Discussion: AMD Ryzen

[-Fran-]

Because the only known silicon is 2.8GHz. to people outside AMD

Because they set the frequency to 3GHz without turbo when running Blender.to make a valid IPC comparison

Because they had to underclock the Broadwell sample.because their ES was running at 3Ghz maybe?

Because if they had final silicon they wouldn't announce a six month delay of the chip.final silicon before the full QA process is called ES still, right?

When pressed about the TDP of final silicon, AMD said "comparable to Broadwell".

I don't see how any of that justifies saying AMD had to overclock an ES and I don't see why that answer from AMD in regards to TDP is so terrible (keeping what I said in perspective).

Cheers!

 

[juanrga]

There is zero evidence an "overclocked ES" was used. Less misinformation please!

I agree ... there is no evidence of this.

Hmm ... J once again you are making statements without the facts to substantiate claims.

J has provided enough evidence and logical thinking to sustain his claims. It is Y who claims that AMD used a 3GHZ ES but he didn't give evidence on the existence of 3GHz samples.

The existence of 2.8GHz engineering samples can be proven. Even the internal code of the 2.8GHz samples, including the steeping iteration, can be pasted here for verification. Can Y prove the existence of a 3GHz samples?

 

[juanrga]

None of this is terribly promising, IMO. If we look at what people are buying, it's the high-clocked, high-IPC quad-core chips that Intel sells the most of. Haswell-E, and Broadwell-E absolutely crush consumer Skylake parts in multi-threaded performance, but that doesn't stop people from choosing the 6700K, a single SKU, more than 5 times as often as all Broadwell-E parts combined and more than 3 times as often as all Haswell-E parts combined. (6700K market share 5.2%, all BW-E 0.9%, all HW-E 1.4% - source: Userbenchmark). They are going to have do better than matching the performance profile of a CPU that no one is buying. If they could release a quad-core Zen clocked at 5GHz, they'd have a slam-dunk, but it looks like we are nowhere near that.

The 65W quad-core ES are clocked at 2.8GHz. I expect final silicon to hit 3.4GHz more or less. 5GHz will not achieved on a mobile oriented process node. Moreover AMD already confirmed at hot chips that Zen uses libraries optimized for efficiency.

 

[-Fran-]

There is zero evidence an "overclocked ES" was used. Less misinformation please!

I agree ... there is no evidence of this.

Hmm ... J once again you are making statements without the facts to substantiate claims.

J has provided enough evidence and logical thinking to sustain his claims. It is Y who claims that AMD used a 3GHZ ES but he didn't give evidence on the existence of 3GHz samples.

No, I am not. I am saying you don't know and no matter how much conjecture based on outgoing facts you gather, there are things you can't use as "fact", since you don't know what AMD is doing, just like most of the internet now.

I am giving you something that is called "reasonable doubt".

Cheers!

 

[8350rocks]

Spoiler

But there isn't much for them to benefit by doing that level of mucking about. It'll just cause bad press and they would still have a lousy chip.

Like when they promised a 1050GFLOPS Kaveri but shipped a sub-900 GFLOPS Kaveri? Like when they promised the moon for Carrizo and showed several designs Carrizo wins that latter didn't appear in any store? Like when they said that certain card was a "overclockers' dream", just before reviews showed how bad was the overclocking? Like when they promised a 16-core Seattle but gave us a 8-core Seattle after a two years delay? Like when they used odd settings benchmarks for hyping the 300 series? Like when they promised us again and again that Zen was a 2016 product, (funny enough I have known for a while it was 2017 and I have been saying it in forums), then Digitimes posted a rumor about AMD delaying Zen to 2017, AMD reacted attacking DigiTimes saying that the rumor was false, and now AMD confirms a delay of Zen to 2017? And that is the short list.

It is evident to me that they have cherry picked the benchmark for Zen vs Broadwell and that final reviews will show lots of benchmarks where Zen is outperformed. I have a 100% certainty on this. Mark my words.

See if your so anti-AMD then maybe you should post elswhere Juan?

The last thing we need is another AMD vs Intel war.

Mark my words you sir will be the first casualty ... along with any of the others who chose to start and engage in any slinging matches.

Since you don't actually have any facts beyond a video AMD released and a couple of slides, it seems highly unlikely that you can accurately extrapolate, in any way, shape or form ... the real performance of these new CPU's.

Whilst you are free here to make predictions, I feel it is important to point out that going on your past history, when you were a committed AMD fanboi, you didn't exactly prove an accurate predictor of the performance results in the end.

If your still angry about being jilted in some way then you need to move on ... accept that all manufacturers spin things up a bit, and spend some time getting lost in a good game.

I recommend FreeLancer ... because I am oldskool geek.

I find the music soothing ...

I am not anti-AMD, see my signature, but I am anti-hype and misinformation. And my comments aren't mean to be interpreted as some AMD vs Intel war. I am replying to demo of Zen vs Broadwell made by AMD. If they had made a Zen vs Piledriver demo and claimed that Zen is ~3x faster clock-for-clock I would be reacting in exactly the same way. To put things in perspective

According to AMD Zen @3GHz would be somewhat in the 100 second mark, whereas the FX-8350 is 320 seconds. This implies that Zen would have ~4.26x higher IPC than Piledriver (recall AMD own slides stating that Zen is ~2x faster than FX-8350 clock-for-clock).

The Stilt have made a similar analysis. He downloaded last version of Blender and benchmarked his Piledriver and Haswell chips. He tested single thread and found that Zen core would be 140% faster than PD core to match his Haswell. According to AMD Zen is ~40% faster than XV; therefore XV would be about 100% faster than PD. Numbers don't match by a huge amount. He claims to be puzzled by the result and he got similar conclusions than me.

We are not talking about a small "spin", about a 10% here and a 5% there to put products in a better shape. We are talking about huge gaps . What are we supposed to do? Say "yes" to anything published/advertised by companies and don't use our brain? Then why the forums? The news section would be enough.

And my facts aren't a video demo, but the details of the microarchitectures, and the good record of predictions that I have made about Zen. A good amount of the information in those slides you mention was posted by me here before those slides were even made by the marketing goods.

That is actually not so hard to believe...

Consider this...

8C zen has twice as many FPUs. So, assuming they are equal, we are already at +100% performance...now...we can assume +40% per AMDs estimates over piledriver, and finally, we can also include the fact that zen can run AVX, but piledriver cannot. Which would put another theoretical 60% blue sky for benchmarks in there.

So, we end up +200% total, which is not far off your calculations (20% difference...).

Doesn't work that way.

First, 200% more is not enough because the performance gap is higher than 4x.

Second, you are counting the same performance gains twice. Zen has 4 ALU per core, Piledriver has 2 ALU. This means that Zen has the ability to execute up to four integer instructions per cycle, which implies twice more throughput than Piledriver. But this is a peak performance, because Zen only has 2 AGUs (the same than Piledriver) and cannot provide data to sustain four ALUs each cycle. This simply implies that sustained performance gain will be inferior to 2x. We can do a cheap estimation.

Zen: 4ALU + 2AGU = 6 execution units
PD: 2ALU + 2AGU = 4 execution units

6/4 = 1.5%, which implies Zen could be about 50% faster, in sustained workloads, than Piledriver. The actual computation is more complex and has to account for other details including the frequency of use of each unit (ALU, AGU) in real code, but ~50% more than Piledriver is close to the expected performance for Zen.

Similar remarks for floating point. Zen is 16FLOP/core and has 2x peak throughput than Piledriver (8FLOP/core), but on sustained floating-point workloads Zen will be ~70% faster than Piledriver. The same happen on Intel side. Haswell core has 2x the number of FP resources than Ivy Bridge core, but it is only 70% faster (sustained performance) clock for clock.

In code that mixes integer and float we can obtain an average, ~60% over Piledriver, which roughly corresponds to the 40% over Excavator officially claimed by AMD.

You are taking the peak throughput gain (2x) and adding average throughput gain (1.4x) on top of that. You are counting twice.

Finally, I don't know what you mean by Piledriver not supporting AVX. AMD has supported AVX since Bulldozer

http://developer.amd.com/community/blog/2009/05/06/striking-a-balance/

If you want to count ALUs...

Zen has 32 ALUs while PD has 16 ALUs

Zen has 16x 128 bit FPUs while PD has 8x 128 bit FPUs

So...there is +200% performance alone right there purely in execution units.

 

[juanrga]

Because the only known silicon is 2.8GHz. to people outside AMD

Because they set the frequency to 3GHz without turbo when running Blender.to make a valid IPC comparison

Because they had to underclock the Broadwell sample.because their ES was running at 3Ghz maybe?

Because if they had final silicon they wouldn't announce a six month delay of the chip.final silicon before the full QA process is called ES still, right?

When pressed about the TDP of final silicon, AMD said "comparable to Broadwell".

I don't see how any of that justifies saying AMD had to overclock an ES and I don't see why that answer from AMD in regards to TDP is so terrible (keeping what I said in perspective).

1) AMD is not a foundry. The engineering samples have to be fabricated outside of AMD and then shipped back to AMD. We can even track the samples using import databases. This is manifesto for engineering samples traveling to India

https://www.zauba.com/import-microprocessors-zeppelin-hs-code.html

There is no information either public or private about 3GHz samples.

2) That doesn't explain why AMD didn't overclock it to 3.2GHz, instead underclocking the other chip.

3) The final version of the ES has the same silicon that is used in the final product.

4) We have known for months the official rating of the AM4 platform: 95W. We have known for a while the TDP of the last engineering samples (you can see it above in the import database where it says "AM4-95W_8C_16T_PROTO-SILA00"). A pair of days before the Blender demo, AMD shared the official roadmap with partners. I have a copy of the whole presentation, but the roadmap was leaked

It shows 95W chips is the maximum for the desktop.

It makes no sense that AMD refuted to answer us which was the TDP of the sample they used in the Blender demo and the pretension that "they said giving the TDP was an strategic thing for them, so they wouldn't inform it until final silicon is ready" doesn't make sense because everyone knows that the TDP is 95W. It makes no sense to hide by "strategic reasons" something is public and well-known.

My alternative explanation however fits on all the available data as a glove:

A) The only existent ES are clocked at 2.8GHz and they can be overclocked.

B) The Broadwell chip was underclocked because the existent Zen sample cannot be overclocked at 3.2Ghz.

C) The sample overclocked at 3GHz would dissipate about 117W, whereas BDW-E underclocked at 3GHz would dissipate about 123W (both computations made by myself). During the demo AMD confirmed that the Zen sample was consuming a bit less power than the BDW-E sample. Well, 117W is a bit less than 123W.

D) AMD refuted to give TDP numbers to the public during the demo, because that 3GHz sample was running above the official 95W rating of the AM4 platform.

E) Their goal is to achieve 3GHz or higher clocks on a 95W rating. This is not possible with current silicon (3GHz <--> ~117W); thus AMD delayed Zen to wait to Glofo to mature the 14LPP node and get higher clocks to make the chip more competitive.

F) My claims are reasonable and backed up by available data (both public and behind doors). I will not insist on this. Anyone is free to believe me or make their own claims.

 

[juanrga]

Spoiler

But there isn't much for them to benefit by doing that level of mucking about. It'll just cause bad press and they would still have a lousy chip.

Like when they promised a 1050GFLOPS Kaveri but shipped a sub-900 GFLOPS Kaveri? Like when they promised the moon for Carrizo and showed several designs Carrizo wins that latter didn't appear in any store? Like when they said that certain card was a "overclockers' dream", just before reviews showed how bad was the overclocking? Like when they promised a 16-core Seattle but gave us a 8-core Seattle after a two years delay? Like when they used odd settings benchmarks for hyping the 300 series? Like when they promised us again and again that Zen was a 2016 product, (funny enough I have known for a while it was 2017 and I have been saying it in forums), then Digitimes posted a rumor about AMD delaying Zen to 2017, AMD reacted attacking DigiTimes saying that the rumor was false, and now AMD confirms a delay of Zen to 2017? And that is the short list.

It is evident to me that they have cherry picked the benchmark for Zen vs Broadwell and that final reviews will show lots of benchmarks where Zen is outperformed. I have a 100% certainty on this. Mark my words.

See if your so anti-AMD then maybe you should post elswhere Juan?

The last thing we need is another AMD vs Intel war.

Mark my words you sir will be the first casualty ... along with any of the others who chose to start and engage in any slinging matches.

Since you don't actually have any facts beyond a video AMD released and a couple of slides, it seems highly unlikely that you can accurately extrapolate, in any way, shape or form ... the real performance of these new CPU's.

Whilst you are free here to make predictions, I feel it is important to point out that going on your past history, when you were a committed AMD fanboi, you didn't exactly prove an accurate predictor of the performance results in the end.

If your still angry about being jilted in some way then you need to move on ... accept that all manufacturers spin things up a bit, and spend some time getting lost in a good game.

I recommend FreeLancer ... because I am oldskool geek.

I find the music soothing ...

I am not anti-AMD, see my signature, but I am anti-hype and misinformation. And my comments aren't mean to be interpreted as some AMD vs Intel war. I am replying to demo of Zen vs Broadwell made by AMD. If they had made a Zen vs Piledriver demo and claimed that Zen is ~3x faster clock-for-clock I would be reacting in exactly the same way. To put things in perspective

According to AMD Zen @3GHz would be somewhat in the 100 second mark, whereas the FX-8350 is 320 seconds. This implies that Zen would have ~4.26x higher IPC than Piledriver (recall AMD own slides stating that Zen is ~2x faster than FX-8350 clock-for-clock).

The Stilt have made a similar analysis. He downloaded last version of Blender and benchmarked his Piledriver and Haswell chips. He tested single thread and found that Zen core would be 140% faster than PD core to match his Haswell. According to AMD Zen is ~40% faster than XV; therefore XV would be about 100% faster than PD. Numbers don't match by a huge amount. He claims to be puzzled by the result and he got similar conclusions than me.

We are not talking about a small "spin", about a 10% here and a 5% there to put products in a better shape. We are talking about huge gaps . What are we supposed to do? Say "yes" to anything published/advertised by companies and don't use our brain? Then why the forums? The news section would be enough.

And my facts aren't a video demo, but the details of the microarchitectures, and the good record of predictions that I have made about Zen. A good amount of the information in those slides you mention was posted by me here before those slides were even made by the marketing goods.

That is actually not so hard to believe...

Consider this...

8C zen has twice as many FPUs. So, assuming they are equal, we are already at +100% performance...now...we can assume +40% per AMDs estimates over piledriver, and finally, we can also include the fact that zen can run AVX, but piledriver cannot. Which would put another theoretical 60% blue sky for benchmarks in there.

So, we end up +200% total, which is not far off your calculations (20% difference...).

Doesn't work that way.

First, 200% more is not enough because the performance gap is higher than 4x.

Second, you are counting the same performance gains twice. Zen has 4 ALU per core, Piledriver has 2 ALU. This means that Zen has the ability to execute up to four integer instructions per cycle, which implies twice more throughput than Piledriver. But this is a peak performance, because Zen only has 2 AGUs (the same than Piledriver) and cannot provide data to sustain four ALUs each cycle. This simply implies that sustained performance gain will be inferior to 2x. We can do a cheap estimation.

Zen: 4ALU + 2AGU = 6 execution units
PD: 2ALU + 2AGU = 4 execution units

6/4 = 1.5%, which implies Zen could be about 50% faster, in sustained workloads, than Piledriver. The actual computation is more complex and has to account for other details including the frequency of use of each unit (ALU, AGU) in real code, but ~50% more than Piledriver is close to the expected performance for Zen.

Similar remarks for floating point. Zen is 16FLOP/core and has 2x peak throughput than Piledriver (8FLOP/core), but on sustained floating-point workloads Zen will be ~70% faster than Piledriver. The same happen on Intel side. Haswell core has 2x the number of FP resources than Ivy Bridge core, but it is only 70% faster (sustained performance) clock for clock.

In code that mixes integer and float we can obtain an average, ~60% over Piledriver, which roughly corresponds to the 40% over Excavator officially claimed by AMD.

You are taking the peak throughput gain (2x) and adding average throughput gain (1.4x) on top of that. You are counting twice.

Finally, I don't know what you mean by Piledriver not supporting AVX. AMD has supported AVX since Bulldozer

http://developer.amd.com/community/blog/2009/05/06/striking-a-balance/

If you want to count ALUs...

Zen has 32 ALUs while PD has 16 ALUs

Zen has 16x 128 bit FPUs while PD has 8x 128 bit FPUs

So...there is +200% performance alone right there purely in execution units.

I already mentioned that 200% more performance cannot explain a performance gap higher than 4x.

I have also explained why having 2x more ALUs per core implies about 50% more performance.

The reason why 8C/16T Zen could achieve in certain benchmarks 2x more performance than 8C/8T Piledriver is not because Zen has 2x more ALUs, that is a coincidence. The reasons are that those 4ALUs can provide about 50% more performance per core plus Zen has SMT and can use the ALUs for additional threads when the main thread doesn't use them, plus Piledriver has a CMT penalty from sharing the front-end (fetch units, decoder) and caches L1, L2, that reduces performance when each module runs two threads. Resume:

2x more ALUs: 50%
SMT: 20%
CMT: 20%

Total: ~2x

This ~2x is for multithreaded workloads. For single thread workloads SMT doesn't work and the CMT penalty doesn't apply, which means that Zen core will be about 50% faster than Piledriver core (despite having twice the number of ALUs). For floating point benchmarks I estimate that Zen core could be about 70% faster than Piledriver (despite having twice wider FPUs).

 

[-Fran-]

Spoiler

1) AMD is not a foundry. The engineering samples have to be fabricated outside of AMD and then shipped back to AMD. We can even track the samples using import databases. This is manifesto for engineering samples traveling to India

https://www.zauba.com/import-microprocessors-zeppelin-hs-code.html

There is no information either public or private about 3GHz samples.

2) That doesn't explain why AMD didn't overclock it to 3.2GHz, instead underclocking the other chip.

3) The final version of the ES has the same silicon that is used in the final product.

4) We have known for months the official rating of the AM4 platform: 95W. We have known for a while the TDP of the last engineering samples (you can see it above in the import database where it says "AM4-95W_8C_16T_PROTO-SILA00"). A pair of days before the Blender demo, AMD shared the official roadmap with partners. I have a copy of the whole presentation, but the roadmap was leaked

It shows 95W chips is the maximum for the desktop.

It makes no sense that AMD refuted to answer us which was the TDP of the sample they used in the Blender demo and the pretension that "they said giving the TDP was an strategic thing for them, so they wouldn't inform it until final silicon is ready" doesn't make sense because everyone knows that the TDP is 95W. It makes no sense to hide by "strategic reasons" something is public and well-known.

My alternative explanation however fits on all the available data as a glove:

A) The only existent ES are clocked at 2.8GHz and they can be overclocked.

B) The Broadwell chip was underclocked because the existent Zen sample cannot be overclocked at 3.2Ghz.

C) The sample overclocked at 3GHz would dissipate about 117W, whereas BDW-E underclocked at 3GHz would dissipate about 123W (both computations made by myself). During the demo AMD confirmed that the Zen sample was consuming a bit less power than the BDW-E sample. Well, 117W is a bit less than 123W.

D) AMD refuted to give TDP numbers to the public during the demo, because that 3GHz sample was running above the official 95W rating of the AM4 platform.

E) Their goal is to achieve 3GHz or higher clocks on a 95W rating. This is not possible with current silicon (3GHz <--> ~117W); thus AMD delayed Zen to wait to Glofo to mature the 14LPP node and get higher clocks to make the chip more competitive.

F) My claims are reasonable and backed up by available data (both public and behind doors). I will not insist on this. Anyone is free to believe me or make their own claims.

That import manifesto is only for India (from what I could read). It also doesn't state anywhere that it represents the *whole* Country nor all of its exports. I see no GF Fab in India. AFAIK, AMD uses the fab in Dresden. I also searched for import and export manifestos in Germany and the USA, but could not find anything that could prove nor disprove your point either. AMD might have not used a 3rd party for transport and just sent someone to pick the ES up (not likely, but not impossible either) since it can be considered "extremely classified" information (or whatever label you want).

For your conjecture around the TDP: yeah, those numbers are really feasible and it points in the direction you suggest. I agree with that for what it's worth. I just don't agree on the "overclocked ES" bit.

And your claims are always reasonable, Juan. They aren't "facts" though.

Cheers!

 

[8350rocks]

Spoiler

1) AMD is not a foundry. The engineering samples have to be fabricated outside of AMD and then shipped back to AMD. We can even track the samples using import databases. This is manifesto for engineering samples traveling to India

https://www.zauba.com/import-microprocessors-zeppelin-hs-code.html

There is no information either public or private about 3GHz samples.

2) That doesn't explain why AMD didn't overclock it to 3.2GHz, instead underclocking the other chip.

3) The final version of the ES has the same silicon that is used in the final product.

4) We have known for months the official rating of the AM4 platform: 95W. We have known for a while the TDP of the last engineering samples (you can see it above in the import database where it says "AM4-95W_8C_16T_PROTO-SILA00"). A pair of days before the Blender demo, AMD shared the official roadmap with partners. I have a copy of the whole presentation, but the roadmap was leaked

It shows 95W chips is the maximum for the desktop.

It makes no sense that AMD refuted to answer us which was the TDP of the sample they used in the Blender demo and the pretension that "they said giving the TDP was an strategic thing for them, so they wouldn't inform it until final silicon is ready" doesn't make sense because everyone knows that the TDP is 95W. It makes no sense to hide by "strategic reasons" something is public and well-known.

My alternative explanation however fits on all the available data as a glove:

A) The only existent ES are clocked at 2.8GHz and they can be overclocked.

B) The Broadwell chip was underclocked because the existent Zen sample cannot be overclocked at 3.2Ghz.

C) The sample overclocked at 3GHz would dissipate about 117W, whereas BDW-E underclocked at 3GHz would dissipate about 123W (both computations made by myself). During the demo AMD confirmed that the Zen sample was consuming a bit less power than the BDW-E sample. Well, 117W is a bit less than 123W.

D) AMD refuted to give TDP numbers to the public during the demo, because that 3GHz sample was running above the official 95W rating of the AM4 platform.

E) Their goal is to achieve 3GHz or higher clocks on a 95W rating. This is not possible with current silicon (3GHz <--> ~117W); thus AMD delayed Zen to wait to Glofo to mature the 14LPP node and get higher clocks to make the chip more competitive.

F) My claims are reasonable and backed up by available data (both public and behind doors). I will not insist on this. Anyone is free to believe me or make their own claims.

And your claims are always reasonable, Juan. They aren't "facts" though.

Emphasis mine.

 

[Cazalan]

Like most people interested in technology I love the info leaks as much as the next person, but you also have to know where those leaks are coming from. Often times it is an unscrupulous technician who is perhaps unhappy with his current pay. An engineer worth his salt will not risk losing their job, and perhaps millions of dollars over a lifetime career, by breaking NDAs or other confidentiality agreements.

 

[jdwii]

Like most people interested in technology I love the info leaks as much as the next person, but you also have to know where those leaks are coming from. Often times it is an unscrupulous technician who is perhaps unhappy with his current pay. An engineer worth his salt will not risk losing their job, and perhaps millions of dollars over a lifetime career, by breaking NDAs or other confidentiality agreements.

I typically go with who has been more accurate over time so far things seem to be adding up and conclusions are starting to be made after Zen releases or they at least do a paper launch we will know. The fact that they underclock the Intel sample is concerning why not just use the same clock speed? Why not tell us the TDP? I personally never trust a company when they are to quiet about a launch. I still remember when GCN came out Amd was in a major hurry to show it off(cause it was great). Now i don't see it. I just hope they don't delay Zen until this time next year waiting for Global foundry to create some magical die.

 

[Inkiad]

I think zen will hit local stores somewhere around march next year.

 

[scuzzycard]

Like most people interested in technology I love the info leaks as much as the next person, but you also have to know where those leaks are coming from. Often times it is an unscrupulous technician who is perhaps unhappy with his current pay. An engineer worth his salt will not risk losing their job, and perhaps millions of dollars over a lifetime career, by breaking NDAs or other confidentiality agreements.

I typically go with who has been more accurate over time so far things seem to be adding up and conclusions are starting to be made after Zen releases or they at least do a paper launch we will know. The fact that they underclock the Intel sample is concerning why not just use the same clock speed? Why not tell us the TDP? I personally never trust a company when they are to quiet about a launch. I still remember when GCN came out Amd was in a major hurry to show it off(cause it was great). Now i don't see it. I just hope they don't delay Zen until this time next year waiting for Global foundry to create some magical die.

If they delay it one more time, it will go down in history as vaporware. It already has that feel. As it stands, AMD is 6 months away from releasing something that might, hopefully, possibly compete with something that Intel released 6 months ago, which was itself a pathetic upgrade of Haswell-E which was released in 2014. The best we can hope for is a price war that gets Intel to lower their ridiculous prices for their 8- and 10-core HEDT CPUs.

 

[Cazalan]

I typically go with who has been more accurate over time so far things seem to be adding up and conclusions are starting to be made after Zen releases or they at least do a paper launch we will know. The fact that they underclock the Intel sample is concerning why not just use the same clock speed? Why not tell us the TDP? I personally never trust a company when they are to quiet about a launch. I still remember when GCN came out Amd was in a major hurry to show it off(cause it was great). Now i don't see it. I just hope they don't delay Zen until this time next year waiting for Global foundry to create some magical die.

There is no financial advantage to releasing final clocks a few months before launch. Intel also underclocks their own processors when comparing cores. It removes the differences of turbo modes. The TDPs are on the roadmaps.

Some people were surprised about how much detail AMD gave at HotChips. Others complained it was too little. Can't please everyone.

 

[Melonious]

If they delay it one more time, it will go down in history as vaporware. It already has that feel. As it stands, AMD is 6 months away from releasing something that might, hopefully, possibly compete with something that Intel released 6 months ago, which was itself a pathetic upgrade of Haswell-E which was released in 2014. The best we can hope for is a price war that gets Intel to lower their ridiculous prices for their 8- and 10-core HEDT CPUs.

Intel has only had minor almost jokelike updates to their architecture since core series first came out. It's impossible for them to do anything else.

Much of their dominance is "on paper" as it is already, with their real advantage being production technology.

All AMD must do is cut the cruft on their processors a bit and they can equal or even surpass Intel, and once they do Intel will be unable to respond except with a massive redesign similar to what AMD has done. Which they have never managed to do in the past and would be foolish to expect of them any time soon.

 

[scuzzycard]

If they delay it one more time, it will go down in history as vaporware. It already has that feel. As it stands, AMD is 6 months away from releasing something that might, hopefully, possibly compete with something that Intel released 6 months ago, which was itself a pathetic upgrade of Haswell-E which was released in 2014. The best we can hope for is a price war that gets Intel to lower their ridiculous prices for their 8- and 10-core HEDT CPUs.

Intel has only had minor almost jokelike updates to their architecture since core series first came out. It's impossible for them to do anything else.

Much of their dominance is "on paper" as it is already, with their real advantage being production technology.

All AMD must do is cut the cruft on their processors a bit and they can equal or even surpass Intel, and once they do Intel will be unable to respond except with a massive redesign similar to what AMD has done. Which they have never managed to do in the past and would be foolish to expect of them any time soon.

Well, we've already seen what Zen can do - it's competitive, but it's not going to run faster than Intel's best. Intel doesn't market an 8-core 4.2GHz CPU because they can't - they just choose not to because they have no competition. What I think we will be able to do at this time next year, is to buy an 8-core Intel CPU for $600 or less, when the price has not budged from $1000 for the last 2 years. Intel's high-dollar Extreme Edition will then finally have to be something much faster than a 3GHz snooze-fest.

 

[gamerk316]

If they delay it one more time, it will go down in history as vaporware. It already has that feel. As it stands, AMD is 6 months away from releasing something that might, hopefully, possibly compete with something that Intel released 6 months ago, which was itself a pathetic upgrade of Haswell-E which was released in 2014. The best we can hope for is a price war that gets Intel to lower their ridiculous prices for their 8- and 10-core HEDT CPUs.

Intel has only had minor almost jokelike updates to their architecture since core series first came out. It's impossible for them to do anything else.

Much of their dominance is "on paper" as it is already, with their real advantage being production technology.

All AMD must do is cut the cruft on their processors a bit and they can equal or even surpass Intel, and once they do Intel will be unable to respond except with a massive redesign similar to what AMD has done. Which they have never managed to do in the past and would be foolish to expect of them any time soon.

Well, we've already seen what Zen can do - it's competitive, but it's not going to run faster than Intel's best. Intel doesn't market an 8-core 4.2GHz CPU because they can't - they just choose not to because they have no competition. What I think we will be able to do at this time next year, is to buy an 8-core Intel CPU for $600 or less, when the price has not budged from $1000 for the last 2 years. Intel's high-dollar Extreme Edition will then finally have to be something much faster than a 3GHz snooze-fest.

Intel doesn't release an 8-core CPU for the masses because they'll never turn a profit on it. None of the programs you run will benefit from it, and the yields on such a chip will be low to the point where Intel will have to sell it north of $1000 to turn a profit on it.

For consumer applications, four cores with 2-way SMT is more then you'll ever need; software you run won't ever scale beyond that.

 

[cdrkf]

If they delay it one more time, it will go down in history as vaporware. It already has that feel. As it stands, AMD is 6 months away from releasing something that might, hopefully, possibly compete with something that Intel released 6 months ago, which was itself a pathetic upgrade of Haswell-E which was released in 2014. The best we can hope for is a price war that gets Intel to lower their ridiculous prices for their 8- and 10-core HEDT CPUs.

Intel has only had minor almost jokelike updates to their architecture since core series first came out. It's impossible for them to do anything else.

Much of their dominance is "on paper" as it is already, with their real advantage being production technology.

All AMD must do is cut the cruft on their processors a bit and they can equal or even surpass Intel, and once they do Intel will be unable to respond except with a massive redesign similar to what AMD has done. Which they have never managed to do in the past and would be foolish to expect of them any time soon.

Well, we've already seen what Zen can do - it's competitive, but it's not going to run faster than Intel's best. Intel doesn't market an 8-core 4.2GHz CPU because they can't - they just choose not to because they have no competition. What I think we will be able to do at this time next year, is to buy an 8-core Intel CPU for $600 or less, when the price has not budged from $1000 for the last 2 years. Intel's high-dollar Extreme Edition will then finally have to be something much faster than a 3GHz snooze-fest.

Intel doesn't release an 8-core CPU for the masses because they'll never turn a profit on it. None of the programs you run will benefit from it, and the yields on such a chip will be low to the point where Intel will have to sell it north of $1000 to turn a profit on it.

For consumer applications, four cores with 2-way SMT is more then you'll ever need; software you run won't ever scale beyond that.

This isn't entirely true. What do you define as a 'consumer'? There are lots of people who use their PC for *content creation* (which, lets be honest, is the main reason to have a full on desktop PC these days- there are plenty of light, fast and cheap options for content consumption). It's currently not much use in *gaming* although that is starting to change (plenty of games can fully use 8 threads now, and whilst I don't think we'll need more than that for rendering any time soon, I can see games using more threads when you take into consideration other systems such as complex AI, networking and all the other plethora of stuff that is becoming expected in modern games).

Content *creation* software often can use the cores. Certainly 8 cores / 16 threads isn't out of the question when your talking video editing, graphics work, audio creation, CAD (rendering) and so on. Also I dispute the notion that 'Intel can't make a profit for less than £1000 / sku for an 8 core part". The die size of an octa core cpu without an igpu is no larger than a quad core with igpu... which they sell for less than £200. I mean if we're looking at high end parts here I think it's safe enough to assume the end user will pair it with a discrete gpu so that is a fair trade off.

I personally think if Zen falls where it appears to, it's going to push Intel to offer some higher core count parts lower down. No that won't improve things for a grandparent who likes to keep in touch over the 'video phone thingy', however it most certainly will benefit many in the target audience for a high end PC. Small businesses is another area where this would be a real boon.

If nothing else I can envisage Intel pushing a hex core part down to the price of the current quad core i7....

 

[-Fran-]

I usually understand "content creation" (or bundle it) in the "prosumer" sector.

So, in that regard, gamerk is not off the mark. If AMD can get a 2C/4T version of Zen out soon, they will be able to cater for the lower end of the spectrum, but I don't think that's their initial plan. Since they're targeting the -E lineup, they'll target "prosumer" and professional markets first. Us, regular consumers, be damned, haha.

Cheers!

 

[ComputerSecurityGuy]

I do think that they will push a 4C/8T and possibly 6C/12T chip initially at the $200 and $300 price points respectively.

 

[gamerk316]

I don't see Zen releasing two or six core variants for reasons already stated.

 

[-Fran-]

I don't see Zen releasing two or six core variants for reasons already stated.

There is substance to that, but I also believe AMD is playing it's cards close to their chest.

I put more weight into the later, just because I don't know the *real* innards of Zen's core design.

Hell, they could pull an Intel and just depend on the BIOS/UEFI microcode for the CPU to achieve it.

Cheers!

 

[juanrga]

Like most people interested in technology I love the info leaks as much as the next person, but you also have to know where those leaks are coming from. Often times it is an unscrupulous technician who is perhaps unhappy with his current pay. An engineer worth his salt will not risk losing their job, and perhaps millions of dollars over a lifetime career, by breaking NDAs or other confidentiality agreements.

I typically go with who has been more accurate over time so far things seem to be adding up and conclusions are starting to be made after Zen releases or they at least do a paper launch we will know. The fact that they underclock the Intel sample is concerning why not just use the same clock speed? Why not tell us the TDP? I personally never trust a company when they are to quiet about a launch. I still remember when GCN came out Amd was in a major hurry to show it off(cause it was great). Now i don't see it. I just hope they don't delay Zen until this time next year waiting for Global foundry to create some magical die.

I smell that "magical die" is not coming... AMD just announced a re-negotiation of the WSA

http://ir.amd.com/phoenix.zhtml?c=74093&p=irol-newsArticle&ID=2198716

 

[juanrga]

There is no financial advantage to releasing final clocks a few months before launch. Intel also underclocks their own processors when comparing cores. It removes the differences of turbo modes. The TDPs are on the roadmaps.

One thing is to give final exact clocks which aren't known becuase depends on the foundry ability, another very different is to not mention clocks during two years. There is no "financial advantage" on hiding clocks. We have known for a while the clocks for designs made by Intel, APM, Cavium, Broadcomm, ARM,... IBM just presented its new Power9 and they also talked about the clocks they expect to hit.

 

[juanrga]

Well, we've already seen what Zen can do - it's competitive, but it's not going to run faster than Intel's best. Intel doesn't market an 8-core 4.2GHz CPU because they can't - they just choose not to because they have no competition. What I think we will be able to do at this time next year, is to buy an 8-core Intel CPU for $600 or less, when the price has not budged from $1000 for the last 2 years. Intel's high-dollar Extreme Edition will then finally have to be something much faster than a 3GHz snooze-fest.

Besides the economic side of things mentioned by gamerk, we also have the technical side. The 8C BDW runs @3.2GHz on 140W slot. It would be increased to about 250W if clocked at 4.2GHz.

Similar thoughts about AMD. They are releasing a 95W 8C chip (instead a 250W 8C chip) due to technical and economical reasons, not "because they have no competition".