Question Anyone else slightly disappointed by the new Ryzen?

[alross016]

Atm I'm slightly disappointed with how the new Ryzens are looking, I was hoping to get a 8/16 but I need to Get a 3800x for that, and there isn't much point in that, 429$. I can buy a 2700 for 200$ atm what is the point apart from the pcie 4.0

 

[xravenxdota]

Depends on the motherboard you bought. MSI said they wont run 3000 series on their first gen ryzen boards.

MSI dont seem to be able to make their mind up
View: https://www.reddit.com/r/Amd/comments/btj9n0/linus_talks_about_zen2_mobo_backward_compatibility/

That msi Rep was not accurate.It's the A320 boards that doesn't have the power phase to run 3xxx series the x370 will be able to run them

 

[xravenxdota]

No offense but it's funny how people biatch over the slightest thing.I personally won't upgrade my 2600 to a 3600 as i don't see the need to.I hardly push this cpu to it's max.But with that being said.Intel charge a premium for HT where amd has smt on about all there cpu's for a much cheaper price.Price to performance intel can't hold a candle compared to amd.I personally would wait for reviews before being disappointing .If they truly have a 15% ipc gain then intel's in real trouble.

 

[InvalidError]

Only possibly questionable upgrade could be from Ryzen 2000series to comparable 3000 like for instance 2600x to 3600x and 2700x to 3700x, all others like 1600 to 3600x or waiting for 3000 instead of buying comparable 2000 series is a no brainer.

It is only a "no-brainer" if you like throwing money at a 20-30% upgrade. Anything under 100% is a no-brainer don't-give-a-damn for me, i5-3470 to Ryzen 3600 is barely passed the threshold where I can be bothered to pay attention.

 

[CountMike]

It is only a "no-brainer" if you like throwing money at a 20-30% upgrade. Anything under 100% is a no-brainer don't-give-a-damn for me, i5-3470 to Ryzen 3600 is barely passed the threshold where I can be bothered to pay attention.

20-30% is quite good nowadays. what were we given lately during last few generations of CPUs 0 to 10% tops and in some cases even some of that was taken away by fixes for security holes. If jumping 3-4 generations is not a no brainer I don't know what is.

 

[nuke20ar]

It is only a "no-brainer" if you like throwing money at a 20-30% upgrade. Anything under 100% is a no-brainer don't-give-a-damn for me, i5-3470 to Ryzen 3600 is barely passed the threshold where I can be bothered to pay attention.

1) Your idea of a "no-brainer" does not define what everyone else thinks of as a "no-brainer."

2) If 100% gain is what you're looking for, you've been very generic. 100% gain in what, IPC? Single-thread? Multi-thread? Frames per second? And if FPS, what game(s)?

A 20-30% upgrade in multi-threaded productivity performance for someone who does even mid-level professional work can easily save hundreds of dollars, or give more free time, or both - and would be worth it for a massive number of people.

 

[InvalidError]

A 20-30% upgrade in multi-threaded productivity performance for someone who does even mid-level professional work can easily save hundreds of dollars, or give more free time, or both - and would be worth it for a massive number of people.

What is the proportion of the world population that does professional stuff at home? Among people I know, that would be under 10% and for the majority of that 10%, most of that "professional" stuff is keeping up with work emails and documentation, nothing that would meaningfully benefit from a 30% upgrade.

As for "100% of what?", that would be overall CPU throughput per dollar for a CPU upgrade and GPU throughput per dollar for GPU upgrade. Ryzen 3600 for ~$260 Canukistan dollars is ~10% higher than my $170 i5-3470 on clocks, gets ~20% more on IPC, ~30% more from HT and 50% more on core count, which makes it only about 63% faster per dollar. Not quite there yet.

 

[nuke20ar]

What is the proportion of the world population that does professional stuff at home? Among people I know, that would be under 10% and for the majority of that 10%, most of that "professional" stuff is keeping up with work emails and documentation, nothing that would meaningfully benefit from a 30% upgrade.

As for "100% of what?", that would be overall CPU throughput per dollar for a CPU upgrade and GPU throughput per dollar for GPU upgrade. Ryzen 3600 for ~$260 Canukistan dollars is ~10% higher than my $170 i5-3470 on clocks, gets ~20% more on IPC, ~30% more from HT and 50% more on core count, which makes it only about 63% faster per dollar. Not quite there yet.

I'm sorry, but your opinion will not dictate the market. If a company is paying someone $80,000 a year to do $120,000 worth of productivity (think smaller to mid-size engineering firms, architecture firms, medium sized businesses, etc), if 5% of their time is spent waiting on a CPU to do work, and you can gain a 20-30% increase in throughput, (1-1.5% increase in productivity), that's a $1,200-1,800 increase per year. That justifies just about any processor.

 

[InvalidError]

I'm sorry, but your opinion will not dictate the market. If a company is paying someone $80,000 a year to do $120,000 worth of productivity (think smaller to mid-size engineering firms, architecture firms, medium sized businesses, etc), if 5% of their time is spent waiting on a CPU to do work, and you can gain a 20-30% increase in throughput, (1-1.5% increase in productivity), that's a $1,200-1,800 increase per year. That justifies just about any processor.

Unless these people are working from home, this is completely irrelevant to home use.

Large enough engineering jobs (at least in the companies I worked at) get delegated to server farms because even a high-end desktop would take too many hours to complete ASIC/FPGA synthesis and simulation jobs. The engineering office PCs are little more than text editors, *nix consoles and Modelsim wave viewers, all the heavy-lifting is done remotely instead of giving every desk an ultra-powerful PC with 64-256GB of RAM that gets grossly under-used most of the time.

 

[nuke20ar]

Unless these people are working from home, this is completely irrelevant to home use.

Large enough engineering jobs (at least in the companies I work at) get delegated to server farms because even a high-end desktop would take too many hours to complete ASIC/FPGA synthesis and simulation jobs. The engineering office PCs are little more than text editors, *nix consoles and Modelsim wave viewers, all the heavy-lifting is done remotely instead of giving every desk an ultra-powerful PC with 64-256GB of RAM that gets grossly under-used most of the time.

> Large enough engineering jobs

cf my post, "think smaller to mid-sized engineering firms, architecture firms, medium sized businesses"

You seem to have taken issue with my statement that "a 20-30% upgrade in multi-threaded productivity performance ... would be worth it for a massive number of people." Can you clarify what you disagree with on that front?

 

[InvalidError]

You seem to have taken issue with my statement that "a 20-30% upgrade in multi-threaded productivity performance ... would be worth it for a massive number of people." Can you clarify what you disagree with on that front?

The initial post I replied to said nothing about productivity or corporate purchases, which is a bit different since it may directly affect productivity to some extent, is tax-deductible and the depreciation is a further tax write-off on top. This forum is 99% individual people discussing their own personal computer use. The average person and even many enthusiasts who pay out of their own pockets for a mostly personal use PC isn't going to bother with a 30% upgrade, which is why there is no shortage of people using 5+ years old PCs around here and 80% of systems on the Steam survey are quad-core or less.

 

[King_V]

> Large enough engineering jobs

cf my post, "think smaller to mid-sized engineering firms, architecture firms, medium sized businesses"

You seem to have taken issue with my statement that "a 20-30% upgrade in multi-threaded productivity performance ... would be worth it for a massive number of people." Can you clarify what you disagree with on that front?

To be fair:

Anything under 100% is a no-brainer don't-give-a-damn for me.

 

[nuke20ar]

The initial post I replied to said nothing about productivity or corporate purchases, which is a bit different since it may directly affect productivity to some extent, is tax-deductible and the depreciation is a further tax write-off on top. This forum is 99% individual people discussing their own personal computer use. The average person and even many enthusiasts who pay out of their own pockets for a mostly personal use PC isn't going to bother with a 30% upgrade, which is why there is no shortage of people using 5+ years old PCs around here and 80% of systems on the Steam survey are quad-core or less.

Now you've shifted the goal-posts, saying that you are referencing only personal use PC. Which, sure, for checking email and watching Ariana Grande videos, isn't going to matter. But for a massive number of people, it will. So your "It is only a "no-brainer" if you like throwing money at a 20-30% upgrade" statement remains problematic. A 20-30% upgrade matters for a massive number of people.

 

[nuke20ar]

To be fair:

You forgot the sentence before that, which was what I took issue with.

It is only a "no-brainer" if you like throwing money at a 20-30% upgrade.

That is a blanket statement. And it is wrong. His next statement, which you quoted, references a 100% upgrade (rather than 20-30% upgrade), and is clearly an entirely separate entity, and is specifically stated as his opinion. The first statement was presented as a statement of fact and is wrong.

 

[drivinfast247]

It is only a "no-brainer" if you like throwing money at a 20-30% upgrade. Anything under 100% is a no-brainer don't-give-a-damn for me, i5-3470 to Ryzen 3600 is barely passed the threshold where I can be bothered to pay attention.

"100%" lol. All I want is double the performance for half the cost. If that isn't possible I'm not upgrading!

Some people use their PC for more than browsing the internet and checking emails.

 

[InvalidError]

"100%" lol. All I want is double the performance for half the cost. If that isn't possible I'm not upgrading!

Some people use their PC for more than browsing the internet and checking emails.

20 years ago, double the performance for half the cost was the norm every two to three years. Now it takes about seven years to double the performance at a given price. 30% over 2.5 years is pathetic compared to what it used to be.

BTW, a seven years old i5-3470 (what I currently have) is more than enough to do just about anything worth doing at home with just a little patience and a thing called multi-tasking - if one program is stuck on a long CPU-intensive operation, I can just tab out to do other stuff until it is done, don't really care about how long it takes as long as it finishes before I run out of other stuff to do. With 32GB of RAM, I have plenty of memory to spare to play a game while that long operation chugs along in the background with whatever CPU-time is left.

 

[-Fran-]

There's a lot of "get off my lawn" in this thread as of late, lol.

Cheers!

 

[DMAN999]

^ LMAO !!!!

I personally am Not disappointed in the new Ryzen 3000 Series CPUs because so far we don't really know how well (or badly) they will really perform.
My take on it is that since I am Very happy with my current R5 2600 rig I'm not that worried about it yet.
I'll be looking at the 3600/3700 CPU's sometime in the next year and if one of them offers me an increase in performance that I feel is worth my money I will upgrade then.

 

[SgtScream]

I'm going to stick with the R7 1700 until Intel counters. I have plenty of processing power to hold me over and there doesn't appear to be any upcoming competition on the blue side any time soon.

 

[TCA_ChinChin]

20 years ago, double the performance for half the cost was the norm every two to three years. Now it takes about seven years to double the performance at a given price. 30% over 2.5 years is pathetic compared to what it used to be.

BTW, a seven years old i5-3470 (what I currently have) is more than enough to do just about anything worth doing at home with just a little patience and a thing called multi-tasking - if one program is stuck on a long CPU-intensive operation, I can just tab out to do other stuff until it is done, don't really care about how long it takes as long as it finishes before I run out of other stuff to do. With 32GB of RAM, I have plenty of memory to spare to play a game while that long operation chugs along in the background with whatever CPU-time is left.

I know it's a dead horse but Moore's law is dead. CPU tech doesn't advanced as quickly as it did before because it can't. Because of that, I would venture that many people are used to this kind of upgrade cycle and are willing, even happy for a 30% upgrade over 2.5 years. In fact, that's better than it was during sandy to coffelake times. People that expect that kind of performance increase from decades past might be disappointed, but times change. Until there is significantly different technology introduced soon, this has been (at least for a decade), is, and will be the norm.

 

[InvalidError]

I know it's a dead horse but Moore's law is dead. CPU tech doesn't advanced as quickly as it did before because it can't. Because of that, I would venture that many people are used to this kind of upgrade cycle and are willing, even happy for a 30% upgrade over 2.5 years.

If people embrace paying more for increasingly small improvements, that would be a problem. The far more likely outcome IMO is that you are going to see even more people holding on to their PCs for 5, 7 or even 10 years instead of "being happy" with 30% upgrades. There is no shortage of people on THG forums still happy with their Sandy/Ivy/Haswell or even older CPUs in some cases, games are barely getting to the point where quad-core i5 may be marginal for a casual gamer.

 

[knickle]

If people embrace paying more for increasingly small improvements, that would be a problem. The far more likely outcome IMO is that you are going to see even more people holding on to their PCs for 5, 7 or even 10 years instead of "being happy" with 30% upgrades. There is no shortage of people on THG forums still happy with their Sandy/Ivy/Haswell or even older CPUs in some cases, games are barely getting to the point where quad-core i5 may be marginal for a casual gamer.

I myself am still rocking a 3770k from November 2012. Every year I say to myself, "This will be the year I upgrade!" But that never happens. Every year I also ask myself, "Is this the year that my PC will stop booting on me when I come home from work?"

One of the things to consider is that grandfathered motherboards don't get BIOS updates. All of those BIOS patches we hear about to address security flaws will never get fixed for us. Even though they may not directly affect people like you and I, it doesn't exactly leave me feeling warm and fuzzy inside when I go to sleep at night.

What I do like about my aged PC is that most of the OS kinks are worked out and glitches are rare. On the negative, I recall the time I played with Blender, and a 10 second animation I made (for fun) took days to render at decent quality. Then there's those personal Blu-ray backups I rendered with handbrake which caused some hair loss. lol

Yeah, so maybe this year will be the year for me.... maybe.

 

[TCA_ChinChin]

If people embrace paying more for increasingly small improvements, that would be a problem. The far more likely outcome IMO is that you are going to see even more people holding on to their PCs for 5, 7 or even 10 years instead of "being happy" with 30% upgrades. There is no shortage of people on THG forums still happy with their Sandy/Ivy/Haswell or even older CPUs in some cases, games are barely getting to the point where quad-core i5 may be marginal for a casual gamer.

I'm not saying it's a good thing to settle for smaller improvements, I'm dimpling saying I think that's the norm, and regardless, people's expectations will lower. There are people that are holding onto sandy-haswell, but this will be the norm. I don't see people another 10 years from now expecting anything close to >30% performance improvement per generation.

 

[King_V]

@knickle - took the words right out of my mouth.

I think the only thing that is tempting me is that my Haswell Dell XPS 8700 doesn't have an M2 slot... but honestly, for my current uses, it's not falling short at all.

What might do it is if my son's PC at his mom's house (Sandy Bridge Dell XPS 8300) finally stops meeting his needs. Then I'll upgrade, and he'll get what I have.

His PC at my place is arguably equal to (but newer than) my machine.

 

[nuke20ar]

If people embrace paying more for increasingly small improvements, that would be a problem. The far more likely outcome IMO is that you are going to see even more people holding on to their PCs for 5, 7 or even 10 years instead of "being happy" with 30% upgrades. There is no shortage of people on THG forums still happy with their Sandy/Ivy/Haswell or even older CPUs in some cases, games are barely getting to the point where quad-core i5 may be marginal for a casual gamer.

Good! This is great for the majority of consumers.

Both the need for increased processing power from a "general PC" standpoint as well as the costs for said processing power (cf Ryzen 1600 for $79) are so low. This is a great time for general PC users because, for general tasks and even most light gaming, we simply don't need more processing power (even if it were advancing as fast).

This is why Intel and AMD are focusing on the rapidly growing HEDT and server customers. Doesn't make sense to target an audience that is likely to see their PCs run well for 5-10 years.

 

[InvalidError]

Now you've shifted the goal-posts, saying that you are referencing only personal use PC. Which, sure, for checking email and watching Ariana Grande videos, isn't going to matter. But for a massive number of people, it will. So your "It is only a "no-brainer" if you like throwing money at a 20-30% upgrade" statement remains problematic. A 20-30% upgrade matters for a massive number of people.

As an employee, would you personally throw $600 at a PC to improve your productivity by 3% by reducing the the 10% of the time you are waiting for your PC for things you can actually just sit in front of the PC and wait for by 30%? No, you'd just keep using whatever the company provided and unless it is grossly inadequate, you wouldn't mind it much. A 30% improvement still does not really matter to you directly, it only matters to the company if it thinks it'll get that much extra productivity out of you. For any operations longer than that, most companies have no shortage of ancillary work to do in the meantime and the net productivity gain is even lower.