Leaked AMD Ryzen Benchmarks?

lolfail9001 said:
If it cannot beat them at 640x480, it cannot beat them anywhere, just tie. Reality of CPU limitations :p
Click to expand...

I'm not quite sure it works like that. My understanding (especially from Kyle's statements at the event yesterday) is that there are more and more games supporting 8 threads. I just can't see a low resolution stressing the CPU enough to take advantage of more cores. You're correct though in that it IS an important metric for IPC. And if clocks are truly separated by 1+ghz, there will be a heavy favor to Intel's chips (and rightly so!).
 
Romeomium said:
I'm not quite sure it works like that. My understanding (especially from Kyle's statements at the event yesterday) is that there are more and more games supporting 8 threads. I just can't see a low resolution stressing the CPU enough to take advantage of more cores. You're correct though in that it IS an important metric for IPC. And if clocks are truly separated by 1+ghz, there will be a heavy favor to Intel's chips (and rightly so!).
Click to expand...


You mean more than 8 threads ;)

Yes that is happening. But so far we have 3 games maybe 4 that do. So its going to be a while before we start really seeing that as a main stream option from a gaming software perspective.
 
razor1 said:
They didn't use 2133, they used 3400mhz ram, it was downclocked automatically by the motherboard.
Click to expand...
so it ran at 2133 anyway, so okay big difference, OJ didn't mean to murder his wife that way, he had it planned differently.
 
That is a problem, if motherboards are automatically doing that, you are stuck with very few boards that can use top rated ram. And those aren't going to be cheap. This B motherboard (MSI so its not a cheapo brand) is probably what 80 bucks, where the Asus motherboards you are talking about are probably the X types which are 200 and up?
 
razor1 said:
This B motherboard (MSI so its not a cheapo brand) is probably what 80 bucks, where the Asus motherboards you are talking about are probably the X types which are 200 and up?
Click to expand...

X370 boards are $150 and up. The B series boards are mainstream.
 
razor1 said:
That is a problem, if motherboards are automatically doing that, you are stuck with very few boards that can use top rated ram. And those aren't going to be cheap. This B motherboard (MSI so its not a cheapo brand) is probably what 80 bucks, where the Asus motherboards you are talking about are probably the X types which are 200 and up?
Click to expand...
ffs its a validation board with ES they got it off the aftermarket, the board has old bios settings.
 
razor1 said:
That is a problem, if motherboards are automatically doing that, you are stuck with very few boards that can use top rated ram. And those aren't going to be cheap. This B motherboard (MSI so its not a cheapo brand) is probably what 80 bucks, where the Asus motherboards you are talking about are probably the X types which are 200 and up?
Click to expand...

I think that's way too speculative - if B boards allow overclocking I don't see why they'd lock the ram down. It probably defaults to 'auto' but allows manual adjustment.

edit: As suspected, the B boards will allow OC on the ram

https://www.asus.com/us/Motherboards/PRIME-B350M-A/specifications/
 
razor1 said:
IF? this wasn't that, I read part of the article before it was taken down. They were retail samples.
Click to expand...

It is a ZD, there are leaks of official Ryzen String CPU's making it a testing sample CPU, the system was bought aftermarket from China
 
http://www.fudzilla.com/news/processors/42986-amd-debunks-ryzen-ddr4-ram-issues

amd-ryzenDDR43400-2.jpg
 
OrangeKhrush said:
It is a ZD, there are leaks of official Ryzen String CPU's making it a testing sample CPU, the system was bought aftermarket from China
Click to expand...


ZD doesn't mean engineering or testing sample, Qualification samples are = to retail samples.

I'm going to go throught this again.

Qualification samples are sent to motherboard manufactures to simulate what the retail samples will be for them to finalize boards, there is almost no difference if any between Qualification and retail samples.

I have here with me, ES of GPU's and had one of Intel CPU's (old P4) till last month when I cleaned out my basement. All of them are very close to retail releases.

This is not something new to me.

If you contact either AMD or Intel and tell them what you are doing, like making a motherboard, they will give you qualification samples which are retail samples without warranty. Qualification samples are the same as Retail. They have to be, because a manufacturer can not base their hardware on a chip that has unknowns.
 
Oh wait there is more


Which was confirmed by my source that AMD updated bios for final reviews and unlocked max frequency DRAM apparently only 1 vender cannot reach 3800 but ASUS have rubber stamped 3800Mhz as achievable.
 
OrangeKhrush said:
Hmmmm lets see

Validation Motherboard - check
DDR4 2133 While Intel runs 3200 - check
ES sample - check

Results are right up in the Intel expensive parts range.....

This ____flute is comparing AIDA quad channel setups and expecting it to reveal anything anyone didnt already know.
Click to expand...

Hey now, he's just doing his JOB. :LOL:
 
thesmokingman said:
Hey now, he's just doing his JOB. :LOL:
Click to expand...

He also ignored every other bench that has the Ryzen competing with Intel's creme on a fuddy motherboard with lowest DRAM possible, like when Kyle does Intel benches he takes H110 motherboards with generic DDR4 2133 and benches Intel CPU's....oh wait no real review will ever do that.
 
OrangeKhrush said:
Oh wait there is more


Which was confirmed by my source that AMD updated bios for final reviews and unlocked max frequency DRAM apparently only 1 vender cannot reach 3800 but ASUS have rubber stamped 3800Mhz as achievable.
Click to expand...

Memory looks good to me. CPC can't see that apparently. Trolls
 
razor1 said:
That is a problem, if motherboards are automatically doing that, you are stuck with very few boards that can use top rated ram. And those aren't going to be cheap. This B motherboard (MSI so its not a cheapo brand) is probably what 80 bucks, where the Asus motherboards you are talking about are probably the X types which are 200 and up?
Click to expand...

How many people do you know buy an $80 motherboard and go out and buy 3200 DDR4 memory :)
It does not help that they are not using a retail CPU but ES, and fook knows what the results will be like if one just leaves the motherboard to default to what it wants on using such sticks.
If the CPU is ES (which is far from ideal), who knows the status of the motherboard.
Not really worth debating these results IMO and better off to wait IMO, about as reliable as what we see on WCCFT :)
Guru3d mentioned they are using ES so just going off them in this instance regarding Iranian review as I could not read it in time.
Hilbert says:
Looking at the photo it is an engineering sample though
Click to expand...
.
Cheers
 
Last edited:
razor1 said:
ZD doesn't mean engineering or testing sample, Qualification samples are = to retail samples.

I'm going to go throught this again.

Qualification samples are sent to motherboard manufactures to simulate what the retail samples will be for them to finalize boards, there is almost no difference if any between Qualification and retail samples.

I have here with me, ES of GPU's and had one of Intel CPU's (old P4) till last month when I cleaned out my basement. All of them are very close to retail releases.

This is not something new to me.

If you contact either AMD or Intel and tell them what you are doing, like making a motherboard, they will give you qualification samples which are retail samples without warranty. Qualification samples are the same as Retail. They have to be, because a manufacturer can not base their hardware on a chip that has unknowns.
Click to expand...

Good luck! I did spend an entire week in another forum explaining the difference between ES and QS. I explained them that QS is technically identical to the commercial chip, and that only differences are "cosmetic" such as packaging (not a fancy box), labeling ("confidential"), or the way the chip identifies itself (via a codified string). They did'nt learn anything and even today they still continue pretending that QS are engineering samples with disabled stuff and on early silicon...
 
From the iranian website themselves:

"Unfortunately, because of problems of motherboard BIOS when doing these tests enable XMP memory configuration, and even there was no manual for it and with the same frequency of 2133 MHz standard tests were done to poor performance. DDR4 memory Kit performance however our used with this processor very closely and even somewhat better than the performance in the same frequency with the Core i7 processor is 6700K and 7700K, hence the charge of each finger to the motherboard.
Because of our use case being applied preterm motherboard manual not boot memory frequency higher 2133 causes the system."
Click to expand...

Basically using a very old test setup before AMD did the 22/2 updates and 26/2 update.....yeah gimped but still runs Intel's premium chips.
 
Shintai said:
http://www.guru3d.com/news-story/first-ryzen-7-1700x-review-finds-its-way-onto-the-web.html

Reality is starting to catch up to the BS.
ryzen1.png


Good luck with the IPC too. You can only fool so many with an increased L2.

ryzen2.png
Click to expand...
Shintai said:
A GTX980 was used.

Looks like we wont see compression benches at least anymore :)
WinRAR-5.3.png

7-Zip-15.14.png
Click to expand...

So 174 fps is inadequate? This is bullcrap and you guys know it. Anything above 120fps and possibly less is totally unimportant and irrelevant. You use pathetic 7 zip and winrar as benchmarks, They are complete crap. Try winzip where multicore support is rewarded. You are clutching at straws here. Poor code . So when there is a superior alternative to these benches you ignore it and parade out benches for programs most people do NOT use or shouldn't use because of superior alternatives. Hang it up!!!
 
Last edited:
CSI_PC said:
It does not help that they are not using a retail CPU but ES, and fook knows what the results will be like if one just leaves the motherboard to default to what it wants on using such sticks.
If the CPU is ES (which is far from ideal), who knows the status of the motherboard.
Not really worth debating these results IMO and better off to wait IMO, about as reliable as what we see on WCCFT :)
Cheers
Click to expand...

Except is is not an ES... because the ES strings start with "1" or "2" depending if they are first gen or second gen ES. The chips that start with "Z" aren't ES.
 
LL


DDR4 3200 kit, already Read/Write is up with the Skylake processors on Dual Channel, Latency is a bit miff but mine are not that fantastic at 3000
 
CSI_PC said:
How many people do you know buy an $80 motherboard and go out and buy 3200 DDR4 memory :)
It does not help that they are not using a retail CPU but ES, and fook knows what the results will be like if one just leaves the motherboard to default to what it wants on using such sticks.
If the CPU is ES (which is far from ideal), who knows the status of the motherboard.
Not really worth debating these results IMO and better off to wait IMO, about as reliable as what we see on WCCFT :)
Guru3d mentioned they are using ES so just going off them in this instance regarding Iranian review.
Hilbert says:
.
Cheers
Click to expand...


Its not an ES, its a qualification sample ;) Guru3d got it wrong.

Yeah I agree with ya on the ram thing though.
 
juanrga said:
According to that graph the power consumption dependence is almost cubic.
Click to expand...
Biquadratic actually, if you assume that stock Ryzen consumes 9W in idle and it had 3.5Ghz clock in P95.
Dullard said:
DDR4 at those speeds has to loosen up a bit, at least they are at 1T. Unless you were /s
Click to expand...
Let me quote someone who actually overclocks stuff:

Now compare that and what AMD shows here.
 
OrangeKhrush said:
Latency is a bit miff but mine are not that fantastic at 3000
Click to expand...
I mean, Skylake has exactly half the latency at very similar settings (looser timings a little, but same MT rate). But a whole 2Gig/sec lower bandwidth.
Another curious thing is that Skylake has way lower latencies than what AIDA has here. But i guess that's the trade-off of cache vs latency.
 
juanrga said:
Except is is not an ES... because the ES strings start with "1" or "2" depending if they are first gen or second gen ES. The chips that start with "Z" aren't ES.
Click to expand...
You have no veracity for anything you say. Credibility less than zero. I want to know why you spend your whole life attacking AMD as if Intel is a source of virtue??? Get a life and dry up. I saw you on Canard PC on twitter yesterday. Your memory claims were debunked completely by Dresden Boy. There are a few Intel fanatics here who you can harmonize with, but your fake news will never be accepted. Your fake science will never be accepted. A scientist has to be objective, a quality you have never had.if you had any integrity you would be dangerous..
 
Last edited:
lolfail9001 said:
I mean, Skylake has exactly half the latency at very similar settings (looser timings a little, but same MT rate). But a whole 2Gig/sec lower bandwidth.
Another curious thing is that Skylake has way lower latencies than what AIDA has here. But i guess that's the trade-off of cache vs latency.
Click to expand...

Like I said the latency is a bit on the high side but that may be more BIOS related, it is not like Haswell and Skylake were without issues with boards upfront, especially if you get early access, we had issues a plenty going from 1155 to 1150 and again with the X99 socket was horrible at first. It is probably the reason for some benches being weird but I am willing to give a new chipset time to teeth in. Not like i will have a motherboard for it anyway until July
 
razor1 said:
Its not an ES, its a qualification sample ;) Guru3d got it wrong.

Yeah I agree with ya on the ram thing though.
Click to expand...
Ah ok.
Well without being able to read the article I could only go by what he said :)
But still, it means it is not the actual retailer product and context still applies it raises questions where and revision of motherboard they have (which controls the microcode).
Memory config they did/let happen is enough anyway to make me leery.
Cheers
 
Last edited:
CSI_PC said:
How many people do you know buy an $80 motherboard and go out and buy 3200 DDR4 memory :)
It does not help that they are not using a retail CPU but ES, and fook knows what the results will be like if one just leaves the motherboard to default to what it wants on using such sticks.
If the CPU is ES (which is far from ideal), who knows the status of the motherboard.
Not really worth debating these results IMO and better off to wait IMO, about as reliable as what we see on WCCFT :)
Guru3d mentioned they are using ES so just going off them in this instance regarding Iranian review as I could not read it in time.
Hilbert says:
.
Cheers
Click to expand...

Hey, I did :), but I needed m-ATX for htpc use, so no choice as there are no Z370 m-ATX boards. As the ASUS Prime supports 3200 ram, I figured why not for $20 more.
 
lolfail9001 said:
Biquadratic actually, if you assume that stock Ryzen consumes 9W in idle and it had 3.5Ghz clock in P95.

Let me quote someone who actually overclocks stuff:

Now compare that and what AMD shows here.
Click to expand...

I wasn't saying they were spot on excellent, but even the 6950X @ 13-18-18-28/3400 isn't radically glorious compared to the 18-17-17-37/3400 that started this conversation (and yeah, I realize that 18-17-17-37 isn't all that tight, but it's been a whole day - give them some time).

And I bet those OCers that hit 12-11-11-28-1T/4200 w/Kaby Lake are running a single 4GB stick or something. Probably as rare as we hear those HW-E/BW-E octos hitting 4.6GHz are. Like mine, for instance - but that can't be, since obviously I don't overclock.
 
You must log in or register to reply here.
Back
Top