Ryzen at stock without turbo beats intel 6900k Goodbye intel

[Clanscorpia]

1 minute ago, samcool55 said:

How many samples was he running?

It defaults to 200 and the benchmark is done with 150. It's a little detail that makes a big difference...

 

There has been a bit of a hassle around it but AMD responded with the sample thing and it seems like that was indeed the issue.

 

Im not sure. This is just what I heard. That no-one could match it but him

[UMxMarky94]

1 minute ago, Clanscorpia said:

Im not sure. This is just what I heard. That no-one could match it but him

So the plot thickens !

[brokencross]

4 minutes ago, Mr.Meerkat said:

So basically, AMD possibly lied about both the Ryzen and 6900K? Damn son

Who would've thought that the same people who used a game to compare perfs would lie about a worlkload perf comparison!

Shocking.

[Mr.Meerkat]

1 minute ago, brokencross said:

Who would've thought that the same people who used a game to compare perfs would lie about a worlkload perf comparison!

Shocking.

It's more the fact that if AMD was lying about the performance, they're lying about the performance of the competitor's CPU as well which defeats the point  

 

@samcool55 Well that solves it

[iRileyx]

2 hours ago, JCBiggs said:

They need a completely new architecture.... There needs to be a third state after  1/0....until then , I'm not that impressed by these chips.  Lol

Zen/Ryzen whatever the hell they're calling it now, is a new architecture...

[brokencross]

4 minutes ago, Mr.Meerkat said:

It's more the fact that if AMD was lying about the performance, they're lying about the performance of the competitor's CPU as well which defeats the point  

 

@samcool55 Well that solves it

Does it, really? The comparison standards they used are shady to say the least. 

I'll wait for the retail cpus to hit the market for some unbiased perf benchmarks. Then we'll see if AMD has really done it or if we'll have to wait for the next """""big""""" thing.

[Clanscorpia]

3 minutes ago, RKRiley said:

Zen/Ryzen whatever the hell they're calling it now, is a new architecture...

Zen is the architecture Ryzen is the chip

[iRileyx]

Just now, Clanscorpia said:

Zen is the architecture Ryzen is the chip

Ah thankyou, i was unsure on it lol.

[Mr.Meerkat]

1 minute ago, brokencross said:

Does it, really? The comparison standards they used are shady to say the least. 

The whole point of it is to either show how your product is just as capable if not more capable that your competitor where increasing the performance of it's competitor is stupid  

[brokencross]

1 minute ago, Mr.Meerkat said:

The whole point of it is to either show how your product is just as capable if not more capable that your competitor where increasing the performance of it's competitor is stupid  

It's more like, I don't even know if they know what they're doing.

[samcool55]

1 minute ago, Mr.Meerkat said:

It's more the fact that if AMD was lying about the performance, they're lying about the performance of the competitor's CPU as well which defeats the point  

 

@samcool55 Well that solves it

Yup, just a small detail that they forgot to pay attention to and caused a little mess.

They said they'll update the profile with the correct amount of samples soon, maybe it's already up tbh i don't know

(just checked, they did update the profile, it's 150 samples now).

 

Also, someone took the time to compile a bunch of benchmarks into a nice sheet. It's far from complete but anyway

https://docs.google.com/spreadsheets/d/1O2Or6XOETZr3a4gLAuCuYEx8m5X_lv23LdwZiCZpBcM/edit#gid=0

 

And, to complete the list, full spec list of used systems!

So the dual-channel whatever stuff people talk about is likely irrelevant, at least that's what i think.

[SLAYR]

53 minutes ago, brokencross said:

You gotta mask that poor single core perf with more cores. It's the AMD way, friend

That and the 220w TDP.

Yeah it isn't going to be anywhere close to 200W, and all the rumors point to a 95W tdp for the eight core cpu, compared to 140W for something like a 6900k it should make intel look like power hungry.

 

If you want to rip on AMD for a 220W cpu here is a 260W, 72 core intel cpu.

http://ark.intel.com/m/products/95831/Intel-Xeon-Phi-Processor-7290F-16GB-1_50-GHz-72-core

[koolerone]

I'm personally not gonna believe any performance benchmarks I see until someone like LTT, Jay, or Kyle (bitwit) does benckmarks against intels CPUs....just sayin...

[brokencross]

1 minute ago, SLAYR said:

Yeah it isn't going to be anywhere close to 200W, and all the rumors point to a 95W tdp for the eight core cpu, compared to 140W for something like a 6900k it should make intel look like power hungry.

 

If you want to rip on AMD for a 220W cpu here is a 260W, 72 core intel cpu.

http://ark.intel.com/m/products/95831/Intel-Xeon-Phi-Processor-7290F-16GB-1_50-GHz-72-core

And to think a single consumer cpu 8 core FX 9590 is only short of 60w when compared to a 72 core intel enterprise chip.

Really makes you think.

 

Yeah, rumors and more rumors. That and the "wait for [insert AMD arc]" are just the bread and butter of any tech-related site. 

[Mira Yurizaki]

31 minutes ago, Mr.Meerkat said:

Was it crap IPC wise? Yep

Was it a completely useless arch? No. In programs that can utilise it's 8 cores (in the FX83xx), it can go toe to toe with a 6600K so...

While eating up measurably more power doing so. Also comparing an 8-core to a 4-core seems kind off.

 

For the desktop consumer, sure, Bulldozer had its appeal of being cheaper while offering similar performance to Intel's offerings if the programs used it. But Intel's offerings had more to offer elsewhere.

[TetraSky]

Considering how bad Bulldozer was and how much it was hyped as the best next thing, I'm gonna be waiting instead of getting hyped for possibly no reason.

It would be great if these claims by AMD were legitimate, but I'm not holding my breath.

[corrado33]

I will never trust benchmark results put on by the company advertising their own product. You know why? They sure as hell cherrypicked that chip. They may claim it's running at X.XX GHz, but when the chip comes out, it'll probably be stock Y.YY GHz (downclocked) so that the bulk of the chips actually work. They probably ran tests for weeks on what type of media would show off their type of chip the best. They may have even messed with the handbrake coding to make it work better with their chips. SO many things could be done to make sure their chip comes out on top.

 

These are all thing I expect to happen at places like this.

 

I'll wait for actual benchmarks by real people doing real workloads. 

 

And even so... it's great if AMD puts out a good chip. I mean, for years they haven't even touched Intel, that's not a lot of bragging rights. It's like the underdog team in sports finally winning a game when the favorite to win has won for the past 20 years. (and will probably continue to win for the next 20)

[brokencross]

Just now, TetraSky said:

Remember the AMD hype train when Bulldozer was coming out?
Wait for actual third party benchmarks before getting hyped over some product that will most likely disappoint. 

Pretty much this. AMD shamelessly cherrypicked blender benchmarks with bulldozer.

Not to mention we haven't yet seen a single core benchmark from this new arc.

[Mira Yurizaki]

So my last post reminded me of a point: AMD poses a serious threat against Intel only if all markets are interested in using them.

 

AMD was a serious threat against Intel during 2000-2006 because AMD was being used in practically all markets, which included the laptop market (until the Pentium M came out) and the server market (which loved the Opteron until the last quarter of the 2000s).

 

If AMD can't win the mobile (i.e laptops), workstation, or server markets over, they're back to square one.

[corrado33]

Wait, they used DOTA2 for gaming performance... I don't play, but quick search says it mainly depends on core speed and not number of cores? 

 

I mean, wtf settings were they using and where were they streaming to that they could properly max out those CPUs?

 

I thought 720p streaming to twitch was useless because of the bandwidth cap?

 

[Mr.Meerkat]

22 minutes ago, M.Yurizaki said:

While eating up measurably more power doing so. Also comparing an 8-core to a 4-core seems kind off.

And? I'm comparing a 4-5 year older CPU to something relatively new which also basically costs double the FX...I don't see why I can't compare something that's a lot cheaper...

[corrado33]

Also, another question.

 

Why was AMD's turbo boost disabled? That seems.... stupid. Almost like it wasn't working... or the chip was already running at the top of its range. 

 

You don't bring a half tuned racecar to the track and say "well if we tuned it all of the way it'd blow the PANTS off of the competition." No one would believe you.

 

And I don't believe AMD. 

 

There was something going on with their turbo boost functionality, that's why it was disabled. 

[creationex]

3 hours ago, JCBiggs said:

They need a completely new architecture.... There needs to be a third state after  1/0....until then , I'm not that impressed by these chips.  Lol

hello quantum chips

 

[dizmo]

3 hours ago, maria2244 said:

if AMD can sell their 6900k RYZEN cpu

for 300 or 400 $

no one will buy i5 or i7 from intel anymore

we have 6900k AMD version

It won't be that cheap. I'd imagine it'll be around $700.

Simple fact is that AMD is a business, and they're in it to make money. It doesn't make sense to price yourself so insanely low, when you can easily price it higher and still garner sales. Especially with how poorly they've done financially over the past few years, and how much R&D likely went into producing Zen.

They'll also I imagine have lower end CPUs, to match the 6700k, 6600k, etc.

3 hours ago, TidaLWaveZ said:

Definitely, although I'm not too sure of the magnitude of their price hikes. I've been paying around the same price (~$300) for my desktop processors for over fifteen years whether it's intel or AMD.  Either way true competition should drive prices down regardless.

I don't think we'll see huge price wars. I think if anything AMD will price themselves slightly lower than Intel.

36 minutes ago, SLAYR said:

Yeah it isn't going to be anywhere close to 200W, and all the rumors point to a 95W tdp for the eight core cpu, compared to 140W for something like a 6900k it should make intel look like power hungry.

 

If you want to rip on AMD for a 220W cpu here is a 260W, 72 core intel cpu.

http://ark.intel.com/m/products/95831/Intel-Xeon-Phi-Processor-7290F-16GB-1_50-GHz-72-core

The TDP isn't a rumor, that's what it is.

Also you can't compare those two CPUs. That's an incredibly unfair comparison. Simple fact is AMD did a pretty poor job with the 9590.

30 minutes ago, brokencross said:

And to think a single consumer cpu 8 core FX 9590 is only short of 60w when compared to a 72 core intel enterprise chip.

Really makes you think.

 

Yeah, rumors and more rumors. That and the "wait for [insert AMD arc]" are just the bread and butter of any tech-related site. 

Fact, actually. 95w TDP for that chip.

I think we should all wait for reviews before saying it's amazing though. AMD tends to bend truths.

[ARikozuM]

2 hours ago, SpaceGhostC2C said:

How many topics do we have about the same AMD event?

Do you want the additive or exponential function?