GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,752
Just Plain Mean
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
AMD Announces BIOS Fix for Ryzen 3000 Boost Clocks - hits September 10th
- Thread starter Sabrewulf...
- Start date
- Status
Just Plain Mean
thesmokingman
Supreme [H]ardness
- Joined
- Nov 22, 2008
- Messages
- 6,617
What's the deal with this crap? One minor issue that is all it is, minor. It will be addressed. Yet that is all it takes for people to piss and moan every single time. Intel screws you over for decades... I don't get it. Go back to Intel?
D
Deleted member 89137
Guest
At the end of the day it's a number. A number that is essentially irrelevant in 2019. Who the hell is using single threaded applications anyways....
Criticalhitkoala
[H]ard|Gawd
- Joined
- Apr 17, 2015
- Messages
- 1,987
I've been happy with both my 3700x and 3900x but neither hit max boost at all. So looking forward to the update.
thesmokingman
Supreme [H]ardness
- Joined
- Nov 22, 2008
- Messages
- 6,617
It could also be your board. The same chip will boost differently on diff boards... bios'.
AMD felt it was relevant enough to print on the box. You'd have to ask them.
Dayaks
[H]F Junkie
- Joined
- Feb 22, 2012
- Messages
- 9,773
A lot of games still have one heavy thread. Not everything can be in parallel.
D
Deleted member 89137
Guest
I'm not saying it's not an issue. I'm just saying people are acting like this is the biggest scandal in the history of the world. Again, who the fuck runs single threaded apps?
purple_monster
Gawd
- Joined
- Jun 1, 2018
- Messages
- 691
i cant think of any without going retro
Criticalhitkoala
[H]ard|Gawd
- Joined
- Apr 17, 2015
- Messages
- 1,987
To be honest it's been the same as most other systems I've built including Intel. Just more unique. Like ram issues. Sticky thermal paste. Etc.
But system problems are always something to consider on all newer architecture. If you really need stability just buy three year old stuff that was once top of the line.
d3athf1sh
[H]ard|Gawd
- Joined
- Dec 16, 2015
- Messages
- 1,233
na just stuff like spectre/meltdown.
The same chip may boost differently if you change the RAM too. Virtually everything can effect boost clocks either way on these CPU's.
Which is it, it will be addressed or people shouldn't complain about it? The former doesn't happen without the latter.
thebufenator
[H]ard|Gawd
- Joined
- Dec 8, 2004
- Messages
- 1,383
3 less than Intel has for Meltdown and Spectre security vulnerabilities.
ahah I have to ask, is this number accurate ? You would win the internet award of the day in the category on how to pour gaz into fire
I like your style lol
EDIT: Wow corrected that typo
Last edited:
What's not to get? We've always known that if you put a cooler on an intel proc that can support that processor's TDP, you'll get to run on at the rated boost clocks (barring weird crap. My work PC is a Dell Optiplex with an i7-8700 (non-K), and if I push it hard (Prime95, for example), it will be strictly limited to the actual 65W TDP, unlike enthusiast motherboards. That means it will not make the all-core boost. IIRC it will actually run about 100MHz below the base speed with a sustained all-core load).
WIth Zen and Zen+, you also pretty much always got the rated clocks, including XFR (and a lot of time coudl beat them. The 1600X was rated for 3.6/4.0 (the latter only on two cores), +100MHz with XFR, and not only could you get 3.7GHz all core all day long (typically with two out of six cores bouncing between 3.7 and 4.1), but you could pretty much guarantee an all-core overclock at around the rated boost speed. Now suddenly with Zen2 a lot of people can't see the rated boost speed on even a single core. I have one of those. I can't get any core to go past 4.25GHz on a 3600X (except ephemerally) with a 360mm AIO. Is it a huge loss? No. Would I support a class action lawsuit? Also no, and arguing for one is kinda dumb. But if AMD said "sorry about that, here's $20" I'd take it.
How many times Intel promised 5GHz and didn't deliver ? Stay tuned for the latest release of 5GHz. There's also a huge difference between all-core at 4.3GHz and having 1 core reach the 4.6GHz. People will get what they asked for thru an update and then complain that AMD slowed their all-core boost only to achieve one-core max boost (mind you) clock. That's my 2 cents, at this point we shall see but I prefer higher all-core OC than 1 core hitting more megahurtz.
May I add, can you define Intel TDP lol ?
I dunno. How many times *did* they release a chip that couldn't hit the rated speed (I have no idea)? This isn't the same thing as reality not matching up with prerelease hype.
That's not what I said. My 3600X will not ever hit the 4.4GHz rated single-core boost clock, ever, on any core. Not even the gold star one listed by Ryzen Master. Try to keep up.
So would I. But, again, my 1600X could do all-core 4.0 all day long. My 3600X can't even hit 4.4 on ONE core at a time. But it does come close and I acknowledge that.
Sure. https://ark.intel.com/content/www/u...-8700-processor-12m-cache-up-to-4-60-ghz.html says the 8700 has a 65W TDP, meaning it should sustain the base clock (3.2GHz) at 65W all day, with a 65W-or-bigger cooling solution. It just so happens that my work Dell can't hit that, but that's probably Dell's fault, and it only misses by 100MHz, and only in torture tests like Prime95 small FFT (nothing I do with the thing in my day job will every hit it that hard, so it typically reaches the 4.3GHz all-core boost), and I didn't pay for it anyway so I don't really care. We all know that's actually PL1 or whichever specific name it is, and everyone here understands that getting the rated boosts blows the rated TDP out of the water.
The only question that is important. Is your 3600 better then your 1600 ? My 3600x also comes close but not quite... still noticeably fastest upgrade I have seen since my first 1ghz thunderbird system ages ago. Yet still being one of the cheapest CPUs I have bought ever. All things considered I am not sweating double digit mhz single thread boost targets. (and of course like everyone else I will install the next bios and be checking lol)
I mean lets not get lost in the weeds on clock speed. Clock speed is not relevant to performance when comparing 2 completely different CPUs. It never has been ever ever... IBM sold Power chips at 5ghz 12 years ago, there is a reason we didn't all switch to power.
Everyone is in for a real shock when 5nm chips hit and clock in sub 4ghz.
Intel when they finally get around to actual next gen parts at 7nm next year... are not going to be clocking much above 4ghz either. Physics are physics... when your dealing with interconnects that close together running them an high mhz leads to electron jumping and bad calculation. You can build better error correction but at some point your solution to one roadblock is a not any better speed bump. Past 5nm (and perhaps even at 5nm) we are going to start seeing quantum tunneling force some innovative design fixes and one simple fix will simply be lower clocks.
Well it runs 200-300MHz faster, so I guess technically the answer is yes.
I, personally, am not, and I've tried to make that clear. I'm just mildly disappointed, because it's my first experience with a chip that couldn't reach the speed on the box. I had hopes that, like Zen, I would be able to get a manual all-core OC close to the boost speed, like my 1600X could do.
I would suggest that Zen and Zen2 aren't so different that "completely different CPUs" is entirely fair (or at least, not as fair is it would be if I were comparing Zen 2 to Bulldozer or Intel or ARM).
Well it runs 200-300MHz faster, so I guess technically the answer is yes.
I would suggest that Zen and Zen2 aren't so different that "completely different CPUs" is entirely fair (or at least, not as fair is it would be if I were comparing Zen 2 to Bulldozer or Intel or ARM).
There is a lot different between zen/zen+ and zen2.
I mean nothing previously housed components in 2 separate not even on the same mfg node silicon bits. Double the L1 L2 and L3 cache spaces... Instruction cache set from 4 way association to 8. the Micro-op cache was upped to 4k... brand new branch prediction engine with a much longer history 512 entries now up from 256 I believe... support for avx256 calculations in one clock cycle (twice the speed of zen)... the alu (arithmetic logic unit) can now calculate 7 operations per clock up from 6 (ok not major but it's a new design)..
They made a lot of little changes... and some pretty significant ones. As well as releasing a completely new packaging method. It's pretty clearly not just a simple die size reduction with a few tweaks. Zen2 destroying zen+ in most benchmarks has little to do with a couple hundred extra mhz.
They really are completely different chips. Its a testament to AMDs engineering that with all the changes they are still running on in the same socket on the same chipsets. I expect over the next year we are going to see some pretty good zen 2 software optimizations drop. AMD actually included the old prediction engine along side their new TAGA based one. (no doubt some software was optimized already for ryzen... there are potential performance gains from optimzations for zen2 yet) AMD claims their new TAGA engine has around 30% fewer mispredicts... I believe in general software optimized for those newer bits should easily be able to see double digit improvements. Assuming that AMDs sales success is going to translate into more software optimization.
Smoked Brisket
Gawd
- Joined
- Feb 6, 2013
- Messages
- 786
Whoa, let's stop anointing AMD as the peoples champ. They are a billion dollar soulless corp. Intel had superior processors for a decade and so could dictate the market. I am happy there is competition, but this consumer on consumer fanboy bullying needs to stop polluting the forums.
This is an appropriate response to the very real issue.
That's right.
The people who are acting like this is a huge scandal are AMD fanboys who patrol the forums in an attempt to shout down any perceived criticism of AMD products.
The Zen2 launch was an incredible success. I am on that bandwagon. My first complete new build since my child was born is all AMD. But there are some real issues with both the Zen and Navi launches. Issues that should be discussed in forums like this because there is a good deal of knowledge here. I continue to be shocked that it is supposed "AMD fans" who are conducting this "shout down the message" style campaign.
Why does this bother me? These types of tactics are detrimental to competition. They are indicative of a pervasive societal tribe mentality that I had hoped the enthusiast community, specifically these forums, would be immune to. These tactics actually tarnish my excitement over my purchase, even though I try not to let that creep in.
And this is why you buy VIA/S3
Gideon
2[H]4U
- Joined
- Apr 13, 2006
- Messages
- 3,548
You must be new to forums.
ManofGod
[H]F Junkie
- Joined
- Oct 4, 2007
- Messages
- 12,863
Hey, VIA was pretty decent, 12 years ago.
S3? Not so much, in my opinion. As for this thread, I just sit back and watch it play out, that is all that is needed. Not going to go about correcting anyone, one way or another, about fan this or fan that, just going to watch and enjoy my systems at home. (I do not upgrade from the 2600 to the 3600 because my games will not benefit enough to be worth it.)Both VIA and s3 have there pros. But s3 seemed to always get in over there heads and fail hard once the cards came out. They did have a few ok'ish cards after the savage line. But good luck finding any. Multi chrome setup was really cool. As far as I know it was the first multi GPU setup to communicate over the pci-e bus.Hey, VIA was pretty decent, 12 years ago.
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,752
ahah I have to ask, is this number accurate ? You would win the internet award of the day in the category on how to pour gaz into fire
I like your style lol
EDIT: Wow corrected that typo
Lol, no. Intel's last microcode was the 3-2019 that was released in June.
Your'e being cynical.
ManofGod
[H]F Junkie
- Joined
- Oct 4, 2007
- Messages
- 12,863
Nah, just open minded and observant.
I get the feeling of been cheated by that max boost fiasco, seems like AMD is owning it and promised a fix. I just hope they won't sacrifice all core turbo for single core higher one.
I wonder how they determine which core is the star one ? I surely hope they don't disable all other cores to measure the highest clocking one with ideal thermals.
For the TDP part, yeah we all understand it but would average joe do ? And let's be honest, no Intel CPU that have advertised boost clock comes in locked at the base clock to meet the TDP, they will boost higher and blow that TDP out of the water... pretty useless metric to put on the box.
Better than mine at 4325. If all threads are loaded I'm at ~4225, and temps in mid-50s.
I really doubt AMD will sacrifice max boost clock for all core just to increase max boost for a single core. I really chalk this up as a minor hiccup for AMD and seem to be owning it (will see in September update).
Red Falcon
[H]ard DCOTM December 2023
- Joined
- May 7, 2007
- Messages
- 12,445
There are still a shit ton of programs out there that are strictly single-threaded, or are lightly-threaded where having that single boost core would greatly benefit them.
Audacity transcoding, Windows updates, clamd, database applications, many updater programs (both Windows and Linux), certain compression tools, etc.
Believe it or not, your OS still has many back-end programs that are strictly single-threaded.
Last edited:
Factory testing, the way I understand it, and no further details given, but it might be "test each core in isolation". I do know that none of my "best" cores appear to boost any higher than the others.
GoldenTiger
Fully [H]
- Joined
- Dec 2, 2004
- Messages
- 29,656
Many games are lightly threaded as well... Single thread perf is still very important.
Gideon
2[H]4U
- Joined
- Apr 13, 2006
- Messages
- 3,548
If you used the chip you would understand. Single thread boost means just that and almost every game uses more then 1 thread as your cpu has to feed the gpu as well which on Ryzen will take your clock down to 4.2 to 4.4 on average so not hitting 4.6 on a single thread on a 3900X is about meaningless in games. Most use 4 threads so yeah your never going to see 4.6 or 4.4, most will be at 4.3 or lower and one of the reasons you dont see a big difference in game benchmarks on Ryzen 3rd gen. If you could get max boost across 4 threads it might be a bigger deal for some, but 1 thread max boost is pointless except for benchmarks. Biggest thing this bios might help is a few people that have all core speed problems and hopefully PBO will be fixed on the 3900X. But reality this will mostly affect benchmark scores and most people will see no difference in actual use.
Red Falcon
[H]ard DCOTM December 2023
- Joined
- May 7, 2007
- Messages
- 12,445
Outside of fucking video games, single-thread performance is extremely important and still very relevant, regardless of the OS.
I'm amazed so many here think these CPUs are used just for video games and that the single-core boost clocks don't matter for anything else.
We aren't just talking about video games, we are talking about all applications, and video games are just a fraction of that.
The extreme ignorance in these threads is just astounding, hot damn.
Also, most modern games are starting to use 8+ threads now, and run like turds with only 4 threads (cores) available when 8 are more optimized or required.
Your statement, while not totally wrong, would be more accurate circa 2013, not in 2019.
Gideon
2[H]4U
- Joined
- Apr 13, 2006
- Messages
- 3,548
Even in productivity apps most are using more then 1 core a few dont but that is getting rare. As soon as you use 2 cores on a Ryzen with any kind of load your going to slow to 4.4 or less no matter what so it's damn pointless to worry about and you would know that if you used the thing. If you got a app that is totally dependent on 1 thread then why the hell did you buy a AMD chip in the first place, Intel clocks higher. Most people use their computer for multiple things and are running more then 1 task almost all the time and why worrying about a single core max boost is stupid on Ryzen, unless all you do is run benchmarks. Like I said this fix will only likely matter to guys that all core boost is not working correctly on. Also if your using 8 cores or 4 on Ryzen 3rd gen you will be running the same speed so it really doesn't matter.
- Status