• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

AMD Ryzen Thread: Affordable Core Act

The lede totally got buried in the official denial that there's any problem with the Win10's scheduler.

AMD said:
The primary temperature reporting sensor of the AMD Ryzen™ processor is a sensor called ”T Control," or tCTL for short. The tCTL sensor is derived from the junction (Tj) temperature—the interface point between the die and heatspreader—but it may be offset on certain CPU models so that all models on the AM4 Platform have the same maximum tCTL value. This approach ensures that all AMD Ryzen™ processors have a consistent fan policy.

Specifically, the AMD Ryzen™ 7 1700X and 1800X carry a +20°C offset between the tCTL° (reported) temperature and the actual Tj° temperature. In the short term, users of the AMD Ryzen™ 1700X and 1800X can simply subtract 20°C to determine the true junction temperature of their processor. No arithmetic is required for the Ryzen 7 1700. Long term, we expect temperature monitoring software to better understand our tCTL offsets to report the junction temperature automatically.

The 1700X and 1800X are deliberately reporting temperatures 20 degrees higher than their true temperatures.

Well, at least we know what's up with Shark Sandwich's temps a couple of pages ago.
 
Funny that you mention that, since Cluster-on-a-Die techniques are a way to make NUMA-aware sw and schedulers perform better on those dual-ring cpus via partitioning of the dual-ring bus into two on-die NUMA nodes.

We need your sage knowledge in threads like these, to keep the negative brigade at bay. God I can't do it all on my tod. Got overwhelmed the other week.
 
Computerbase ran their game benchmark set on Windows 7, Windows 10, Windows 10 with HPET (high precision event timer) disabled (as per AMD guidelines), and Windows 10 in "High Performance" energy mode ("Höchstleistung"):
ryzen_wincwuex.png

https://www.computerbase.de/2017-03/ryzen-windows-7-benchmark-core-parking/

It seems like overall there is no large difference, though for a few individual benchmarks Win 7 and Win 10 in "high performance" mode show significantly improved results. Note that those two always track together, so it seems like the differences between Win 7 and 10 are really entirely attributable to more aggressive energy optimization and not scheduling changes. (And you can get the same or better performance as Win 7 in Win 10 by enabling the "high performance" profile)

I'm not familiar with this benchmark, but the important thing to note about power management is that it might not actually show up in a traditional benchmark and in all use-cases.

Where I have historically seen the biggest gains with power management for Intel processors is with applications that quickly spike the processor up to a high CPU state, and then fall back to idle. It makes a bigger difference to "responsiveness", as the system is able to respond sooner due to the quick ramp up in frequency, rather than, say, max FPS.
 

blu

Wants the largest console games publisher to avoid Nintendo's platforms.
We need your sage knowledge in threads like these, to keep the negative brigade at bay. God I can't do it all on my tod. Got overwhelmed the other week.
There's nothing sage in my knowledge - it's common knowledge in the HPC world. We're just on a gaming forum. Some markets outside of gamers have been weeping for a product like Ryzen, as Intel have been practising extortion in those markets for many years now. In contrast, gamers are fine with an overclocked 4c. And since gaming is just a hobby for me, and my occupation lies elsewhere, I have no issue recognizing the advantages of something like Ryzen.
 

Durante

Member
My job is in HPC, and it seems like a better fit for a datacenter chip to me (so the parameter study type workload rather than the communication-heavy one which might benefit from AVX throughput). Of course, this will also be a function of price, there's a lot of room there considering the absurd pricing on higher-end Xeons.

Anyway, scientifically it's rather interesting, since if you take the core cluster results on the 8-core chip and extrapolate that upward to a 2 socket system with 32 each you might get something like 3-4 levels of significantly distinct communication latency even on a single shared memory node with "just" two sockets. "Interesting" HW is rarely what domain scientists want in their HPC nodes though, in my experience.
 

Akronis

Member
My job is in HPC, and it seems like a better fit for a datacenter chip to me (so the parameter study type workload rather than the communication-heavy one which might benefit from AVX throughput). Of course, this will also be a function of price, there's a lot of room there considering the absurd pricing on higher-end Xeons.

Anyway, scientifically it's rather interesting, since if you take the core cluster results on the 8-core chip and extrapolate that upward to a 2 socket system with 32 each you might get something like 3-4 levels of significantly distinct communication latency even on a single shared memory node with "just" two sockets. "Interesting" HW is rarely what domain scientists want in their HPC nodes though, in my experience.

Does AMD have anything equiv to Xeon at this point? I know in the past, their Opteron line was the main competitor but I don't think AMD has made any newer ones.

And have AMD or any manufacturers revealed any multiple CPU socket motherboard? We're an Intel shop for our HPC stuff now but it'd be pretty neat to see AMD come back into the server market.
 
Übermatik;232039537 said:
AMD: Tips for Building a Better AMD Ryzen™ System

https://community.amd.com/community...ilding-a-better-amd-ryzen-system?sf62303579=1

Not sure if there's anything new in this, but nice to have a guide.
Thanks, I meant to post this earlier myself along with the 20c temp. update. Added to OP.

All current and prospective owners should read the above link.


Headings:
● Update Your Firmware
● Memory Matters
● Mind Your Power Plan
● The Observer Effect
● Overclocking!
● PUTTING IT ALL TOGETHER



Highlights:

● Memory Matters

AMD Ryzen™ processors have an appetite for faster system RAM, but it’s important to ensure that you have a solid setup before proceeding.


  1. The AMD Ryzen™ processor does not offer memory dividers for DDR4-3000 or DDR4-3400. Users shooting for higher memory clocks should aim for 3200 or 3500 MT/s.

  2. Memory vendors have also begun to validate 32GB (4x8GB) kits at 3200 MT/s rates for select motherboards.

  3. Ensure that you are programming your BIOS with the recommended timings (CAS/tRCD/tRP/tRAS/tRC/CMD) and voltages specified on the DRAM packaging.

  4. To ensure reliable POST, the AMD Ryzen™ processor may fall back to a DIMM’s JEDEC SPD “safe” timings in the event an overclock proves unreliable. Most DIMMs are programmed to boot at DDR4-2133 unless otherwise instructed by the BIOS, so be sure your desired overclock is in place before performance testing. Use CPU-Z in Windows to confirm.

  5. For speed grades greater than DDR4-2667, please refer to a motherboard vendor’s memory QVL list. Each motherboard vendor tests specific speeds, modules, and capacities for their motherboards, and can help you find a memory pairing that works well. It is important you stick to this list for the best and most reliable results.

  6. We have internally observed good results from 2933, 3200, and 3500 MT/s rates with 16GB kits based on Samsung “B-die” memory chips. Potential kits include:
    • Geil EVO X - GEX416GB3200C16DC [16-16-16-36 @ 1.35v]
    • G.Skill Trident Z - F4-3200C16D-16GTZR [16-18-18-36 @ 1.35v]
    • Corsair CMK16GX4M2B3200C16 VERSION 5.39 [16-18-18-36 @ 1.35v]
  7. Finally, as part of AMDs ongoing development of the new AM4 platform, AMD will increase support for overclocked memory configurations with higher memory multipliers. We intend to issue updates to motherboard partners in May that will enable them, on whatever products they choose, to support speeds higher than the current DDR4-3200 limit without refclk adjustments. AMD Ryzen™ processors already deliver great performance in prosumer, workstation, and gaming workloads, and this update will permit even more value and performance for enthusiasts who chose to run overclocked memory.

  8. AMD’s officially-supported DRAM configurations are below for your reference:
ryzen-7-official-drammejao.png




● Mind Your Power Plan

Make sure the Windows® 10 High Performance power plan is being used (picture). The High Performance plan offers two key benefits:


  1. Core Parking OFF: Idle CPU cores are instantaneously available for thread scheduling. In contrast, the Balanced plan aggressively places idle CPU cores into low power states. This can cause additional latency when un-parking cores to accommodate varying loads.

  2. Fast frequency change: The AMD Ryzen™ processor can alter its voltage and frequency states in the 1ms intervals natively supported by the “Zen” architecture. In contrast, the Balanced plan may take longer for voltage and frequency changes due to software participation in power state changes.

In the near term, we recommend that games and other high-performance applications are complemented by the High Performance plan. By the first week of April, AMD intends to provide an update for AMD Ryzen™ processors that optimizes the power policy parameters of the Balanced plan to favor performance more consistent with the typical usage models of a desktop PC.



● PUTTING IT ALL TOGETHER

To test the performance impact of all of these various changes, we threw together a brand new Windows 10-based system with the following specifications:


  • AMD Ryzen™ 7 1800X (8C16T/3.6-4.0GHz)

  • 16GB G.Skill (2x8) DDR4-3200
    • Clocked to 2133MT/s: 15-15-15-35-1t
    • Clocked to 2933MT/s: 14-14-14-30-1t
  • ASUS Crosshair VI Hero (5704 BIOS)

  • 1x AMD Radeon™ RX 480 GPU (Radeon Software 17.2.1)

  • Windows 10 Anniversary Update (Build 14393.10)


Throughout this process we also discovered that F1™ 2016 generates a CPU topology map (hardware_settings_config.xml) when the game is installed. This file tells the game how many cores and threads the system’s processor supports. This settings file is stored in the Steam™ Cloud and appears to get resynced on any PC that installs F1™ 2016 from the same Steam account. Therefore: if a user had a 4-core processor without SMT, then reused that same game install on a new AMD Ryzen™ PC, the game would re-sync with the cloud and believe the new system is also the same old quad core CPU.


Only a fresh install of the game allowed for a new topology map that better interpreted the architecture of our AMD Ryzen™ processor. Score one for clean computing! But it wasn’t a complete victory. We also discovered that the new and better topology map still viewed Ryzen™ as a 16-core processor, rather than an 8-core processor with 16 threads. Even so, performance was noticeably improved with the updated topology map, and performance went up from there as we threw additional changes into the system.


As an ultimate maneuver, we asked the question: “Can we edit this file?” The answer is yes! As a final step, we configured F1™ 2016 to use 8 physical CPU cores, rather than the 16 it was detecting by default. Performance went up again! After all was said and done, we gained a whopping 35.53% from our baseline configuration showing how a series of little changes can add up to something big.


The picture tells the story clear as day: configuration matters.
pastedimage_92eckae.png
 

Durante

Member
Does AMD have anything equiv to Xeon at this point? I know in the past, their Opteron line was the main competitor but I don't think AMD has made any newer ones.

And have AMD or any manufacturers revealed any multiple CPU socket motherboard? We're an Intel shop for our HPC stuff now but it'd be pretty neat to see AMD come back into the server market.
No, they don't have anything competitive right now.

They will once they get Naples out, which will scale up to 64 cores per shared memory node with two sockets.
 

shark sandwich

tenuously links anime, pedophile and incels
The lede totally got buried in the official denial that there's any problem with the Win10's scheduler.



The 1700X and 1800X are deliberately reporting temperatures 20 degrees higher than their true temperatures.

Well, at least we know what's up with Shark Sandwich's temps a couple of pages ago.
Haha man what a relief.

Still very aggravated about my memory (which I specifically chose from Asus' QVL) not running at its proper speed. Contacted Asus tech support Friday night and have not received a response yet.

Oh and the BIOS and Asus AI software both report differing XMP profiles and they're both wrong. CPU-z reports the correct #s, which are supposedly the ones they were validated at on this motherboard.
 
Does AMD have anything equiv to Xeon at this point? I know in the past, their Opteron line was the main competitor but I don't think AMD has made any newer ones.

And have AMD or any manufacturers revealed any multiple CPU socket motherboard? We're an Intel shop for our HPC stuff now but it'd be pretty neat to see AMD come back into the server market.
They had largely vanished from this space. You can heavily discount less-efficient and under-performing Piledriver consumer desktop parts, but that doesn't translate elsewhere.

For one, the upper end Opterons didn't get updated to the Piledriver+, Steamroller or Excavator core. They were faster and more efficient, but not sufficient enough to overcome AMD's lost market-share, or particularly feasible given their incredibly limited financial resources. They had to make a sacrifice in not being able to invest in several GPU and CPU products to put out fires elsewhere, while preparing for Zen.


The OP will have Server/HPC updates here: http://www.neogaf.com/forum/showpost.php?p=231164109&postcount=2

I hadn't yet included the most recent public showing, though it had been posted earlier in this thread.

Forrest_Naples_Embargoed%20Until%203_7_17-page-010_678x452.jpg


AMD Naples Performance First Demo 2/21 - AnandTech.com
https://www.youtube.com/watch?v=15TcolqloKI


AMD Naples Benchmarks vs Intel Xeon E5-2699 V4
https://www.youtube.com/watch?v=ZtYTPcshyA0


AMD Prepares 32-Core Naples CPUs for 1P and 2P Servers: Coming in Q2
http://www.anandtech.com/show/11183...aples-cpus-for-1p-and-2p-servers-coming-in-q2


AMD Ryzen Is Strong On The Desktop But Naples Could Really Put A Chink In Intel's Server Armor
https://www.forbes.com/sites/daveal...ld-really-put-a-chink-in-intels-server-armor/


AMD Naples server processor: More cores, bandwidth, memory than Intel
https://arstechnica.com/information...essor-more-cores-bandwidth-memory-than-intel/
 
Haha man what a relief.
Sorry, I thought I had shared my own temperature findings when you initially posted, but looks like I never hit submit...


The lede totally got buried in the official denial that there's any problem with the Win10's scheduler.

The 1700X and 1800X are deliberately reporting temperatures 20 degrees higher than their true temperatures.

Well, at least we know what's up with Shark Sandwich's temps a couple of pages ago.
So it was not the Asus BIOS getting the temps wrong lol
Except, this is an issue that can be addressed in the BIOS.

I speak from first hand experience.

My Gigabyte Aorus X370 Gaming 5 shipped with F3 BIOS which is very close to what most reviewers used, and why so many commented on "high CPU temps" unless they had been more observant.

Both that and the F4 BIOS have the same 20c temperature discrepancy, causing the fans to spins far faster than they should and seeming to alter the the turbo/XFR states (note, I don't have extensive figures of my own on this).

After launch Gigabyte released the F5c beta BIOS, which finally resolved the issue. Before AMD confirmed this or reading Shark Sandwich's experience here, I was able to see this for myself.

Temps are reporting 20c lower and the fans aren't spinning unnecessarily from the inaccurate readings. The system is dead silent.


The other board manufacturers should be able to address this as well. My AX370's UEFI is laughably spartan, but at least Gigabyte's update got this aspect right.


Any other owner can confirm what I've said about temps. Since updating, I had downgraded to both F3 and F4 again for testing, and it went back to 20c higher temps and higher fan speeds. Currently on F5c/F5d awaiting the next proper update from AMD to all board makers and anything Gigabyte has before then.


ーーー

Here's chew*'s post on the review BIOS for the Gaming 5:

http://www.xtremesystems.org/forums...-of-the-Jedi&p=5254923&viewfull=1#post5254923

chew* @ XtremeSystems said:
F3f is old has base AMD agesa .
Despite date f3 is newer, f3n is newer.

F5c is newest with a change that did not work so can possibly cause an issue F5b is probably the safest and newest.

F3f was reviewers bios fyi.

I ditched it for publically available F3 for review.


F3n boots easier with certain modules

Bios after base agesa will have better effeciency
 
Excuse the consecutive posts, but none of these are remotely related.


Biostar X370-GTN Mini-ITX
http://www.neogaf.com/forum/showpost.php?p=231452345&postcount=1001


Gigabyte Japan showed a slide of their B350 mITX board which is "coming soon."

Gigabyte AB350N-Gaming WiFi Mini-ITX
http://www.4gamer.net/games/259/G025981/20170314076/

Google Translate said:
Although there was no actual machine, it was introduced only by a slide, but the B350 equipped motherboard "Mini-ITX specification" GA-AB350N-Gaming WiFi "is also preparing for domestic release. For those who want to make small PC with Ryzen, it may be a long-awaited product

http://pc.watch.impress.co.jp/docs/news/1049064.html

Google Translate said:
In addition, he announced that he will also introduce the Mini-ITX motherboard "AB350N-Gaming Wi-Fi" compatible with Ryzen soon. At present, the strongest motherboard reigning in the Mini-ITX world is ASRock's "X99E-ITX / ac", but the original potential possessed by the LGA2011-v3 processor such as two memory slots and one PCI Express x16 slot It is a fact that you must struggle to cool down to a high TDP.

On that point, Ryzen 7 is a sufficient suspension with Mini-ITX, and Ryzen 7 1700 can realize a high-spec environment of 8 cores and 16 threads with TDP 65W. For that reason, AB350N-Gaming Wi-Fi will be one of those who was looking forward to the user who wishes to set up with the Mini-ITX form factor. I would like to expect early market launch.
 

turmoil

Banned
The 4c/8t parts not hitting 4Ghz turbo is market segmentation or a technical incapability of the architecture?

Given all the talk about the hardships of getting high clocks on parts with many cores I thought that it would be easier to achieve in the 4c/8t R5s.
 

Renekton

Member
The 4c/8t parts not hitting 4Ghz turbo is market segmentation or a technical incapability of the architecture?

Given all the talk about the hardships of getting high clocks on parts with many cores I thought that it would be easier to achieve in the 4c/8t R5s.
Seems to be Samsung's process limitation.
 

Locuza

Member
The 4c/8t parts not hitting 4Ghz turbo is market segmentation or a technical incapability of the architecture?

Given all the talk about the hardships of getting high clocks on parts with many cores I thought that it would be easier to achieve in the 4c/8t R5s.
Well the architecture can hit ~ 4 Ghz as seen with Ryzen 7.
The reason is probably the binning, where AMD uses worse chips for R5.


Silicon Lottery made a rough statistic for R7 SKU's (sample size per SKU 30, 90 in total for the comparison):

Ryzen 7 1700
93% reach 3.8GHz @ 1.376V
70% reach 3.9GHz @ 1.408V
20% reach 4.0GHz @ 1.440V

Ryzen 7 1700X
100% reach 3.8GHz @ 1.360V
77% reach 3.9GHz @ 1.392V
33% reach 4.0GHz @ 1.424V

Ryzen 7 1800X
100% reach 3.8GHz (assumed)
97% reach 3.9GHz @ 1.376V
67% reach 4.0GHz @ 1.408V
20% reach 4.1GHz @ 1.440V

https://www.reddit.com/r/Amd/comments/5xybp7/silicon_lottery_ryzen_overclock_statistics/
 
The same test with 5960x

ping-intel.png



Ryzen as lower physical core-to-core latency which may explain why sp to mp ratio beats intel in some benchs, maybe.

So Ryzen has half the latency when only using cores in the same CCX compared to the 5960x, but higher latency when crossing CCX?

Does this mean with appropiate thread managing and mostly avoiding crossing CCX, the latency performance will be a lot better or am I extrapolating too much?

Looking at those 4-core only and SMT off tests, I'm pretty sure I'll get a Ryzen build in a few months. I'll just change core affinities for the games that work better with less threads or something, if I even need to do that (I have both 60 and 144 Hz screens, but don't need 144 for everything).
 

LordCiego

Member
Nice, only one month away and some of the ryzen launch problems will be solved when the R5 are released. I think that for gaming workloads the 1600X and 1600 are not going to have very different results than R7 chips with the same clocks.
 
There's nothing sage in my knowledge - it's common knowledge in the HPC world. We're just on a gaming forum. Some markets outside of gamers have been weeping for a product like Ryzen, as Intel have been practising extortion in those markets for many years now. In contrast, gamers are fine with an overclocked 4c. And since gaming is just a hobby for me, and my occupation lies elsewhere, I have no issue recognizing the advantages of something like Ryzen.

I hardly game these days becauses I am so busy with work and use my computer for it, so productivity takes precedence over gaming for me.

That said Intel have also been practising extortion in the gaming enthusiast market with 4-cores since Sandy Bridge up until Kaby Lake (with incremental perf increases along the way) occupying the top spot. Only now with AMD bringing affordable 8-cores Intel are moving their high-end 6-core HEDT down the product stack with Kaby Lake X.

The pure negativity in this thread doesn't match up with the reality of what a great set of CPUs these Ryzen R7s are. They should not be compared with a 7700K but unfortunately that is how the narrative has been set now. Ryzen R5's should be and although they won't match a 7700K, one of the R5's will be close to half the price.

Nice, only one month away and some of the ryzen launch problems will be solved when the R5 are released. I think that for gaming workloads the 1600X and 1600 are not going to have very different results than R7 chips with the same clocks.

Yes but the prices will be game-changers.
 

Paragon

Member
Nice, only one month away and some of the ryzen launch problems will be solved when the R5 are released. I think that for gaming workloads the 1600X and 1600 are not going to have very different results than R7 chips with the same clocks.
I wouldn't be surprised if the 1600/X have the least consistent gaming performance of the entire line-up with two 3-core CCX units, since gaming seems most affected by the dual CCX structure of the R7 CPUs.
One 4-core CCX (R5 1400/1500X) or two 4-core CCX (R7 1700/X) seem like they would be the better choice for gaming.
 

Datschge

Member
·feist·;232094117 said:
Guru3D posted early and deleted that link because the official AMD announcement is later. "Cached" here.
Interesting to see that the TDP of the 4c8t Ryzen models are still 65W. That points to them using two binned CCX as well, and models with only one CCX may well only appear once the APU models arrive (the announced Ryzen Mobile laptop line and some desktop APUs).
 

blu

Wants the largest console games publisher to avoid Nintendo's platforms.
My job is in HPC, and it seems like a better fit for a datacenter chip to me (so the parameter study type workload rather than the communication-heavy one which might benefit from AVX throughput). Of course, this will also be a function of price, there's a lot of room there considering the absurd pricing on higher-end Xeons.
Pricing and density, both, IMO, will be a drawing factor for Ryzen. We will see how it will fare in the HPC world (which I was referring to in the context of NUMA knowledge, not that necessarily Ryzen will storm that market). That said, it seems AMD themselves will be positioning Ryzen more towards the datacenter marker, so there are multiple factors at play here.

Anyway, scientifically it's rather interesting, since if you take the core cluster results on the 8-core chip and extrapolate that upward to a 2 socket system with 32 each you might get something like 3-4 levels of significantly distinct communication latency even on a single shared memory node with "just" two sockets. "Interesting" HW is rarely what domain scientists want in their HPC nodes though, in my experience.
I don't think we need to straight extrapolate that way. For one, IF topologies will likely vary much from the desktop variant.
 

ethomaz

Banned
With 65W TDP I guess they will use 8C failures to 6C and 4C.

For 6C I understand because you really need 2 CCX but for 4C is weird.

Let's see after launch.
 
The 1600 sounds like the CPU to get. Will probably overclock to 3.8-4 Ghz, has 6 cores + SMT. In games, it'll probably be very, very close to the 1800X@stock.
I dunno, the price gap between the 1600 and 1600X is small enough that paying the premium for better binning isn't a bad idea. With the R7s, the price gap between the 1700 and the 1800X is big enough that you're better off taking your chances with the silicon lottery than buying the highest binned chip, outside of scenarios where you need 16 threads at ~4Ghz.
 
The pure negativity in this thread doesn't match up with the reality of what a great set of CPUs these Ryzen R7s are. They should not be compared with a 7700K but unfortunately that is how the narrative has been set now. Ryzen R5's should be and although they won't match a 7700K, one of the R5's will be close to half the price.

Yeah it really is shocking that gamers forum judges cpu by benchmarks in games.

Being great for rendering/workstation/server workloads doesn't mean anything to majority of PC gamers because they won't use anything more advanced than office suite on their pc so that leaves gaming as the only cpu demanding task.

If we are talking about narritives it's sad to see that some enthusiasts try to sabotage years of effort to make people understand that cpu matters for their partisan goals.
 

nubbe

Member
I imagine R5 & R3 could suffer more from the CCX latencies if they only disable broken cores on both clusters
You could end up with 1+3 or 3+3 if they do it like that
 

strata8

Member
I imagine R5 & R3 could suffer more from the CCX latencies if they only disable broken cores on both clusters
You could end up with 1+3 or 3+3 if they do it like that
That wouldn't make any sense, there'd be massive performance variation between chips in that case. 3+3 is likely for the 1600 but there's no way we'll see a mix of 1+3, 2+2, 4+0 for the 4 core parts, at least within the same model number.
 

shark sandwich

tenuously links anime, pedophile and incels
1600 + 1600X should be pretty great. Gaming performance will be nearly indistinguishable from 1800X. And at that price it's competing against i5 7600k, so we could see a pretty big gap for games/workloads that make good use of >4 threads. Plus AM4 motherboards are cheaper and will likely support the next 1-2 CPU generations.

Still not an Intel killer, but it will be a strong contender for a lot of gamers.
 

nubbe

Member
That wouldn't make any sense, there'd be massive performance variation between chips in that case. 3+3 is likely for the 1600 but there's no way we'll see a mix of 1+3, 2+2, 4+0 for the 4 core parts, at least within the same model number.

You wouldn't get massive variation in production software
But games would suffer
 
I imagine R5 & R3 could suffer more from the CCX latencies if they only disable broken cores on both clusters
You could end up with 1+3 or 3+3 if they do it like that

It seems likely the 1600's will be 3+3 and if that's true, they may end up being the worst gaming chip since they'll have the fewest threads per CCX to deal with applications that don't like crossing CCX's like games. The 1500's are the wild card here, a 4+0 should consistently provide performance between and i5 and an i7 across applications and this is the rumored configuration for those chips. If the chips actually end up being 2+2, performance can range anywhere between a Pentium to an i7 depending on the app, for the same amount of cores/threads.
 

shandy706

Member
BPKVG6N.png


4 cores / 8 threads at $169

DAYUM

Oh yes....I'm itching so bad to build myself a new gaming system. It's "killing" me.

I've had parts in my cart multiple times this week and haven't checked out. (i5s and i7s). I just need to hold out a few more weeks. Hopefully the R5s will perform well and the i5/i7s will drop in price.

I'm willing to buy AMD or Intel. I'm going to be all about price vs performance with this build. I have my case, my PSU, and my GPU. I just need a Motherboard, RAM, the CPU and few small things.

Exciting.
 
Yeah should thoroughly obsolete the 7700K...

No. Posts like this aren't doing anyone any favours. The 7700K still clocks higher, the 7700K still has higher IPC and the 7700K will remain the go to option for a high end gaming focused machine.

The R5s are squarely targeted at the i5 range and they have a lot to offer in that segment. Including SMT and overclocking when compared to the locked i5 products is a tremendous value add for that end of the market.
 

Papacheeks

Banned
Oh yes....I'm itching so bad to build myself a new gaming system. It's "killing" me.

I've had parts in my cart multiple times this week and haven't checked out. (i5s and i7s). I just need to hold out a few more weeks. Hopefully the R5s will perform well and the i5/i7s will drop in price.

I'm willing to buy AMD or Intel. I'm going to be all about price vs performance with this build. I have my case, my PSU, and my GPU. I just need a Motherboard, RAM, the CPU and few small things.

Exciting.

I'm in the same boat lol. i'v got 2 builds on PC part picker and two wishlists created on newegg just waiting to push that Add to cart button.
 

Paragon

Member
http://www.anandtech.com/show/11202/amd-announces-ryzen-5-april-11th
Anandtech has a good write-up and confirms 3+3 for the hexas and 2+2 for the quads. Benches will be interesting. I wonder if it'll be possible to unlock some extra cores.
Oh no.
I mean, I guess the 1500X gets double the cache per core - Anandtech still report 16MB total - and they're say it has 200MHz XFR (from 3.7GHz) but 2 cores per CCX seems like it's going to have the biggest penalty when gaming.
I'd say the R7 1700 is the CPU to get if you're building a Ryzen system for gaming, based on this information.

And I'm still considering it; it's just no longer an instant-buy for me like I was hoping. I'm waiting to see what Intel do to try and counter this, as AMD's pricing is very aggressive.
And I'm not even sure that Intel can counter this. They can compete on performance but I don't know about price.
 
Yeah should thoroughly obsolete the 7700K...
No. Posts like this aren't doing anyone any favours. The 7700K still clocks higher, the 7700K still has higher IPC and the 7700K will remain the go to option for a high end gaming focused machine.

The R5s are squarely targeted at the i5 range and they have a lot to offer in that segment. Including SMT and overclocking when compared to the locked i5 products is a tremendous value add for that end of the market.
I agree, even if it comes across as fun police.
 
Top Bottom