So much for "8 cores is more future proof!"

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

rainy

Senior member
Jul 17, 2013
508
427
136
Because AMD was a painful experience for more than a decade. People have not forgotten that experience.
Over decade it's a serious exaggeration: Athlon 64 was very OK till July 2006 (Core 2), then in 2009-2011 various models of Phenom II were released and it was a decent architecture and finally in March 2017 Zen architecture was released by AMD.
 
Last edited:

H433x0n

Golden Member
Mar 15, 2023
1,066
1,247
96
This age old debate is irrelevant in 2023. The only new info is that when looking back at ComputerBase’s Zen 1 review it turns out that they’re borderline delusional and their commentary isn’t on the level at all.
 

Ranulf

Platinum Member
Jul 18, 2001
2,407
1,305
136
Why are we exhuming a horse that has been dead for ~10 years just to beat it some more?

Because Bryan at TechYesCity needed content? I don't know, I read the video title/description and some comments but can't be bothered to watch. Kinda burned out on this issue. At this point I can't really complain about either my fx8350 which was an upgrade for a phenom 955 quad core in 2014 that lasted me until I retired it in 2021 and my now 8 year old 4790k.

I have more regrets getting into AM4 early than anything AM3 related and regret not getting a 2600k over the 2500k in 2011.
 

Abwx

Lifer
Apr 2, 2011
11,166
3,862
136
RA Tech has a video about it. Watch and listen to the benchmarks running. In Shadow of the Tomb Raider, Ivy Bridge i5 had audio issues, and failed to render assets and NPCs. It will spit out good numbers at the end though. RA was playing games and using discord on windows 10 (you know how many gamers actually use their gaming PCs) and experiencing audio issues on those old similarly priced Intel CPUs, the FX was fine. That said, there were and still are games where the better single thread IPC of those i3 and i5 Intel CPUs makes a positive difference to playability vs FX, which IMO is all that matters, the gaming experience. Not bigger bar better on charts. Most test suites were automated and or unattended. And certainly not jumping in 64p MP and playing a few rounds then reporting back. Instead testing early in games which is usually less CPU taxing. In game test runs, when not using canned, were and often still are, under a minute each, often under 30 seconds.

That was due to i5 of any gen including Haswell not being multitasking capable, once you threw more than one app the perfs of the respective apps would tank contrary to the 2600K FI, but the FX8350 was much better than both in this respect, you could do a rendering and still game without much losses in the FPS.

This didnt appear in benches that were made in clean rooms with only the game running and not the slightest background task performed, this is not visible in the 1800X review i linked because Computerbase didnt do their usual multitasking tests but it can be seen in previous reviews like this one :

 

BFG10K

Lifer
Aug 14, 2000
22,709
2,979
126
So feel free to make all the sarcastic threads you want, but you are cheering for a time when we were held back by capitalism rather than technology. Disgusting.
Nobody's cheering for Bulldozer/Piledriver except those that still refuse to believe(tm). The thing was absolute garbage, AMD's version of Prescott.

Now the pendulum's swung the other way and Intel's the one shipping furnaces specifically built for Cinebench. Suddenly gaming and power consumption doesn't matter, it's all about MOAR CORES!!! And you guessed it, certain people from the other camp are now defending this.

These things are revisited because some things never change. Indeed, in the video forum right now we have someone parading around claiming 8GB VRAM isn't an issue ("you're using the wrong settings!") despite piles of objective evidence already here to the contrary. Go back far enough and we had the same thing with 4GB vs 8GB, 2GB vs 4GB, etc. Look how those turned out.

And don't forget this forum also told us "Mantle was going to allow AMD to dominate the Android market", and "DX12 would provide automatic performance gains without any effort just by flipping a simple switch".

Yep, I also started threads "revisiting" those too, heh.
 
Last edited:

ondma

Platinum Member
Mar 18, 2018
2,768
1,350
136
Nobody's cheering for Bulldozer/Piledriver except those that still refuse to believe(tm). The thing was absolute garbage, AMD's version of Prescott.

Now the pendulum's swung the other way and Intel's the one shipping furnaces specifically built for Cinebench. Suddenly gaming and power consumption doesn't matter, it's all about MOAR CORES!!! And you guessed it, certain people from the other camp are now defending this.

These things are revisited because some things never change. Indeed, in the video forum right now we have someone parading around claiming 8GB VRAM isn't an issue ("you're using the wrong settings!") despite piles of objective evidence already here to the contrary. Go back far enough and we had the same thing with 4GB vs 8GB, 2GB vs 4GB, etc. Look how those turned out.

And don't forget this forum also told us "Mantle was going to allow AMD to dominate the Android market", and "DX12 would provide automatic performance gains without any effort just by flipping a simple switch".

Yep, I also started threads "revisiting" those too, heh.
But Intel *is* competitive in gaming. In fact they have a slight edge except for the purpose built vCache chips. Although Intel power consumption is definitely a drawback, it is far from the Bulldozer vs Intel situation.
 

TESKATLIPOKA

Platinum Member
May 1, 2020
2,423
2,914
136
Why are we exhuming a horse that has been dead for ~10 years just to beat it some more?
Because @BFG10K had some personal vendetta against Bulldozer, so he had to make a thread twelve years later, where he could finally vent his pent up frustration.
But this is seriously pointless thread at this time.
 
Last edited:

TESKATLIPOKA

Platinum Member
May 1, 2020
2,423
2,914
136
Because AMD was a painful experience for more than a decade. People have not forgotten that experience.
And you were one of those? If not, then I would rather hear that from someone who actually owned a Bulldozer system.
But considering you are talking about some nonsensical "more than a decade of painful experience" BS, It's pretty clear you didn't own It.
 

Zor Prime

Golden Member
Nov 7, 1999
1,023
588
136
Because AMD was a painful experience for more than a decade. People have not forgotten that experience.
I went from Phenom II X6 to Ryzen 1700. Don't see a problem. The reason why I upgraded was because some games required SSE(3-4?) instructions the Phenom II X6 simply didn't have. I was still rocking most of the latest games on that Phenom II X6. Damn, what a fine chip that was.
 

Hans Gruber

Platinum Member
Dec 23, 2006
2,217
1,153
136
I went from Phenom II X6 to Ryzen 1700. Don't see a problem. The reason why I upgraded was because some games required SSE(3-4?) instructions the Phenom II X6 simply didn't have. I was still rocking most of the latest games on that Phenom II X6. Damn, what a fine chip that was.
There was a chart a few years ago that showed the jump from Bulldozer to the Zen architecture was a 60% improvement in IPC performance. That illustrates just how bad the AMD architecture was before Zen. The Ryzen CPU's really hit their stride when they switched from GloFlo to TSMC silicon with Zen 2.
 
Reactions: lightmanek

gdansk

Platinum Member
Feb 8, 2011
2,489
3,379
136
Bulldozer to Zen was 60% more IPC.
But Excavator+ to Zen was more like 30-35% more IPC.
Really, one of the problems AMD had was their inability to update their desktop platform with improved cores every year like Intel did. Most of their limited efforts were rightly focused on Zen.

Edit: Oh actually I'm wrong according to AMD Zen to Excavator is still about 1.5x .
 
Last edited:

Abwx

Lifer
Apr 2, 2011
11,166
3,862
136
There was a chart a few years ago that showed the jump from Bulldozer to the Zen architecture was a 60% improvement in IPC performance. That illustrates just how bad the AMD architecture was before Zen. The Ryzen CPU's really hit their stride when they switched from GloFlo to TSMC silicon with Zen 2.

That was on Cinebench wich is FP, what about Integer wich is more in line with standard usage, have you a number..?..
Bulldozer to Zen was 60% more IPC.
But Excavator+ to Zen was more like 30-35% more IPC.
Really, one of the problems AMD had was their inability to update their desktop platform with improved cores every year like Intel did. Most of their limited efforts were rightly focused on Zen.

The improvement vs Excavator was 52%, but this CPU had a cache that was halved in respect of the previous core, and as said it was in Cinebench, i dont think that it was as massive in Integer.

Edit : In 7 ZIP ST a FX8350 score 3600MB/s at 4.2, comparatively Zen 4 score 6000MB/s@5Ghz, so about 5040MB/s at same 4.2 frequency and 40% better IPC in Integer for Zen 4, it could be somewhat different in other Integer codes.
 
Last edited:

NostaSeronx

Diamond Member
Sep 18, 2011
3,688
1,222
136
Bulldozer's lineal development was revolving door-style. Where as Zen development has had the same person running the ship.

1st Gen: 1997-2002 = Witt/Keller -- Low Power
2nd Gen: 2003-2004 = Glew(Architect)+Weber(Director) -- High Performance
3rd Gen: 2005-2007 = Moore(Architect)+Hester(Director) -- Low Power/Server-scale
4th Gen: 2006-2011 = Butler(Architect)+Meyer(Director(before launch) -> Co-Director(@ launch)) -- High Performance/Mobile/Server-scale
5th Gen: 2012-2014 = Clark(Architect)+Fair(Director) -- Mobile, Server-scale canned with 20LPM/SHP

If we see this then the performance improvements are going to get eaten by new direction(director as well) forced onto each new Chief everytime.

Also, there has been a development recently.
"Revisiting Clustered Microarchitecture for Future Superscalar Cores: A Case for Wide Issue Clusters"
"Huawei Technologies Company Ltd., Shenzhen, China" -> "He is currently the CPU Chief Architect with Huawei"

If the Huawei custom core comes out and it's a clustered microarchitecture. I warned you all many times over, btw. All monolithic cores will lose to the CMT Kunpeng.
 
Last edited:

gruffi

Member
Nov 28, 2014
35
117
106
Nobody's cheering for Bulldozer/Piledriver except those that still refuse to believe(tm). The thing was absolute garbage, AMD's version of Prescott.
No, definitely not. Netburst was a complete failure. The concept, the implementation, just everything. It was destined to fail. The biggest garbage in the x86 history. Bulldozer is a different story. The concept was actually really good. But it failed because of mainly two factors. First, the design goals were way too conservative. An AMD engineer back then talked about it. The goal was like +20% clock speed at similar IPC (or -5% maximum) compared to 10h family. But they should have gone for something like at least +10-20% clock speed and >=20% IPC. And second, the initial implementation was pretty miserable, also caused by wrong decisions. They achieved the clock speed improvement of 20%. Deneb started with 3 GHz, Zambezi started with 3.6 GHz base clock. And the MT scaling was quite good (which wasn't that impressive at all due to the low ST IPC). But that's it. The rest was more than underwhelming. They not only lost 5% IPC maximum, but more like 15% compared to Phenom II. Some decisions also were simply wrong. They should have implemented a WB cache, not WT. They should have implemented horizontal MT only, not this awkward mix of horizontal and vertical MT. They should have focused more on legacy code optimization, not on new extensions like SSE5 or XOP. Etc. Intel also had the process node advantage. Even their 32nm process was quite a bit better than Glofo's 32nm process. Let alone Intel's 22nm process, used with Ivy Bridge and Haswell. That also has a significant impact comparing power consumption. The only decent Bulldozer design was Excavator. But too little too late. Zen was just around the corner at that time.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
28,805
21,542
146
It was a settlement, so AMD paid the court system to do the opposite. You in fact can't deny that it doesn't have eight cores. (There is only ever eight-cores for FX-8, six for FX-6, four for FX-4, etc)

Note the website is eight cores and eight threads. You can't actually lawsuit them for this claim because they won the settlement by buying out any further litigation/trials. Think of it AMD strategically/tactically lost the battle to win the war. Sure they pay people a few dollars but this issue can never legally come up again.
Setting aside the technical discussion; AMD took a major L from that lawsuit. They were already hemorrhaging cash and in debt. Another 12 million+ and legal fees was like death by a thousand cuts. Getting to declare FX as 8 cores from then on is most certainly not a W either. In fact, it makes it seem somehow even worse as in the TYC comparison above. Had it not been for forming the semi-custom division to work with MS and Sony I don't think AMD exists anymore. The name might have survived, but the company would be vastly different, perhaps entirely irrelevant.
 

Schmide

Diamond Member
Mar 7, 2002
5,588
719
126
The big difference here is not the cores but the instruction set. Haswell has AVX2 which is the basis for most of the modern libraries DX11-12. Without those extensions modern software is going to run on legacy SSE. AVX was replaced to fast to ever get full adoption.

It would still lose still but sandy/ivy bridge would be proper test.

Tech YES acts like the z87 motherboard and haswell CPU is lesser because it's an OEM and starts making excuses for it. (this is just your bottom of the barrel)

Seriously you can find 3 Bulldozer cpus but can't find a sandy bridge?

Shenanigans
 
Last edited:

gruffi

Member
Nov 28, 2014
35
117
106
But Intel *is* competitive in gaming. In fact they have a slight edge except for the purpose built vCache chips. Although Intel power consumption is definitely a drawback, it is far from the Bulldozer vs Intel situation.
Except? You are funny. Take away Intel's cache advantage over non V-Cache Ryzens and compare at the same memory specs and AMD still at least trades blow with Intel's offerings on average. In fact we are not far from the Bulldozer situation. It's almost the same, just vice versa. If you want the best gaming experience then there is no other choice than AMD's V-Cache models at the moment. They are faster and use only half or even less energy to achieve this. You must be a die hard Intel fan to believe they are competitive in gaming. They are definitely NOT at the higher end of the scale. Just like Bulldozer was.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
28,805
21,542
146
Shenanigans
Precisely. His channel hasn't grown much in the last 5yrs. Meanwhile Dawid does tech is strictly for the LULZ and has already passed him by. Bryan's been click farming hard in an attempt to grow the channel. E.G. he and his buddy Greg Salazar claiming Ryzen 3600s are dying like flies. Or his attempt to make latency some kind of hot button issue in the community.

We shot the messenger which is usually poor debating, but sometimes they have it coming.
 
Reactions: Tlh97 and Schmide

TESKATLIPOKA

Platinum Member
May 1, 2020
2,423
2,914
136
Except? You are funny. Take away Intel's cache advantage over non V-Cache Ryzens and compare at the same memory specs and AMD still at least trades blow with Intel's offerings on average. In fact we are not far from the Bulldozer situation. It's almost the same, just vice versa. If you want the best gaming experience then there is no other choice than AMD's V-Cache models at the moment. They are faster and use only half or even less energy to achieve this. You must be a die hard Intel fan to believe they are competitive in gaming. They are definitely NOT at the higher end of the scale. Just like Bulldozer was.
I wonder who is biased here.
Raptor is pretty competitive in performance even against V-cache models, actually only 7800X3D is faster. The only thing you can criticize is that horrible power consumption, but that's It.

Here is a chart to Min FPS at different power limits and power consumption for anyone who is interested. Link
 

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
No, definitely not. Netburst was a complete failure. The concept, the implementation, just everything. It was destined to fail.

That is uninformed nonsense. It had great ideas that evolved x86 cores and also not so great ideas as well. I think corporate policies at Intel were equally important, as they were throwing everyone with pulse and two hands on Itanium project and P4 was not given mindshare and transistors it required. Then gigahertz stupidity happened and the rest is history.

What does not align with that now retrospective test of history is Your bold "claims". Don't take my word for it, i'd strongly suggest You read Chips and Cheese piece on P4:


TLDR version:


So in fact plenty of ideas lived on when executed properly in Sandy Bridge and continue to live on to this day.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
28,805
21,542
146
I wonder who is biased here.
Raptor is pretty competitive in performance even against V-cache models, actually only 7800X3D is faster. The only thing you can criticize is that horrible power consumption, but that's It.

Here is a chart to Min FPS at different power limits and power consumption for anyone who is interested. Link
It's more complicated than throwing up bar charts of what W1z tested. It's too easy to change a game suite and watch those bigger bar better change. For example: Ages of Empire IV is single thread heavy which helps Intel hide power issues and why the 14900K doesn't lose performance at 95W. It's such an outlier that including it is sus. Spiderman and CSGO also prefer Intel. Change those 3 out for Horizon Zero Dawn, Assetto Corsa, and MS Flight Sim and it'd tell a very different story. Conversely you could add even more titles that favor Intel and do the same. Even aggregate scores suck because not everyone test stock everything, equal ram speeds, boost clocks on GPUs, most have massive overlap on titles tested, etc.

It has been about bang for buck and what games you play for a long time now. Nothing has changed there. What has changed is the industry consensus of which company is having issues with power efficiency. No amount of image management and reform efforts are going to get most people to believe power and heat are not issues for Intel on desktop. The other factor is platform longevity. Which matters, regardless of the very vocal small minority that have created talking points claiming it isn't. If any of their talking points held water, AM4 would not still be killing it in retail sales all these years later. It's that simple, there is no bamboozling anyone with the slightest clue while that fact exists.

I am enjoying this thread though. Dunking on FX is shooting fish in a barrel. Something no one sporting would ever do.
 

TESKATLIPOKA

Platinum Member
May 1, 2020
2,423
2,914
136
It's more complicated than throwing up bar charts of what W1z tested. It's too easy to change a game suite and watch those bigger bar better change. For example: Ages of Empire IV is single thread heavy which helps Intel hide power issues and why the 14900K doesn't lose performance at 95W. It's such an outlier that including it is sus. Spiderman and CSGO also prefer Intel. Change those 3 out for Horizon Zero Dawn, Assetto Corsa, and MS Flight Sim and it'd tell a very different story. Conversely you could add even more titles that favor Intel and do the same. Even aggregate scores suck because not everyone test stock everything, equal ram speeds, boost clocks on GPUs, most have massive overlap on titles tested, etc.
I think what I posted more than enough shows that Raptor Lake is not another Bulldozer. Performance is there, It's just not power efficient.
Regrettable that there is no similar test for 14600K or AMD chips with different power limits.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
28,805
21,542
146
I think what I posted more than enough shows that Raptor Lake is not another Bulldozer.
I wasn't contending that point as I agree. I hope we never have that big of a difference between vendors again. It's the best CPU market in forever because AMD got their groove back. They need to sustain another 5yrs to really get us where we need to be though. Wintel is still most of the PC pie. I'd like to see AMD and Linux get a significantly bigger slice.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |