Are you an AMD or Intel guy?

Who has your brand loyalt?

  • Intel

    Votes: 7 16.7%
  • AMD

    Votes: 15 35.7%
  • Cyrix

    Votes: 0 0.0%
  • No brand loyalty here, whoever is faster

    Votes: 23 54.8%
  • ARM

    Votes: 1 2.4%
  • PowerPC

    Votes: 0 0.0%
  • Other

    Votes: 2 4.8%
  • Pancakes

    Votes: 9 21.4%

  • Total voters
    42
Not all of them were bad. But I went through several motherboards before I found a good one for my Athlon Thunderbird 1.33GHz CPU. It was an A7V-133 actually. Even the ABIT boards I used prior to that were a mess. That setup was bullet proof but after all it took to get there, I took a break from AMD CPU's for some time until the Athlon 64 came out. That 3200+ system I had was one I had for a spare machine at the time, not my main rig. It wasn't as stable as I would have liked.

For the life of me, I can't remember what board I had for my A64 3000+. I remember I had a DD loop and a BFG 6800GT or something like that. Sold the whole system on Ebay because I was a broke teenager and I needed paintballs.
 
For the life of me, I can't remember what board I had for my A64 3000+. I remember I had a DD loop and a BFG 6800GT or something like that. Sold the whole system on Ebay because I was a broke teenager and I needed paintballs.

I'm not sure what my Athlon XP 3200+ had. I do remember the boards I had for my Athlon 64's. I had the ASUS A8N SLI Deluxe for my 3800+, Tyan S2895 (K8WE) for my Opteron 254's. I had a Tyan S2882 for Opteron 246's. I had a 3800+ (Venice) on an A8N-E. I remember most of the motherboards I have, but it gets fuzzy around the Pentium IV days. I changed them out quite frequently at the time excluding my P4C800 Deluxe. I never updated to the P4C800-E Deluxe as the change was so minor, it wasn't worth buying the new model. I also had a number of them in the Pentium II, III, and Athlon days. In my 486 days, and even somewhat up into my Cyrix 6x86 days, I didn't track that information too much so I couldn't tell you too much about the boards I had back then. I just looked at the chipset and cache configuration above all else.
 
I'm not sure what my Athlon XP 3200+ had. I do remember the boards I had for my Athlon 64's. I had the ASUS A8N SLI Deluxe for my 3800+, Tyan S2895 (K8WE) for my Opteron 254's. I had a Tyan S2882 for Opteron 246's. I had a 3800+ (Venice) on an A8N-E. I remember most of the motherboards I have, but it gets fuzzy around the Pentium IV days. I changed them out quite frequently at the time excluding my P4C800 Deluxe. I never updated to the P4C800-E Deluxe as the change was so minor, it wasn't worth buying the new model. I also had a number of them in the Pentium II, III, and Athlon days. In my 486 days, and even somewhat up into my Cyrix 6x86 days, I didn't track that information too much so I couldn't tell you too much about the boards I had back then. I just looked at the chipset and cache configuration above all else.

This reminds how I can't wait till you do another best/worst motherboards story again. Those have always been fun reads especially when they covered both sides of the fence going back 10+ years.
 
Most people looking at my rigs would think I'm simply teams blue/green but that not's true at all.

I've had both Intel and AMD over the years and I'm really looking forward to doing another AMD build in the next few years. If I had the money to burn I'd do one now. It'll be hilarious if I end up with a AMD CPU and Intel GPU.
 
Both and neither.

I pick the one for the job. AMD is usually cheaper if it's something that does not require speed.
Though AMD has caught up quite a bit and is very close, I do want to build and AMD system. It's hard since I already
have a 5960x OC, so not sure how much better it would be.
 
I'd like to say I have no loyalty as there are a few retired AMDs around the house, but AVX performance is critical for me, and while I'm glad that AMD finally caught up (in theory, I'll wait for some actual performance comparisons) with Intel's AVX2 implementation, there's no AVX512, yet, which is even faster. And for that I'll probably be buying Cascade Lake X in the fall, and I seriously hope that Intel's pricing falls into better line with Zen 2 to sweeten the deal, now that they seem to be close to standard performance parity.
 
Both and neither.

I pick the one for the job. AMD is usually cheaper if it's something that does not require speed.
Though AMD has caught up quite a bit and is very close, I do want to build and AMD system. It's hard since I already
have a 5960x OC, so not sure how much better it would be.

I went from an Intel Core i7 5960X @4.5GHz to an AMD Threadripper 2920X @4.2GHz. It really seems like a lateral move most of the time. I really like the X399 platform, so the upgrade has that going for it at least.
 
Ack.. where do I begin?

My list (similar to Dan's):
1985-1986 Zilog Z80A - 4mhz (Amstrad CPC 464)
1986-1990 NEC V20 (8088) 10mhz - was originally an intel 8088
1990-1993 AMD am386dx-40 (first self built)
1993-1994 Intel 486sx-25 (overclocked to 33mhz)
1995 (briefly) AMD K5-75
1995-1997 Cyrix 6x86 PR-166
1997-1998 AMD K6 200 (overclocked 225)
1998-1999 AMD K6-2 450 (overclocked 480)
1999-2001 AMD Athlon 600 (overclocked 635)
2001-2004 Dual Athlon 1200 MP, which I upgraded to dual 2600+ after crushing a core
2004-2007 Athlon 64/754 G0 Clawhammer - 3ghz (overclocked to 3.2)
2007-2008 Intel Q6600 (overclocked to 3.0ghz)
2008-2012 Intel C2D 2.66 ("Penryn") - Macbook pro (went with this for a while, still have it in the house)
2012-2014 Intel i7-3770 (non k)
2013 Intel i7-4570s (home server)
2014 Intel i7-4770 (non k) (home server)
2015-2017 Intel i7-5775C
2017-2019 AMD Ryzen 7 1700 (non x, overclocked to 3.7ghz on all cores) - Motherboard died.

With a bit of thinking I reckon I could name nearly all of the motherboards, in recent history I've had the most luck with Asrock, older history (from about 1999-2008) Gigabyte motherboards served me best. Asus has always died on me.

Currently:
Intel i7-9700k (desktop, overclocked 4.8ghz/0 AVX offset - with all the sleep states and power saving on, cause it's idling a lot) +
Intel i7-4790 (server - upgraded only as I got a really good deal)
Intel Pentium N3700 (notebook).

The 9700k requires obscene amounts of airflow to keep cool (I have it under a D15), even at stock settings. Fan filters are required, somewhat dusty environment.

If you count sheer computing time spent, I'd probably have more time on intel chips than AMD, I have a preference for the underdog but need absolute platform stability at the moment and that's why I'm on intel

I've been a gamer since 1985.. so I guess that's 34 years of gaming. These days I get an hour here or there.. had a lot more time before I had a kid, but I'm having a great time revisiting 80s classic games with him (Space quest 1 EGA currently).
 
Last edited:
Here are some CPU's I have in my collection. Most are obviously Intel. But this gives you some idea of how many processors I've had as I discovered a couple I don't remember getting.

KClbhOih.jpg


From top to bottom, left to right.

Intel 80186
IBM Blue Lightning DX2-66MHz (Cyrix)
Intel 486 DX2-66MHz
Intel Pentium 75
Cyrix 486 DX2 80MHz
Intel Pentium 60MHz
Intel Pentium 90MHz
Intel Pentium Pro 150MHz (I have some 180MHz chips as well, but they are on a motherboard)
Pentium MMX 166MHz (Left)
Pentium MMX 233MHz (Right)
Pentium II 233MHz
Pentium III 500MHz
Pentium III (Coppermine) 1.0GHz
Pentium IV 1.5GHz
Pentium 4 965 3.73GHz Extreme Edition
Core 2 Duo E6400 (I didn't know I had one of these)
Core 2 Quad Q6600
Core i7 5960X
Core i7 7740X (Most retarded CPU ever)
AMD Athlon (Thunderbird) 1.33GHz
AMD Athlon XP 2600+ (I have no idea where this CPU came from)

Occasionally, when friends and families give me their old systems, I pull the processors out for the collection. That's where some of these came from. A few of them are the ones I used in my own personal machines back in the day, but very few of them are examples of this. The Q6600, 5960X and Athlon 1.33GHz are probably the only CPU's here that were pulled from previous "main" PC's. Although, some of the earlier 486 era stuff could be as well. I was late to the party and was using that stuff well into the Pentium era. By the time I got a Pentium, they were still current but the 486's weren't worth anything. I can't recall if those 486 DX2's were actually the ones in my rig or pulls from other systems years later.
 
Last edited:
Alright if we're listing past CPU's from out past rigs.

Atari 6502-used Atari 400 but then significantly modded. Funny story, I bought this from a friend of my dad with newspaper deliver money.
Intel 8088 7.16Mhz store bought Tandy 1000ex
Pentium II 233Mhz-had me down from Dad and modded w/ anything I could get my hands on
Unknown AMD-a hand me down from Dad and I don't remember the specs didn't last too long since I got the p4 shortly after.
Pentium IV 3.06Ghz store bought VPR Matrix
Pentium IV 3.40Ghz-self built completely rebuilt and upgraded the Matrix
Core 2 Quad Q9300 2.50Ghz-store bought Gateway FX7026 and upgraded ram, GPU, HDD's over the years
Core I7 620M-Sony VAIO laptop
Core I7 2630QM-Toshiba Qosmio x775 3d laptop upgraded ram and drives
Core I7 2600k-self built
Core I7 4930k-self built
Core I7 5670Q-MSI Titan GT80 w/ many upgrades
 
I'm a fan of AMD products, and I'm not ashamed of it. While they have not necessarily had the best performance, historically, the performance/price ratio has always tilted in AMD's favor, and that is more important to me than having the absolute fastest. I personally purchased my first new Intel processor during the Sandy Bridge era, as the performance was so heavily skewed in Intel's favor that the value proposition actually tilted their way in my mind. I have always appreciated that AMD did not lock down features of products just for market differentiation, but simply allowed performance and price to make that difference.
 
I have always appreciated that AMD did not lock down features of products just for market differentiation, but simply allowed performance and price to make that difference.

I'm sorry, but this is simply not true. AMD locked down its triple core processors by disabling the fourth core. More often than not, the fourth core was absolutely fine. Intel doing things like this is far more rampant, granted. Most of the time I think Intel is fine doing so as building fewer products and segmenting out one thing to many markets makes sense. Using processors that would otherwise be scrap and laser cutting traces to remove faulty cache or whatever is fine. However, when it comes to its X299 platform and vROC, I'm right there with you.
 
I'm sorry, but this is simply not true. AMD locked down its triple core processors by disabling the fourth core. More often than not, the fourth core was absolutely fine. Intel doing things like this is far more rampant, granted. Most of the time I think Intel is fine doing so as building fewer products and segmenting out one thing to many markets makes sense. Using processors that would otherwise be scrap and laser cutting traces to remove faulty cache or whatever is fine. However, when it comes to its X299 platform and vROC, I'm right there with you.
I was always under the impression that the 4th core was more of a binning issue, kinda like overclocking - maybe you will, maybe you won't get that 4th core. I seem to recall varied reports from people who purchased the tri-core processors (I actually have one, but did not purchase a motherboard known to 'unlock' them so I couldn't test myself). I'm thinking more specifically of things like ECC support and virtualization support, which often there didn't seem to be any reason or rhyme to Intel's level of support - prime example being on the "C" chipsets, you could have ECC support with a Xeon or an i3, but not an i5 or i7? For full virtualization support, some i5's would have it, some wouldn't; some i7's would have it, some wouldn't. I realize this is less and less the case with more recent products, but still have to be careful.
 
I was always under the impression that the 4th core was more of a binning issue, kinda like overclocking - maybe you will, maybe you won't get that 4th core. I seem to recall varied reports from people who purchased the tri-core processors (I actually have one, but did not purchase a motherboard known to 'unlock' them so I couldn't test myself). I'm thinking more specifically of things like ECC support and virtualization support, which often there didn't seem to be any reason or rhyme to Intel's level of support - prime example being on the "C" chipsets, you could have ECC support with a Xeon or an i3, but not an i5 or i7? For full virtualization support, some i5's would have it, some wouldn't; some i7's would have it, some wouldn't. I realize this is less and less the case with more recent products, but still have to be careful.

I did a test at HardOCP with 10 of those processors and 8 of them unlocked without issue. Binning may have had allot to do with it but the unlock rates were high, implying that they weren't all necessarily disabled because of binning. Also keep in mind that Threadripper is basically a market segmented Epyc CPU. Fewer CCX complexes and half the memory channels. AMD separates its products somewhat artificially as well. Intel does this far more often than AMD does, but both companies are guilty of it. I think it's a smart business decision.

Having said that, I do agree that Intel's segmentation is a bit rigid when it comes to things like virtualization support and ECC support on i7 and i9 CPUs.
 
I caved... the 9700k is going and the 9900k is coming

I place the blame firmly on Dan_D's shoulders for the cpu pics.
 
Last edited:
"No brand loyalty here, whoever is faster"

Now for the longest time that was Intel, so naturally I went there. But, I had my Athlon days and CPUs of that era that competed well. But then of course after a certain time Intel just took off, and AMD had really nothing. But now it's swinging back around after so long.

My next upgrade this July is a Ryzen 3000 CPU/X570 system for my main computer, upgrading from a 4770K that I recently upgraded to 7700K. Now I'll be upgrading to a Ryzen 7 3800X or Ryzen 9 3900X, I want the 3900X.
 
"No brand loyalty here, whoever is faster"

Now for the longest time that was Intel, so naturally I went there. But, I had my Athlon days and CPUs of that era that competed well. But then of course after a certain time Intel just took off, and AMD had really nothing. But now it's swinging back around after so long.

My next upgrade this July is a Ryzen 3000 CPU/X570 system for my main computer, upgrading from a 4770K that I recently upgraded to 7700K. Now I'll be upgrading to a Ryzen 7 3800X or Ryzen 9 3900X, I want the 3900X.

For all but about a five year period in the last two decades or so, AMD was that "other" CPU company. A company that produced CPU's which were at best, budget alternatives and at worst, cheap knock offs of Intel's CPU's. It has at times even been in third place as far as prolific PC CPU vendors go. People often fail to realize that AMD has never truly beat Intel on equal footing. It seems AMD has to both produce a good product and Intel has to falter in some way to create conditions favorable enough for AMD to win when it comes to performance. When you compare the size of each company and their capital, R&D budgets and manufacturing, it's always a David vs. Goliath type story.

It's certainly never competed financially. AMD only gained the ability to design a better performing processor than some of Intel's offerings due to circumstance and making the right business decisions. Its acquisition of NexGen Systems and the hiring of former DEC Alpha engineers allowed AMD to create the wildly successful K6, K7 and K8 processors. At the same time, Intel happened to bet on the wrong horse with Netburst and they ended up being stuck with the high clocking, inefficient design for several years allowing AMD to win the performance crown. Intel had bet on its superior manufacturing and deep pipeline design to allow for clock speeds upwards of 5GHz and beyond to counter AMD's more efficient design. This obviously never happened as even reaching 4GHz proved to be beyond Intel's capabilities at the time.

I've often said, it wasn't a case of the Athlon being that good, but more about Intel's Netburst being that bad. Its obviously a combination of both, and the Athlon series (especially, the Athlon 64) were certainly innovative, but Intel made several blunders which allowed AMD to come out looking much better than it otherwise would have. Intel choosing Rambus' RDRAM was one blunder, Netburst itself was another. You had chipset fiasco's with i820 and its MTH issue during this time. NVIDIA's nForce chipsets helped give AMD what it was sorely lacking at the time which was a quality platform. Something that's hard to believe given how bad NVIDIA's Intel chipsets were.

Today, security mitigation not withstanding, AMD is still at an IPC and clock speed deficit compared to Intel. AMD is in a unique position once again, but its less about how good its products are and more about how Intel has faltered and made bad decisions once again. Intel's design allowed for outstanding performance while leaving their CPU's vulnerable. Intel drove the market for so long that it became complacent and overly greedy, even by its own standards.

It's general greed and complacency created a climate which allowed AMD to come in with a core count advantage and good enough IPC to beat Intel in some applications and provide good enough performance for gamers and general enthusiasts. AMD made all the right decisions regarding platform features and pricing. X399 motherboards are just as expensive as their X299 counterparts, but they provide more features for less money as you don't need special license keys or processors topping $1,000 to leverage the full capabilities of the platform. You also don't need to step up to the Xeon to gain access to ECC memory support. Of course, Intel's security issues and discovered flaws have allowed AMD to close the performance gap and move up the food chain. Not only does the performance deficit get reduced, but Intel's reputation has taken massive hits lately.

Now, things will certainly get more interesting once we get Ryzen 3000 series CPU's in hand and do some testing. We'll see how its IPC advantage over Ryzen 1000 and 2000 series CPU's stacks up and whether or not it can fully close the gap (or surpass) Intel's newer CPU's.
 
Hmm I don’t agree with parts of your post, which is unusual Dan.

Core2 was based on pIII. It was not technically superior to AMD’s option, it just happened that there were enough tweaks possible to make it faster.

More recently with the piledriver/bulldozer mess, I think it was a case of AMD making the wrong trade offs.

I think the recent intel blunder with 10nm is a big one. Relying on a somewhat outdated architecture is hurting intel. But to their credit they are pushing it further and are staying competitive (unlike the p4 era).

I don’t actually think ice lake is going to be a huge jump from skylake, just an iteration on the theme.

Intel really need a new architecture, ideally with something like tensor cores.
 
Hmm I don’t agree with parts of your post, which is unusual Dan.

Core2 was based on pIII. It was not technically superior to AMD’s option, it just happened that there were enough tweaks possible to make it faster.

More recently with the piledriver/bulldozer mess, I think it was a case of AMD making the wrong trade offs.

I think the recent intel blunder with 10nm is a big one. Relying on a somewhat outdated architecture is hurting intel. But to their credit they are pushing it further and are staying competitive (unlike the p4 era).

I don’t actually think ice lake is going to be a huge jump from skylake, just an iteration on the theme.

Intel really need a new architecture, ideally with something like tensor cores.

Core 2 was heavily reworked over P6 and it had a massive IPC advantage over AMD's Athlon X2 and Phenom / Phenom II processors. Even without an integrated memory controller it was way ahead of AMD's offerings for quite some time. Intel has had a lead that AMD is only now catching up to and it's taken massive security mitigation to close that gap. Keep in mind that technically, Ryzen only trades blows with Haswell which is over 5 years old. Core 2 being built on P6 doesn't change the fact that it was superior to AMD's offerings at the time. What AMD did with Bulldozer is essentially what Intel did with Netburst. It bet on the wrong horse and it didn't pay off.

Yes, Intel's blunder with 10nm is a big deal. Probably a bigger deal in the long run, but the process node is far from all their is when it comes to performance. Intel's making claims that Sunny Cove has got an 18% IPC advantage over Skylake, but obviously, its made statements about IPC vs. previous processors before which were erroneous at best.
 
Much love for the commentary! (Wish I could have this “debate” in real-time with you, maybe we should collaborate on some stuff, seems we have a lot in common...)

Ryzen 2700x is at a broadwell level of performance, more if you count SMT. It’s not as far behind as you make out, clock for clock. The big thing with Zen2 is AVX/AVX2, in fact I would say the single biggest thing in processors these days is AVX.

While used somewhat lightly in games, productivity software relies on it heavily, this is increasingly being used by consumers . You see this when you compare the 8 core/8 thread 9700 against say the 1800x or even 2700x in things like visual studio, obs, davinci, adobe, autoCAD etc. Content creation is becoming a much bigger thing for the average consumer.

High school subjects over here (in Australia) now have non linear video editing and autoCAD (inventor, cad, maya) as part of the curriculum.

No longer is it just the pros using this software. Especially when anyone can download davinci for free, or get autoCAD as part of their university course for free.

I would like to see more reviewers using Autocad/Davinci/Adobe Premiere/Excel (with monte carlo)/Visual studio etc, as right now puget systems seems like the only place to get this sort of info. https://www.pugetsystems.com/all_articles.php

Personally, as a qualified digital marketer, and as an embedded systems designer, game performance matters a little (as mentioned I'm still a gamer, albeit not as much these days), but productivity performance matters more.
 
Last edited:
Become a Patron!
Back
Top