It will take more than 3 months for the market to show increased confidence in AMD. They are arrow pointing up though. All the new OEM deals are a sign of that.

Agreed. Even the fact there's a laptop with a socket AM4 1700 in it shows a pretty promising future. The fact Intel created an i9 after so many years is another sign that AMD is a serious contender. The problem currently is convincing the consumers; people look at the gaming benchmarks, dismiss everything else, and deem Ryzen an inferior product despite not needing to play their games at 90FPS. That's currently hurting their image.

Hopefully we will se re-reviews once all the BIOS and game bugs are ironed out. Ryzen is a great product with the typical AMD launch. (Sloppy is to kind of label)

@TieSKey: so i flashed the latest bios but with the older bios i had achieved 44 seconds for the same file so that was 2 seconds better than a i5 6500. With the newer bios i have 46 seconds so that's a tie. 🙁 So i loose on the hdd benchmark too with a pci-e 256 gb nvme vs a 64 gb sata 3 Ocz Vertex 4. 🙁 The only thing that i got the R7 1700 was because it was better in compression vs a 7700k. If i just bought the 7700k i would have saved the cost of the motherboard and the videocard. But i think i was better if i just kept my 6600k and forgot the update and i would have 750 $ now in my pocket. :bang: The only good thing about the latest Agesa 1.006 is that i can keep my ram @ 1.35v now (down from 1.43v).

I think u are missing the point. Did u check cpu utilization? if it's not 100% on your 1700 (@ the thread count u want), then u can't use it to compare CPU cuz the bottleneck is somewhere else.

I think u are missing the point. Did u check cpu utilization? if it's not 100% on your 1700 (@ the thread count u want), then u can't use it to compare CPU cuz the bottleneck is somewhere else.

That's not true... If all threads are below 100% utilization, then the bottleneck is elsewhere. But if any one thread is maxed out, the CPU is the bottleneck.

That's not true... If all threads are below 100% utilization, then the bottleneck is elsewhere. But if any one thread is maxed out, the CPU is the bottleneck.

In that case, the outdated, poorly created program is the bottleneck. Not the CPU, as the CPU would have more to offer, but the program is too basic to understand how to use it.

I have the latest 7zip, even this board's review (of the R5 1600) shows maximum 27 mb/s for the R7 1700 : http://www.guru3d.com/articles_pages/amd_ryzen_5_1600_review,11.html I usually get 26 mb/s and 1 time i saw 27 mb/s but the i5 6500 is hitting 25-26 mb/s too. I will test to see cpu utilisation but i think i've got a poor R7 or my brother's got a golden i5. 🙁

Boy you are focusing on one piece of software aren't you?Is that all you do compress files?I can understand it being part of what people are testing not being the sole focus though.It all ends up Ryzen ended up being a great chip at competitive prices and everyone wins even the ones that don't buy.

In that case, the outdated, poorly created program is the bottleneck. Not the CPU, as the CPU would have more to offer, but the program is too basic to understand how to use it.

That's one way of looking at it, but not the only way. Just because a process is single-threaded, that doesn't mean it is inefficient. Just because a process maxes out any number of CPU cores, that doesn't mean the CPU is slow, either. Amdahl's Law is a good example of why more cores isn't always better. Here's one reference: https://www.cs.uky.edu/~jzhang/CS621/chapter7.pdf In the case of something like most games, adding too many CPU threads can actually end up hurting performance. Keep in mind - multi-threaded CPUs are good at multitasking, but not so much at parallelization (especially when SMT/HT is involved). GPUs are good for paralellization, which is why CUDA and OpenCL were created, and, why games often do physics processing on the GPU rather than CPU. @sarsar Background processes, the filesystem (and anything done to configure it), file fragmentation, the drive controller, and the drive itself can collectively make enough of a difference to lose a MB/s or two when comparing alike systems.

That's one way of looking at it, but not the only way.

It's the only important way to look at it if we actually want to advance anything.

It's the only important way to look at it if we actually want to advance anything.

That is objectively false - again, consider Amdahl's Law. Not all tasks can be improved by adding more threads to it, that is a provable fact. If making tasks more multithreaded inherently always made them faster, Intel would've been pushing for more cores long before AMD, especially when you consider GPUs (which are better at paralellization) are stealing tens of billions of dollars from them. We're in an age where people have pissing contests over meaningless framerates, and developers know this. Depending on the CPU, scheduler, and background tasks, a game could actually lose FPS when adding more threads, because the parent process must spend time synchronizing each thread. This delay becomes far more apparent when something like HT is used, since those threads aren't run in parallel. When too many threads need to be synchronized, sometimes the parent process could be "tapping its feet" for a very long time (in a computer's perspective). That parent process could've been better off just doing the work itself. The player would likely get better minimum FPS, and the CPU could possibly use less wattage (since the load is more evenly divided and therefore the frequency can lower) but the maximum FPS would drop. To "hardcore enthusiasts", that's a turnoff. Keep in mind too, programming something (like a game) to be multi-threaded is relatively difficult and tedious. Considering adding more threads rarely actually increases maximum FPS (at least for overclocked Intels...) I can see why devs would rather take the easy way out. EDIT: For consoles, it makes sense for them to use many lower-freq cores. That allows the devs to more evenly divide the load (and therefore lower the power consumption). For console games, achieving the highest possible frame rate is totally irrelevant, so any losses in maximum FPS can be ignored. Consoles also have the advantage of having a core or two dedicated to background processes. This is good, since it prevents timing issues (in games) with child threads from synchronizing with the parent. However, devs can't depend on that for PC, which I'm guessing is why many console ports are still 4 threads or fewer.

Buy a 1700/X, wait for threadripper, or go 2nd hand X99?

Depends on how much you can get the 2nd hand X99 for.

You really can't use Newegg as a judge of AMD pricing....or pricing from any company for that matter. Newegg adjusts their prices based on competition from other etailers and product sales performance.

I have the latest 7zip, even this board's review (of the R5 1600) shows maximum 27 mb/s for the R7 1700 : http://www.guru3d.com/articles_pages/amd_ryzen_5_1600_review,11.html I usually get 26 mb/s and 1 time i saw 27 mb/s but the i5 6500 is hitting 25-26 mb/s too. I will test to see cpu utilisation but i think i've got a poor R7 or my brother's got a golden i5. 🙁

How much time do you really spend compressing/decompressing files?

Hopefully we will se re-reviews once all the BIOS and game bugs are ironed out. Ryzen is a great product with the typical AMD launch. (Sloppy is to kind of label)

Was there as much hype surrounding Zen performance as there was Bulldozer? I know there was still more than there should have been, but was it on the same level?

I have noticed some sweet sales of late. A couple of days ago I was able to get 5 (yes 5) Ryzen 5 1600 CPUs for $154 each. I was pretty damned pleased with that deal. :cheers:

Was there as much hype surrounding Zen performance as there was Bulldozer? I know there was still more than there should have been, but was it on the same level?

From what I remember, there was much greater hype from Bulldozer, but fewer people pushing it. People's expectations were much lower of Zen, but I'd say a lot more people were hoping for it to be "the end of Intel" or at least something that brings balance to the market. The ridiculous thing is despite the shakey launch, Ryzen really is right where it should've been (performance-wise) yet people were still disappointed.

Was there as much hype surrounding Zen performance as there was Bulldozer? I know there was still more than there should have been, but was it on the same level?

There was a lot of hype but it was tempered. People were cautiously optimistic.

That's not true... If all threads are below 100% utilization, then the bottleneck is elsewhere. But if any one thread is maxed out, the CPU is the bottleneck.

I'm not an expert in 7z internals but afaik it spreads the load quite evenly among cores and the compressing algorithm scales pretty well. I've never seen a compressor maxing out only 1 core when given more.

From what I remember, there was much greater hype from Bulldozer, but fewer people pushing it. People's expectations were much lower of Zen, but I'd say a lot more people were hoping for it to be "the end of Intel" or at least something that brings balance to the market. The ridiculous thing is despite the shakey launch, Ryzen really is right where it should've been (performance-wise) yet people were still disappointed.

I'm not sure if I'd say people are disappointed. There was a survey a while ago that said the Ryzen 5 was the most well-received CPU in years (something like 84% positive). Considering that Kaby Lake got a stunningly low 12%, AMD must be doing something right. Ryzen 7's score was a bit lower than Ryzen 5 (74%), I'm guessing mostly due to the negative gaming benchmarks vs the 7700K, but it was still largely positive. As for the price drop, AMD may be making room for Threadripper, or it could just be retailers jostling for sales.

Buy a 1700/X, wait for threadripper, or go 2nd hand X99?

Do all three? 😀

Do all three? 😀

I'm thinking the latter would more fit what I want at this moment in time, and be the cheapest option.

I'm not sure if I'd say people are disappointed. There was a survey a while ago that said the Ryzen 5 was the most well-received CPU in years (something like 84% positive). Considering that Kaby Lake got a stunningly low 12%, AMD must be doing something right.

People still seem to like Ryzen, but still expected it to be better. Most reviewers deemed Ryzens to be "just a good value" but very few actually called it good or recommendable. The results about Kaby Lake don't surprise me - Intel charged more money for a product with a 0% performance improvement. Despite this, I still see people looking at the i9 series like "oooh shiny!" and completely ignoring the existence of Threadripper (which as we know was pretty much the inspiration to i9). To clarify - I personally am satisfied with Ryzen. I do wish higher overclocks could be achieved but I otherwise think it is as good as it should've been. But whether it's Reddit, Youtube, or even here on Guru3D, I've seen a lot of people who opted for Intel because they found Ryzen disappointing or underwhelming.

It was to be expected considering the average performance and low overclockability.