Core i5 and Core i7 processors.
Source: WCCFTech
Add your own comment

126 Comments on AMD "Summit Ridge" ZEN CPU at 2.80 GHz Beats 3.40 GHz Core i5-4670K

#101
Ubersonic
Solaris17seriously, I want to know what these numbers are compared to the skylakes everyones currently buying, not haswell CPUs
Well look at it this way, on that benchmark the Z8 was ~10% quicker than the i5-4670K while clocked ~15% slower. According to Anandtech's CB R10 charts the i5-6600K is ~14% quicker than the i5-4670K while clocked the same. It's only an assumption here but extrapolating that it would then mean that the i5-6600K is ~4% quicker than the Z8 while clocked 18% quicker.

Of course that's only apples to bananas napkin math. But, it's safe to say that if AMD price the new 8 core against the i5 like they have done previously and it comes out clocked higher than the prototype then they will have a real dragon slayer on their hands.
Posted on Reply
#102
jabbadap
Solaris17seriously, I want to know what these numbers are compared to the skylakes everyones currently buying, not haswell CPUs
Not really sure can you compare these results very reliably to anything. That ashes cpu fps depends so much from gpu and version of the bench that between different comps, result are very different(i.e. if you underclock your graphics card your cpu fps will rise).

This is the result that I founded closes to Okidna's pic:
www.ashesofthesingularity.com/metaverse#/personas/75dd84a8-d8c0-47cc-ad4f-10e88c397845/match-details/c3ea33e6-b8ef-4723-bec0-6bc51b0c971c
1080 Standard, 1.30.21168.0
[INDENT]i5-6600K
cpu fps: 68.9
NB: 74.1
MB: 71.8
HB: 62.2
vs
Zen ES
cpu fps: 58.0
NB: 64.4
MB: 60.5
HB: 50.8
[/INDENT]
Just go through the ashes benchmark dB, you might get the picture or not how reliable this benchmark is.
Posted on Reply
#103
G33k2Fr34k
UbersonicWell look at it this way, on that benchmark the Z8 was ~10% quicker than the i5-4670K while clocked ~15% slower. According to Anandtech's CB R10 charts the i5-6600K is ~14% quicker than the i5-4670K while clocked the same. It's only an assumption here but extrapolating that it would then mean that the i5-6600K is ~4% quicker than the Z8 while clocked 18% quicker.

Of course that's only apples to bananas napkin math. But, it's safe to say that if AMD price the new 8 core against the i5 like they have done previously and it comes out clocked higher than the prototype then they will have a real dragon slayer on their hands.
Except your comparison is pointless. The AOS benchmarks makes little to no use of the extra cores in the Zen chips. The i7 4790 beats the 8-core i7 5960X in this benchmark too. Does that mean that the i7 4790 is faster than i7 5960X? No.

As developers become better at utilizing threading using the new graphics APIs the difference between 8-core and 4-core CPUs will become more and apparent. BTW, Vulkan seems to be much better in that department.
Posted on Reply
#104
Vayra86
G33k2Fr34kThe i7 4790 beats the 8-core i7 5960X in this benchmark too. Does that mean that the i7 4790 is faster than i7 5960X? No.
Yes - situationally. Quad cores clock higher, and in games are evidently faster. That doesn't limit itself to AoTS at all, but is a universal given within gaming. Games just like fast cores and high clocks.

So, if Zen can achieve similar perf with higher core counts but *lower* actual clocks on those cores, and they can put the double amount of cores at the same price point as an i7... AMD is going to destroy gaming CPU benches. I could care less about what they put on the box in terms of Ghz, I want to know if it clocks beyond 4 Ghz on that core count. Then we've got something to talk about :)

Per-core performance in the end is irrelevant, if AMD can put more cores to work in actual games, and all their efforts are pointed towards that - in the end, Intel's mainstream is STILL very much focused on quad core products. Even if AMD manages to get a foot in the door this way in terms of sales and relative performance, they can already push Intel into moving their hexacores towards the mainstream market which will be a win and actual progress for once, after a decade of stagnation.
Posted on Reply
#105
xorbe
Nobody99But you won't be able to buy AMD laptop and expect the battey to last.
I have an apu laptop now and yeah, my eye is on an intel+pascal laptop later this year.
Posted on Reply
#106
arbiter
JismThat does'nt matter. What matters for AMD is to offer a product in both enterprise and consumer-market, that is able to compete. And when it competes in performance and price, AMD has a winner. Just as it had with that 480X.
Well pair that RX 480 with a more budget cpu, if some results i seen are right performance goes off a cliff if you aren't using a good fast cpu with the 480.
Posted on Reply
#107
Steevo
I forsee clocking issues with GloFlo 14nm process and power consumption being 10-20% higher than Intel chips at the same performance level. But that being said, 40% IPC is equal at 40% lower clocks, plus SMT on each cores during pipeline stalls may help mask multi-threaded issues. it would be good to see some Prime and SuperPI numbers, and any advanced hardware they may have put in, and hopefully the software to use it isn't the vaporware that AMD/ATI is known for spinning in the PR machine.

If they are within 5-10% of the IPC of Intel they have a winner, first in that they already are selling chips at a price point, and if they keep with that strategy, of being an IP vendor with some goods on showcase who are hungry for the market, who can make custom chips with everything, and are on the spot when it comes to support.. They could be in place for growth instead of the stagnated still born company they have been for the last few years.
Posted on Reply
#108
Caring1
arbiterWell pair that RX 480 with a more budget cpu, if some results i seen are right performance goes off a cliff if you aren't using a good fast cpu with the 480.
Same can be said for pairing a 1080 with a sandy bridge CPU, performance suffers.
Posted on Reply
#109
TheGuruStud
AsRockAnd yet many claim running 120Hz with 120fps is better than running 60Hz with 60fps.

How ever you look at it to me 7 fps is a lot imo for the games i play and keeping it above 30-35fps is a priority, were some claim they are not happy gaming with any thing less than 60fps.

In the end not all people are the same or require the same, i remember i was not able to use CRT screens lower than 75Hz back in the days( i can see the flicker very easy ). But gaming i am perfectly fine as long as it's over 30fps with LCDS.

In the end it don't matter until real benchmarks are released. As hell if i am going buy a AMD cpu just because it runs A game well which to me this benchmark don't really show that.

Time will tell if Zen makes it or not, i hope so as AMD needs it.
Most LCDs heavily blur. That doesn't fool me, though, it gives me a headache. 60 is fine, but over 85 is so smooooth. 120 lets me kill bitches in CS like ants.
Posted on Reply
#110
chuck216
It looks promising at least the 1D sample does, I'm guessing the 2D sample has some things disabled.

While everyone is busy comparing the Zen to the I5 and i7 I'd like to bring up a couple points... First, There is no way to tell how many threads the chip is running whether 8 or 16 and if 16 how much of the processor is being used for each thread.

The Bench that impresses me is the comparison with the FX 8350. It is doing 38% more performance with 70% of the processor clock speed. Meaning if everything scales and they can get 3.5 or 4.0 Ghz out of the new cores it will literally double the performance if the Vishera chips clock for clock.

Again without knowing how many threads were utilized for the benchmark it's hard to tell if it's actually doubling clock for clock performance on 8 threads or 16 threads.
Posted on Reply
#111
arbiter
Caring1Same can be said for pairing a 1080 with a sandy bridge CPU, performance suffers.
Um not in same way 480 suffers. posted it many times so far. I will just link it, but if you buy gtx1080 you are not dropping it in no budget 100$ cpu

i.imgur.com/JF7ngP5.png
Posted on Reply
#112
lanlagger
for some "8C/16T > 4C/4T 10% victory" does not sound impressive and promising, but it is!!! because prior this AMD would pack how many cores he wanted and clock them as high as AIO coolers could not even cool them (past 5.0 ghz or whatever)... and still wont beat intels i3 2C/4T... now AMD can beat even i5. As long as AMD can price them low (and they can - look at this FX series - how many cores priced so low) we will have upper midrange CPU competition again (since bulldozer there was zero... there fore intels progress was also sub 5% and prices also managed to bump up).
Posted on Reply
#113
64K
lanlaggerfor some "8C/16T > 4C/4T 10% victory" does not sound impressive and promising, but it is!!! because prior this AMD would pack how many cores he wanted and clock them as high as AIO coolers could not even cool them (past 5.0 ghz or whatever)... and still wont beat intels i3 2C/4T... now AMD can beat even i5 as long as AMD can price them low (and they can - look at this FX series - how many cores priced so low) we will have upper midrange CPU competition again (since bulldozer there was zero... there fore intels progress was also sub 5% and prices also managed to bump up).
As far as pricing goes, it depends partly on recouping development cost and how much AMD may have spent on R&D on Zen and how much AMD has to spend to get companies that manufacture PCs and servers to use more of their Zen chips. That seems to be their biggest hurdle. Getting more PC and server manufacturers to buy and use their chips.

Probably the Zen chips will be priced pretty low but that's bad long term. AMD goes further into debt quarter after quarter with that policy. They need to get out of the red if they can and have some money also for R&D for future CPU generations.



Edit: This is only one article that I have read on the subject of AMD's decline in the CPU department according to financial analysts but it may be of some interest.

www.investopedia.com/stock-analysis/032515/3-major-problems-facing-advanced-micro-devices-amd.aspx

Taken from article:

"AMD competes with Intel in PC and server processors but has been bleeding share in both markets. In servers, AMD won as high as 25% market share in 2006, making it a major player in the industry. Today, AMD is essentially nonexistent in the segment, claiming a low single-digit share. Meanwhile, Intel has built a near-monopoly, allowing it to charge high prices and extract extremely high margins."

"In PCs, the story is much the same. Intel has continued to steal market share from AMD, especially at the low end with its Atom chips, despite already controlling most of the segment. During the second quarter of 2014, Intel generated nearly 95% of PC processor revenue, shipping 84% of all desktop processors and 88% of all laptop processors."
Posted on Reply
#114
argon
from a 1090t .. to what you would upgrade ?

---6800k
or
this ?
Posted on Reply
#115
hapkiman
argonfrom a 1090t .. to what you would upgrade ?

---6800k
or
this ?
Ha - I still have a rig based around a Phenom II x6 1090T and it runs fine (it's not my main rig, but I still use occasionally). But I understand that you would want to upgrade. If you need the cores, the i7 6800k or i7 6850k would be an excellent choice, but don't forget you'll need a new mobo, RAM, etc...
Posted on Reply
#116
Jism
argonfrom a 1090t .. to what you would upgrade ?

---6800k
or
this ?
If you really need it, go out and buy whatever you need. Perhaps a second handed 8350 or something is a better upgrade on the 1090T.

If you can wait, wait the results for ZEN, because they will be cheaper and if not intel drops their prices as well.

I've upgraded from a 1055T with 4.2GHz clocks to a 8230 now on 4.6Ghz. It's defenitly faster, but that half baked 4 /8 cores thing design on the vishera lacks on some benches and / or applications. It is technically not a real 8 core with a shared FPU causing problems with some apps. For instance, Linx benchmark runs far better (400% better) on a Thuban X6 compared to a 8320 / 8 cores / 4.6Ghz.

Does'nt mean that they are bad CPU's ; no, they will work in any day application just fine. But coming from a 1055 standpoint, it's different.
Posted on Reply
#117
argon
JismIf you really need it, go out and buy whatever you need. Perhaps a second handed 8350 or something is a better upgrade on the 1090T.
.
my gpu is an used fury got used two week ago, i upgraded from a barely playing 6950 that has made his life... yesterday i see lots of bottleneck and i'm living a week with shuttering.
sincerely i dont like what amd did with fx8150 .
I don't want to upgrade to an old again platform.. i want those new 14nm . 6800k already got it.
but my question is, if this win by not so lot on an old i5-4xxx series, the 6800k will be on top of zen or it will lose?
Posted on Reply
#118
TheGuruStud
argonmy gpu is an used fury got used two week ago, i upgraded from a barely playing 6950 that has made his life... yesterday i see lots of bottleneck and i'm living a week with shuttering.
sincerely i dont like what amd did with fx8150 .
I don't want to upgrade to an old again platform.. i want those new 14nm . 6800k already got it.
but my question is, if this win by not so lot on an old i5-4xxx series, the 6800k will be on top of zen or it will lose?
This benchmark (even if real) is pretty worthless. If real, intel is likely going to be pissed (who am I kidding, intel probably has a dozen spies at AMD).
Posted on Reply
#119
Jism
argonmy gpu is an used fury got used two week ago, i upgraded from a barely playing 6950 that has made his life... yesterday i see lots of bottleneck and i'm living a week with shuttering.
sincerely i dont like what amd did with fx8150 .
I don't want to upgrade to an old again platform.. i want those new 14nm . 6800k already got it.
but my question is, if this win by not so lot on an old i5-4xxx series, the 6800k will be on top of zen or it will lose?
It's one benchmark, proving nothing basicly. Yes we see some numbers but not the total value of the chip, pricing and all.

You have a AM3+ platform, if you can 'score' a 8320 or anything in particular for cheap (used that is) call it a day. The vishera chip is pretty good, however on some benchmarks it's just not fast enough as a 1090T would be. It is due to the design and shared FPU. But it does'nt make it a bad CPU in particular. The vishera runs everything great basicly, and games with Vulkan should even do better.

So. If you need it > go for whatever your budget allows you to.
Posted on Reply
#120
TheGuruStud
JismIt's one benchmark, proving nothing basicly. Yes we see some numbers but not the total value of the chip, pricing and all.

You have a AM3+ platform, if you can 'score' a 8320 or anything in particular for cheap (used that is) call it a day. The vishera chip is pretty good, however on some benchmarks it's just not fast enough as a 1090T would be. It is due to the design and shared FPU. But it does'nt make it a bad CPU in particular. The vishera runs everything great basicly, and games with Vulkan should even do better.

So. If you need it > go for whatever your budget allows you to.
I've seen them for 100 bucks a lot. That's a steal. OC to 4.8 or so and call it a day.
Posted on Reply
#121
Jism
I'm not sure what motherboard he has in particular, but oc'ing is never a guarantee. I suggest that he is running a 1090T stock, while they should OC perfectly to 4.2 up to 4.4Ghz.
Posted on Reply
#122
TheGuruStud
JismI'm not sure what motherboard he has in particular, but oc'ing is never a guarantee. I suggest that he is running a 1090T stock, while they should OC perfectly to 4.2 up to 4.4Ghz.
It's as close to a guarantee as you're going to get this late in production. I hit 4.7 on air with my 8350. It was 1st batch and not a good OCer (a little V hungry).
If it can't hit 4.7, then return that POS (just like I should have returned ALL of the haswell chips, bought IVY and mailed F YOU letters to every intel employee alive).
Posted on Reply
#123
argon
JismI'm not sure what motherboard he has in particular, but oc'ing is never a guarantee. I suggest that he is running a 1090T stock, while they should OC perfectly to 4.2 up to 4.4Ghz.
i have an 890gx msi, i have oc'ed at 3.8ghz with a prolimatech genesis and a single fan....i dont think this mobo support the fx8350.
I'll play again for a month in this condition then I'll look out to see the i7-6800k(6core) or just wait the new years for this amd zen?
I need a new platform fast!! my monitor is 120hz is so shitty play in this condition right now...
Posted on Reply
#124
Nobody99
I would like to point out that multiple threads are fine but given that a lot of applications can't even use all the threads it is still better to have less threads when the threads are weaker (i5 vs Zen). Even Handbrake can't use all the threads if you have a lot of them like 40 on a single running encode.
Posted on Reply
#125
Shou Miko
Ashes of the Singularity has always been a AMD favor game I would like to see some encoding instead without GPU acceleration to see if it can really beat a i5-4670k without breaking a sweat at 2,80ghz.

Plus games has soo many variables to actually count.
Posted on Reply
Add your own comment
Nov 25th, 2024 10:03 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts