AMD Ryzen HAS BEEN ANNOUNCED!!!

[N3v3r3nding_N3wb]

11 hours ago, AnonymousGuy said:

Here's the ways AMD can fuck it up:

 

• IPC being shit.  Hey no 6700K or 7700K single threaded comparisons?  Mmmmkay.  Y'all realize you pretty much never utilize 8 cores?
• Terrible yields.  March 2nd launch becomes June delivery.  Historically happens to their graphics cards.
• Being clocked already to the edge.  100Mhz "turbo"?  Really bruh?
• Mobo's lacking features.  First time AMD has put out anything natively supporting >USB2.0.

1. Basically on par with Broadwell (Broadwell is about 3% behind Skylake), according to AMD -- with no indication that they're lying.  

I'm going to assume you mistyped and meant single-threaded performance, not IPC. Assuming that, obviously the higher clocked part with marginally better IPC will be superior in single-threaded tasks, but the benefit of the multiple cores will most likely outweigh slightly better single-core performance for the future.

 

2. Iirc, they're using the same manufacturing process as on Polaris, so it's a mature process with reportedly very good yields, and they've stated (however much faith you wanna put in them) that there should be no availability issues whatsoever.

 

3. The 100 MHz "boost" isn't the boost, it's the guaranteed XFR benefit.

 

4. Motherboards are basically on par with Intel's, but should be cheaper because Ryzen has most of the IO integrated into the chip.

 

 

I understand and support still being a bit skeptical of performance and waiting for reviews, but even just paying attention to the announcements rule out three of your potential problems, and the other one was debunked by a small amount of research.

Edited February 23, 2017 by N3v3r3nding_N3wb
Minor wording edit

[Techstorm970]

Fun fact: The i3-6300 scores a 162 on Cinebench R15 single-thread, the same as the Ryzen 7 1800X at AMD's event today.

[JoostinOnline]

2 minutes ago, techstorm970 said:

Fun fact: The i3-6300 scores a 162 on Cinebench R15 single-thread, the same as the Ryzen 7 1800X at AMD's event today.

Matching an i3 in single threaded processes is actually a pretty big accomplishment.

[Techstorm970]

2 minutes ago, JoostinOnline said:

Matching an i3 in single threaded processes is actually a pretty big accomplishment.

It also proves that Ryzen will be a great option for gamers! 

[MadyTehWolfie]

13 hours ago, DocSwag said:

Well of course .

 

But cinebench already gives us a good idea of the approximate performance, so I wouldn't be surprised if the 1800x is basically the equivalent of the 6900k in games.

True but if it ends up like what I think happened in fallout 4 where something with AMD cpu's wasnt working right it may not provide the performance the synthetic benchmarks show. Think Ryzen is going to deliver though finally some competition for a company that got too big for their boots.

[MakeAMDGreatAgain]

4 minutes ago, techstorm970 said:

It also proves that Ryzen will be a great option for gamers! 

Are you being sarcastic?

[Techstorm970]

1 minute ago, MakeAMDGreatAgain said:

Are you being sarcastic?

No....

[MakeAMDGreatAgain]

3 minutes ago, techstorm970 said:

No....

Yes..? 

[Techstorm970]

2 minutes ago, MakeAMDGreatAgain said:

Yes..? 

[Drak3]

6 hours ago, Orangeator said:

Generally the conscientious is the more cores you have within the same architecture the less the single core performance is. I could prove this over and over again.

Except that the 6700K and 6800K are not the same architecture. The first is Skylake, the second is Broadwell, and Skylake has better per clock performance than Broadwell, if only a slight edge.

The Hexacore Broadwell is at two disadvantages when compared to the quadcore Skylake.

 

6 hours ago, techstorm970 said:

NOT TRUE

At stock, there's some truth to modern Intel quad core parts having better single core performance than their hexa, octa, and potentially decacore counterparts. This is exclusively due to lower clock speeds to meet power consumption and heat output targets, and this becomes untrue the nanosecond one brings up overclocking.

I mean, look at the 4770K vs the 5820K. Same architecture, but the 4770K's single core performance is higher due to it being able to hit a slightly faster clock speed whilst meeting TDP requirements

[Sprawlie]

Very interesting. Lot's of 'IF''s .

 

 

IF, what they say is right, and that the r7-1700 is "On Par" with the i7-7700k, (which is what they often charted it beside in their press conference for multi-core.

 

and IF: it also matches on par the single threaded performance of the 7700k for gaming purposes

 

and so far, in CAD$ is looking to be almost $100 less than the 7700k. That is going to be a very, Very compelling argument. Though, Now for the price i was willing to spend on the i7 before it went on sale of $560, i'm thinking why not go 1800x.

 

But i'm not purchasing a damned thing until real world benchmarks, performance tests, and good quality real world QA happens.

 

I'm picky 

[Drak3]

Also, in case anyone is interested, there's been a single benchmark run on the 1700X on UserBenchmark. Should it be credible, then it looks bright on AMD delivering on their promises: Between Haswell and Broadwell levels of IPC.

 

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-Intel-Core-i7-5960X/m243187vs2580

[Orangeator]

25 minutes ago, Sprawlie said:

Very interesting. Lot's of 'IF''s .

 

 

IF, what they say is right, and that the r7-1700 is "On Par" with the i7-7700k, (which is what they often charted it beside in their press conference for multi-core.

 

and IF: it also matches on par the single threaded performance of the 7700k for gaming purposes

 

and so far, in CAD$ is looking to be almost $100 less than the 7700k. That is going to be a very, Very compelling argument. Though, Now for the price i was willing to spend on the i7 before it went on sale of $560, i'm thinking why not go 1800x.

 

But i'm not purchasing a damned thing until real world benchmarks, performance tests, and good quality real world QA happens.

 

I'm picky 

In multi-threaded the 1700 outperforms the 7700k by 50%... What the hell are you are you talking about on par?!?!

[Techstorm970]

50 minutes ago, Drak3 said:

Also, in case anyone is interested, there's been a single benchmark run on the 1700X on UserBenchmark. Should it be credible, then it looks bright on AMD delivering on their promises: Between Haswell and Broadwell levels of IPC.

 

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-Intel-Core-i7-5960X/m243187vs2580

Single-core performance is on par with i3-6300, just like in Cinebench R15.

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-Intel-Core-i3-6300/m243187vs3536

Based on this, it's probably credible.

[Drak3]

2 minutes ago, techstorm970 said:

Single-core performance is on par with i3-6300, just like in Cinebench R15.

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-Intel-Core-i3-6300/m243187vs3536

Based on this, it's probably credible.

It's a tiny tad slower than the Sky i3, about the same as a similarly clocked Haswell or Broadwell. Which, again, all AMD actually promised.

[Techstorm970]

53 minutes ago, Drak3 said:

Also, in case anyone is interested, there's been a single benchmark run on the 1700X on UserBenchmark. Should it be credible, then it looks bright on AMD delivering on their promises: Between Haswell and Broadwell levels of IPC.

 

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-Intel-Core-i7-5960X/m243187vs2580

RYZEN 1700X ON PAR WITH i5-7500!!! (single & quad core performance)

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-Intel-Core-i5-7500/m243187vs3648

[Techstorm970]

Ryzen 7 1700X vs FX-9590

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-AMD-FX-9590/m243187vs1812

[Drak3]

3 minutes ago, techstorm970 said:

Ryzen 7 1700X vs FX-9590

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-AMD-FX-9590/m243187vs1812

I'm mostly amused by the fact that the 9590's turbo speeds in the benchmark are lower than the base clocks.

[DocSwag]

2 minutes ago, techstorm970 said:

Ryzen 7 1700X vs FX-9590

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-AMD-FX-9590/m243187vs1812

Dat 52 % IPC doing its work  

[N3v3r3nding_N3wb]

11 hours ago, Prysin said:

sure it will be close, in some cases, others not. But you have to look at teh whole ecosystem. Look at the boards, the partners...

With this big a launch, you expect the board partners to bring their BEST boards out at launch. You know, the "Maximus EXTREME" brand. The top end boards. The expensive boards...

And what is on display? two tiers/generations behind intels current offerings... Sure there will be more feature rich boards eventually. But nobody is going to wait 6-18 months for a better board. They want RYZEN now, when there is hype.

 

RYZEN is doomed to flop. Even if the chip is pretty good, the boards, the overclocking, the actual TDP (please, as if 8 cores 16 threads is 95w in Prime95 or during Folding at home).  AMD also raised the bar for cooling. Given their decent stock cooling, you cannot just grab a 50$ cooler, because it will only improve things marginally, instead you need to go directly to 75-100+$ coolers... Which means any serious overclocking is going to cost you more then what you currently need on Intels offerings.

"The top end boards. The expensive boards...

And what is on display? two tiers/generations behind intels current offerings..." From what I can see on every single thing that I've read, they have practically reached parity with Intel. (http://www.pcworld.com/article/3154720/computers/amd-reveals-an-army-of-ryzen-pcs-and-am4-motherboards.html, https://www.asus.com/us/Product-Compare/?products=s7csF4UqmJykbmtW,Jp5t0js638s6sbJG&b=0).

 

"please, as if 8 cores 16 threads is 95w in Prime95 or during Folding at home" Just because it hasn't happened yet doesn't mean it can't happen, and even if it does have a higher real TDP, I'm sure it won't be 45 watts more than 95 (I can assert that with no sources because you didn't have any sources, and a claim with no evidence can be dismissed without evidence).

 

"AMD also raised the bar for cooling. Given their decent stock cooling, you cannot just grab a 50$ cooler, because it will only improve things marginally, instead you need to go directly to 75-100+$ coolers... Which means any serious overclocking is going to cost you more then what you currently need on Intel's offerings." Are you saying that AMD's chips, with a lower TDP (so presumably lower temps) will require better cooling to be as cool?  Better stock cooling is a GOOD thing, not a bad thing.  Their better stock cooling, if anything, can completely remove the need for an aftermarket cooler -- it (actually the original Wraith, the older version of what're shipped with these) is roughly as good as a 50 dollar aftermarket cooler (http://www.tomshardware.com/reviews/amd-wraith-cpu-cooler,4450-2.html).

[Techstorm970]

10 minutes ago, Drak3 said:

I'm mostly amused by the fact that the 9590's turbo speeds in the benchmark are lower than the base clocks.

Should I say "Agree"

or "Funny"?

...

...

...

WHY CAN'T I CHOOSE BOTH?!?!?! 

 

[goodtofufriday]

21 minutes ago, techstorm970 said:

Ryzen 7 1700X vs FX-9590

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-1700X-Turbo-Disabled-vs-AMD-FX-9590/m243187vs1812

I say AMD should give all 9590 users a free 1800x for sticking with it.... 

 

I actually asked on their twitter lol. 

[Lays]

2 hours ago, N3v3r3nding_N3wb said:

1. Basically on par with Broadwell (Broadwell is about 3% behind Skylake), according to AMD -- with no indication that they're lying.  

I'm going to assume you mistyped and meant single-threaded performance, not IPC. Assuming that, obviously the higher clocked part with marginally better IPC will be superior in single-threaded tasks, but the benefit of the multiple cores will most likely outweigh slightly better single-core performance for the future.

 

2. Iirc, they're using the same manufacturing process as on Polaris, so it's a mature process with reportedly very good yields, and they've stated (however much faith you wanna put in them) that there should be no availability issues whatsoever.

 

3. The 100 MHz "boost" isn't the boost, it's the guaranteed XFR benefit.

 

4. Motherboards are basically on par with Intel's, but should be cheaper because Ryzen has most of the IO integrated into the chip.

 

 

I understand and support still being a bit skeptical of performance and waiting for reviews, but even just paying attention to the announcements rule out three of your potential problems, and the other one was debunked by a small amount of research.

 

2 hours ago, techstorm970 said:

Fun fact: The i3-6300 scores a 162 on Cinebench R15 single-thread, the same as the Ryzen 7 1800X at AMD's event today.

 

I hope it clocks well, because I wouldn't mind trying the high-end chip out.   But a LOT of the games I play only use a few threads, and my single thread is absolutely huge with my 24/7 OC.  So if these chips don't clock well, having "similar" IPC doesn't really matter a whole lot in tasks that can't make full use of their extra cores. 

 

[VanayadGaming]

13 minutes ago, Lays said:

 

 

I hope it clocks well, because I wouldn't mind trying the high-end chip out.   But a LOT of the games I play only use a few threads, and my single thread is absolutely huge with my 24/7 OC.  So if these chips don't clock well, having "similar" IPC doesn't really matter a whole lot in tasks that can't make full use of their extra cores. 

 

Well, I'm very interested in the Multi-threading part and how it goes. Being able to stream properly will be a big plus. Also as I'm usually playing Grad Strategy games... does are very CPU bound... so win-win for me  As in all, it really is dependent on the task at hand. 

[Vode]

The Ryzen 1800x seems to be exactly the same as my 5960x @ 4Ghz in Cinebench. Single core was 158 also @ 4Ghz.