![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
[Passmark] Ryzen 5 5600X claims the first place in single-thread performance in Passmark
- Thread starter luisxd
- Start date
D
Deleted member 89018
Guest
Hmm. I’ve got a 3700x in an sff build. I don’t do as much with multi-core workloads as I thought I’d be doing when I put this box together, maybe dropping to 6 cores could put me in better temps and better performance for the few older games I still play.
interesting...
/omoshiroi
interesting...
/omoshiroi
legcramp
[H]F Junkie
- Joined
- Aug 16, 2004
- Messages
- 12,365
When are official reviews out.... this is getting juicy.
cybereality
[H]F Junkie
- Joined
- Mar 22, 2008
- Messages
- 8,789
This is huge. AMD is going to kill it.
Rockenrooster
Gawd
- Joined
- Apr 11, 2017
- Messages
- 955
What's the 10900k's boost clock? 5.3ghz?
And the 5600x beats it by 10% at a lower clock (4.6ghz was it?)
And the 5600x beats it by 10% at a lower clock (4.6ghz was it?)
fullvietFX
[H]ard|Gawd
- Joined
- Sep 1, 2004
- Messages
- 2,015
Athlon XP all over again.
So when Intel had the lead in gaming it did not matter as it was "close enough". Now that AMD will have the lead nearly everyone will claim it "destroys" Intel. I think people need to go back and look at AMD slides as other than League of legends and CSGO the lead AMD has in gaming with even the 5900x over the 10900k in the other games it showed was around 5%. Bottom line is that Intel had a bigger overall lead before but now all of sudden an even smaller lead is such a big deal just because its AMD. Hell all Intel as to do is get a modest 10% IPC increase with Rocket Lake and it will be in the overall lead in gaming again. I am thinking about getting a 5900x myself but these silly exaggerations based on a couple of outliers needs to stop.
Attachments
Well here is the thing... Gaming was the ONLY win Intel was previously claiming. So AMD was not just the better value proposition they where flat out the best option for everything but single threaded performance. Which means little to very few users really. Now Intel looses in every check box.
This was the last benchmark Intel could put on a Marketing slide without resorting to lying.
How does Intel marketing dept spin this ? Rocket Lake doesn't matter cause it doesn't exist yet... by the time Intel is shipping those in 4-6 months from now, AMD will probably be ready to drop a Zen3+. Back ported 14nm+^3 is not going to cut it. Even if Rocket lake pulls ahead a gain slightly in single threaded its not likely going to win multi threaded... its almost for sure going to be going into a dead end board, and likely AMD will already have Zen3+ rumors swirling as a last hurrah for the AM4 platform.
I suspect Rocket Lake won't ship in major volume. They will release it make a bunch of the mid range model and produce just enough of the highest end to perhaps regain the single thread crown and have something on offer before they get to Alder Lake. At this point Intel is in the back seat... and should really focus everything they can on making sure Alder Lake and their move to DDR 5 is ready for what AMD has cooking for the Zen 4 gen. (this reminds me very much of the Athlon days... Intel needs to take the Knee on their 14nm respins and backports of failed 10nm parts. They seem to be falling in to the same pattern of mistakes they made with P4... they propped P4 for to long and ended up behind back then as well.)
Last edited:
AMD will have to have Zen3+ fairly soon because I have no doubts Intel will easily take back the overall gaming crown with Rocket Lake. And then they have Alder lake to contend with too before Zen4 launches. It is going to be hard to make decisions on what to get as about every 6 months we will be seeing one company or the other do something it seems. I guess its not like I am going to suffer with a 9900k but I did want to go ahead go pcie-4 before too long.
The funny thing is that slide shows 1080p performance. For the people on 1440p or 4k that difference means nothing. I'll keep my 10700k running 5.2ghz, not even considering going to AMD
Its really not a hard decision for almost every gamer I have talked to the last few months. AMD is going to OWN Intel for the next year at least. What decision is there to make... at 4k resolution which is finally realistic now with this next gen of GPUs, Zen 2 is basically = with Intel in gaming today. The only flag Intel has been able to plant was single threaded super high FPS low resolution gaming. Which matters to very few gamers. Now Zen 3 looks to own that crown as well. Intel wins in nothing accept cost at this point.
I have talked to a few gaming friends.... and at least a couple are planning to go all out on massive AMD 12/16 core Zen3s with fancy new BNavi or 3080. Another option a few of my friends have been throwing around... is grabbing Zen 2 parts at a discount. Great thing about AMD its super easy to throw a fire sale 36/37/3800 into a good 570 board and wait for a good deal or a + refresh. I know its anecdotal but I would say of the 20 or so folks I game with regularly almost half of them are planning to build something new this year and everyone of them is leaning AMD. Another few people myself included are likely to drop a Zen 3 into their older AMD boards bios willing.
I do hope Intel pulls it together for Alder Lake... if not AMD is going to charge us up the nose for Zen 4. I hope Intel isn't wasting a ton of resources on Rocket Lake frankly... it going to max out at 8 cores, its going to run hot, it MAY win single core perhaps. Its a waste of resources. Intel should be focused 100% on getting Alder lake out as soon as possible. They can't afford for that gen to ship later then AMDs Zen 4 platform. AMD beating them to PCIe 4 was bad PR but not the end of the world... if AMD beats them to DDR 5 that is another can of worms.
I dont need to upgrade from a cpu point anytime soon. I was simply referring to the fact that you can have something better to look forward to around every 6 months with each company leap frogging each other in gaming which had not been the case. So if gaming is the main concern then someone that does not go Zen3 right at launch can look forward to Rocket Lake and then Zen3+ and then Alder lake and then Zen4. My only need is going with faster pcie at some point so I am in no hurry but the 5900x looks tempting...
Krenum
Fully [H]
- Joined
- Apr 29, 2005
- Messages
- 19,189
That's HUGE for AMD which has always been plagued by single core under performance. I will definitely consider this chip if those numbers are true. Good price point too.
Krenum
Fully [H]
- Joined
- Apr 29, 2005
- Messages
- 19,189
In the case of Intel backports to 14nm++++++++++++++++++++++++++++++++++ I don't think anything will matter much. Intel hasn't had anything but tiny IPC increases for years now but IPC increases are the only way they'll be able to gain any performance because 14nm is played out and I don't see any increases in clock speed coming in the future which is how Intel has kept the single thread performance crown. Also keep in mind that if there are any noticeable increases in IPC it may negatively affect max clock speeds. There are many examples over the years of more efficient architectures backsliding on clock speed due to limitations. Netburst was a very inefficient architecture but could clock high to mask the inefficiencies for a time. Core2 came out as a massively more efficient architecture with much higher IPC but took a clock speed hit to due so.
Another factor to consider is most major IPC advancements tended to come with new or massively revised technology. The A64 on-die memory controller was a huge one. I doubt Intel has something new up their sleeve for architectures in the near future which is going to hamper any chance for major IPC increases which leaves them needing high clock speeds which the likelihood of being able to increase clock speeds in any noticeable amount is non-existent.
Zen is a perfect example of this. First generation was efficient but had low clocks. Zen+ increased IPC and clock speeds but was little more than a slight iteration. Zen2 was a bigger redesign which increased IPC again and clocks again but to do so required a new process node to make all that extra cache affordable while being able to keep power needs in check. Now with Zen3 we're seeing increases in IPC and clock speeds again. How much of this is due to changes in architecture or simply by going from 4 core CCX to 8 core CCX we don't know yet. We do have hints that the 8 core CCX alone will be a nice boost by looking at the 3300x which eliminated a lot of the latency penalties of the 6+ core CPUs.
I don't think Intel is going to have anything decent to compete with until they can get the 7nm node working. 10nm is dead in the water and can't seem to scale in number of cores or clock speed and 14nm+++++++++++++++++++ is way too long in the tooth and played out to be able to realize any additional gains. At best we're looking at an Intel who continues to stagnate for at least another year if not two.
Unless all these leaked benchmarks are lies, AMD is looking to be in a really good spot for a while. Zen has been scaling well in IPC and clock speed and looks to be amazing with the upcoming release of Zen3. What we've been seeing is exciting and AMD has not been overhyping products for a while and never really with Zen but I will do no more than remain cautiously optimistic until the release of Zen3 and I see trusted reviews on what it can and can't do.
spacedrone808
Limp Gawd
- Joined
- Mar 28, 2017
- Messages
- 195
Nice to see fat intel falling apart
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,669
Looks fake.
Ryzen 5600 line, no frequency specified. Sloppy photoshop?
The Intel 10900 looks underclocked? Even more sloppy photoshop.
Granted, I have never used that benchmark, maybe those other lines are built in goalposts. But if so, a typical 5.1Ghz 10900x will score 4376, or 25% higher than that Ryzen 5600 score...
C'mon people, you have brains, use them.
Ryzen 5600 line, no frequency specified. Sloppy photoshop?
The Intel 10900 looks underclocked? Even more sloppy photoshop.
Granted, I have never used that benchmark, maybe those other lines are built in goalposts. But if so, a typical 5.1Ghz 10900x will score 4376, or 25% higher than that Ryzen 5600 score...
C'mon people, you have brains, use them.
I'd be much more skeptical of the single sample size but you can check the live chart yourself. https://www.cpubenchmark.net/singleThread.html. Found that by using my big brain for 3 seconds on google.
And seriously you aren't familiar with passmark?
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,669
No and I don't measure my penis either...
But thanks for that link, looks like built in goalposts. So point #2 still valid. And why no other results? A bench that leaves out overclocked results seems less useful to me.
/dev/null
[H]F Junkie
- Joined
- Mar 31, 2001
- Messages
- 15,182
The multi-core score is pretty much idential to the 3700x which is pretty interesting.
Comixbooks
Fully [H]
- Joined
- Jun 7, 2008
- Messages
- 21,473
I have no reason to think its fake. Just has a winner on their hands. Going AMD next time around.
Proper benches are only a few days away.... and real hardware enthusiasts will be uploading their scores. I expect the win for AMD is actually going to be much more convincing. Expect the higher end AMD chips to score even higher with faster clocks.
We'll know in 4 days.
I think everyone agrees synthetic benches are useless... outside of PR. After a year or more now of Intel spin PR, Cinebench doesn't count but our synthetic is legit type crap marketing. Its just nice to see the last legit legs Intel PR has to stand on get kicked out.
Yes the proof will be in the real world benchmarks. I have a feeling Intel is going to be even more disapointed on that score... still I expect Intel is going to still win a few games anyway. Games heavily optimized for Intel and not AMD are still going to swing in their favor. However the more recent the games the more likely they will be just as optimized for team Red. This isn't like Zen 1s launch where software developers had been ignoring AMD for years.
Rockenrooster
Gawd
- Joined
- Apr 11, 2017
- Messages
- 955
Well, if you are a gamer first then yes, Synthetics and such don't mean much. Only games would matter...
However most any benchmarks are good to have as it allows us to make informed decisions. The more the better.
I'm not a gamer first, but I do like to play them, I have lots of workloads, so I ended up with a 3900x. I just wish I had more money lol.
My point was about the members with such extreme stances. Not walking the middle ground like many people do. Like yourself who find benches agreeable but not all emcompassing. Many on the flipside think benches are totally useless and have no indication of real world performance while offering zero evidence to substantiate thier claim. Almost certainly these stances are fueled by nothing more than internet poison.
blackmomba
Gawd
- Joined
- Dec 5, 2018
- Messages
- 773
It's called coping
Ill also add that synthetic benchmarks never matter unless they feed your confirmation bias.
Some are like "oh look AMD beats intel in synthetic - glad synthetics dont mean shit vs real world" after having just bought a 10900k. When a week earlier, had they not purchased a chip, they would be like "oh wow AMD smoked Intel in benches - glad Im getting an AMD"
Better performance maybe, better temps? Doubtful, they're both 65w parts and will both pull ~88w under full load. I don't see you getting much different temps either way, but single core should be better with a 5600x. Enough to sell your 3700x and buy a 5600x? I don't think I'm going to bother, I have 2 3700x builds and I'm sure they'll still work just fine when the 5600x releases. Maybe if they ever release a 5700x I'll bite, but I'm not stepping up to the cost of a 5800x and I'm not dropping 2 cores out just for a little bit of extra single core that only affects me slightly. Sounds like you're a bit more reliant on single core, so benchmarks should tell you if it's a good change or not. Figure you can probably get ~$200 for a 3700x, but we still don't know availability of 5600x or if it'll be in enough stop to be close to MSRP. I would expect to come out of pocket at least ~$100 so only you'll know if that $/perf is worth it.
I think the argument was AMD was close in most cases, but won hands down in threaded loads, so it was ok that it was a little slower sometimes, but faster in others. *IF* AMD is able to beat Intel in both single and multi, then this is a much bigger deal than only winning in just gaming like Intel has been claiming for some time. If AMD can win in gaming, Intel has no leverage left to play. There is no doubt AMD will continue to win in threaded work loads. This is all with large grains of salt as we don't know the actual performance yet, just a few leaked benchmarks that are pointing to the gap disappearing and AMD possibly even winning more so. That said, Intel can/will just adjust prices to match performance, but to the people that say "all I do is game so I want the fastest at all cost", if AMD can flip this group (or at least some of this group) then it's a huge win for them... it's not a huge win for Intel because Intel is the current leader. This in no way means Intel will be going anywhere, but it'll be hard to recommend Intel for any use case if this is correct. Before it was kind of split, if you gamed only, go Intel, if you have more heavy threaded work loads, go AMD. If one company can take both of these titles, it's a huge win and means the only way to compete would be on price.
It is interesting but it follows the same pattern as last time with the 3600x beating the 2700x in single threaded performance but trading blows in multithreaded performance(though it was a bit behind overall).
MavericK
Zero Cool
- Joined
- Sep 2, 2004
- Messages
- 31,742
No, they're not. No one buy this...I don't want it selling out in 3 seconds like the fuckin' Ampere, or scalpers getting ahold of it.
MavericK
Zero Cool
- Joined
- Sep 2, 2004
- Messages
- 31,742
Huh? That's the Big Navi reveal, I don't know if we'll hear any more about Zen 3 until release on Nov. 5.
Your right I got mixed up on the dates. Still Nov 5 isn't far away. Less then 2 weeks anyway. Probably see real legit reviews before then.
cybereality
[H]F Junkie
- Joined
- Mar 22, 2008
- Messages
- 8,789
If it is what we hope, it will sell out in 3 seconds, yes.
I don't even need any new parts, but I might buy one anyhow for a future Zen 3 rig in case the rumors of a limited run are true.
MavericK
Zero Cool
- Joined
- Sep 2, 2004
- Messages
- 31,742
Hopefully not, it's the last piece I need to complete this build.
I guess worst-case I can buy a Zen 2 someone is selling off and wait until later to upgrade to Zen 3. But I'd like to just get a 5900x or 5800x and be done with it.
cybereality
[H]F Junkie
- Joined
- Mar 22, 2008
- Messages
- 8,789
Sorry, I meant that the 6800 XT might be limited.
I'm looking at the 5950X, should help a lot with compile times and rendering.
I'm looking at the 5950X, should help a lot with compile times and rendering.