|
 That chart is nuts. By the time you've hit the price point of a 4C/8T Intel part, AMD's already given you 4C/8T, 6C/12T, and is now showing you 8C/16T for almost the exact same price.e: also reposting for new page
|
#
?
Feb 13, 2017 02:51
|
|
|
|
| # ? May 21, 2024 13:41 |
|
|
Given the horrible shading, neck musculature, and finger deformity, pleas tell me that's a Rob Liefeld original. Evidence against: no glowing/scarred eye, fewer than 80 teeth shown.
|
|
#
?
Feb 13, 2017 03:02
|
|
|
x86-64 Platfrom Discussion - Intel: laptop CPUs on LGA2066, AMD: RIP AND TEAR YOUR THREADS
|
|
#
?
Feb 13, 2017 03:10
|
|
|
So I created a chart of all of the leaked prices that I know of: Yes, the USD prices for the last leak are correct, any number * 1.25 * 0.8 = that exact same number. I don't know what it means though, other than being amusing. Everything is collected in this imgur album.
|
|
#
?
Feb 13, 2017 03:20
|
|
|
SpelledBackwards posted:Given the horrible shading, neck musculature, and finger deformity, pleas tell me that's a Rob Liefeld original. Evidence against: no glowing/scarred eye, fewer than 80 teeth shown. Close, but no cigar. Drawn by Tom Grindberg, written by Steve Behling and Michael Stewart.
|
|
#
?
Feb 13, 2017 04:53
|
|
|
I want to see IPC vs Intel for a single core. Pronto.
|
|
#
?
Feb 13, 2017 04:57
|
|
|
This is what we have so far, not the best benchmark but it's a data point. http://wccftech.com/amd-ryzen-7-1700x-389-8-core-cpu-benchmarks-leaked/ 
|
|
#
?
Feb 13, 2017 05:14
|
|
|
That R5 1500 lookin like a pretty tasty upgrade to my FX8350
|
|
#
?
Feb 13, 2017 06:19
|
|
|
MaxxBot posted:This is what we have so far, not the best benchmark but it's a data point. Trying to wrap my head around that. Looking great for AMD pretty much. To me the biggest problem for a long time has been AMDs power usage per performance, not so much overall usable performance. This really does seem to fix that problem. I hope the platform itself is stable and compatible. In the past things like AMD's SATA and USB drivers were subpar.
|
|
#
?
Feb 13, 2017 07:56
|
|
|
redeyes posted:Trying to wrap my head around that. Looking great for AMD pretty much. To me the biggest problem for a long time has been AMDs power usage per performance, not so much overall usable performance. This really does seem to fix that problem. I hope the platform itself is stable and compatible. In the past things like AMD's SATA and USB drivers were subpar. The wild card is "how fast was the processor actually running" because nobody's sure if it's running at the 3.4 GHz base clock or more like 3.8 GHz boost clock. The motherboard doesn't appear to be turbo'ing the processor, but it should be if it were a retail board, and it's not like we have mature measurement tools either. Keeping up with a ~3.1 GHz Haswell-E/Broadwell-E is doing pretty OK if it's running at 3.4 GHz - that would put them only ~10% behind Intel in IPC. It would be less impressive at 3.8 GHz or higher, but still not bad for engineering samples/revision 1.0. There is probably some room to tweak it in later revisions just like Piledriver/Steamroller/Excavator tweaked Bulldozer. Especially if the price is right - bearing in mind that the 5820K (6C/12T) has been priced around $330 for a year or two at this point. The 8C+ Intel chips have always been a ripoff and everyone likes to focus on them when proclaiming what a great deal Ryzen will be, but the 6C12T chips are priced very close to the 4C8T chips and are very appealing for throughput-oriented stuff like video encoding. The downside is that AMD has a problem with their 4C and 6C dies and they will probably run hotter and slower than the 8C chips. Paul MaudDib fucked around with this message at 08:26 on Feb 13, 2017 |
|
#
?
Feb 13, 2017 08:21
|
|
|
X is for XFR, the automatic overclocking https://www.techpowerup.com/230609/amd-ryzen-xfr-frequencies-revealed
|
|
#
?
Feb 13, 2017 08:55
|
|
|
Anime Schoolgirl posted:x86-64 Platfrom Discussion - Intel: laptop CPUs on LGA2066, AMD: RIP AND TEAR YOUR THREADS What, you don't like multiple threads in TYOOL 2017? (One thread for the two platfroms would be good).
|
|
#
?
Feb 13, 2017 13:35
|
|
|
Measly Twerp posted:Yes, the USD prices for the last leak are correct, any number * 1.25 * 0.8 = that exact same number. I don't know what it means though, other than being amusing. It just happens that the currency conversion rate and the tax are multiplicative inverses 
|
|
#
?
Feb 13, 2017 16:00
|
|
|
HalloKitty posted:X is for XFR, the automatic overclocking I haven't had an AMD processor since the Athlon XP 2500+, what are the odds you have to mess around with delidding to get the most out of this? Have they been better than Intel when it comes to TIM/solder?
|
|
#
?
Feb 13, 2017 16:15
|
|
|
Icept posted:I haven't had an AMD processor since the Athlon XP 2500+, what are the odds you have to mess around with delidding to get the most out of this? Have they been better than Intel when it comes to TIM/solder? They've always been soldering as far as I know, I don't know about Ryzen since it's not out yet, but I can't imagine they'd change
|
|
#
?
Feb 13, 2017 16:53
|
|
|
Icept posted:I haven't had an AMD processor since the Athlon XP 2500+, what are the odds you have to mess around with delidding to get the most out of this? Have they been better than Intel when it comes to TIM/solder? It depends how big the chip is probably. e: Every question asked is also having the side effect of making me even more intensely curious for real review sites to get their hands on and bench/review/do architecture deep dives on this chip. It's hard not to be excited just for something new, even if it's not something that actually impacts our typical PC use in any real way. Gwaihir fucked around with this message at 17:22 on Feb 13, 2017 |
|
#
?
Feb 13, 2017 16:58
|
|
|
What I enjoy the most currently is how AMD is literally using the numbering convention similar to the ATI 1800/1700/1600 series, and also a bit of a throwback to the Athlon XP's but lesser so since Intel isn't selling a 1.8Ghz P4 anymore. Still, can't wait for this Platfrom to Threadrip all my task with pure MHZ!  I seriously want to build a Ryzen system just to see what AMD can do when this launches. The last AMD I built was a 1090T system that wasn't....bad. Could have used an SSD but the buyer was on a major budget and had a bad habit of catching Cryptolocker at least once a month until I put an Ad Blocker on their Router.
|
|
#
?
Feb 13, 2017 17:53
|
|
|
EdEddnEddy posted:Still, can't wait for this Platfrom to Threadrip all my task with pure MHZ! intel is the company running up the clock speed but as kaby lake seems to show that has a high rate of diminishing returns, as before
|
|
#
?
Feb 13, 2017 20:13
|
|
|
I think i must have fallen victim to rx480 overdraw. The pcie slots on my board wont work with either the new 480 or my old gtx780 now. The both light up and fanspin but nothing appears on screen and iget hardware missing beeps. Anyone know if its possible to get compensation for this? Im so miffed, had no clue this was a thing with these cards
|
|
#
?
Feb 13, 2017 20:25
|
|
|
Surprise Giraffe posted:I think i must have fallen victim to rx480 overdraw. The pcie slots on my board wont work with either the new 480 or my old gtx780 now. The both light up and fanspin but nothing appears on screen and iget hardware missing beeps. Anyone know if its possible to get compensation for this? Im so miffed, had no clue this was a thing with these cards When did you get the 480? If it's after July of last year your motherboard was never exposed to the power issue, because they fixed it in drivers at that point. edit: AMD is doing a thing on the 28th: https://www.eventbrite.ca/e/amd-capsaicin-cream-developer-sessions-tickets-31885769194 repiv fucked around with this message at 21:02 on Feb 13, 2017 |
|
#
?
Feb 13, 2017 20:29
|
|
|
repiv posted:When did you get the 480? If it's after July of last year your motherboard was never exposed to the power issue, because they fixed it in drivers at that point. I never got to the point of installing drivers. It just never worked. Maybe I borked the board when I slotted my 780 back in while the machine was off but the power still plugged in and turned on.
|
|
#
?
Feb 13, 2017 21:05
|
|
|
Surprise Giraffe posted:I think i must have fallen victim to rx480 overdraw. The pcie slots on my board wont work with either the new 480 or my old gtx780 now. The both light up and fanspin but nothing appears on screen and iget hardware missing beeps. Anyone know if its possible to get compensation for this? Im so miffed, had no clue this was a thing with these cards I'm pretty sure there's no way it was drawing too much in BIOS unless the card was actually broken and faulty somehow. The power draw issue would have only been present under load.
|
|
#
?
Feb 13, 2017 21:11
|
|
|
What month should Ryzen/Vega come out? I've been looking all over for dates but there doesnt seem to be a general consensus other than "2017".
|
|
#
?
Feb 13, 2017 21:21
|
|
|
March is the expected full reveal month, probably GDC which I think is the 3rd.
|
|
#
?
Feb 13, 2017 21:24
|
|
|
Last rumor I heard was Ryzen motherboards on February 24, CPUs on February 28. Edit: but I think you are talking about APUs so nevermind ignore me
|
|
#
?
Feb 13, 2017 21:49
|
|
|
3DMark Physics Scores? https://twitter.com/OC3D/status/831287903741411329
|
|
#
?
Feb 14, 2017 01:03
|
|
|
Phone posting or else I'd do it myself, but could someone make a plot of those scores vs the price?
|
|
#
?
Feb 14, 2017 01:20
|
|
|
https://twitter.com/CPCHardware/status/830851176853823488 this may not seem to make sense if you never realized that AMD also historically frequency binned their chips, SKUs in the PII era had progressively lower clocks all the way down the stack. Anime Schoolgirl fucked around with this message at 02:49 on Feb 14, 2017 |
|
#
?
Feb 14, 2017 02:46
|
|
|
Anime Schoolgirl posted:https://twitter.com/CPCHardware/status/830851176853823488 Isn't AMD's 4C/4T competing with i3's and not intels i5's though.
|
|
#
?
Feb 14, 2017 03:17
|
|
|
Intel will still beat them on the high end for single threaded. I think this will be a wash in the end honestly. BUT price will be driven down. I will still buy Intel because AMD blows at Q/A.
|
|
#
?
Feb 14, 2017 03:19
|
|
|
yep lucky that the i3-7350k is priced the same as an i5 lmao redeyes posted:Intel will still beat them on the high end for single threaded. I think this will be a wash in the end honestly. BUT price will be driven down. I will still buy Intel because AMD blows at Q/A.
|
|
#
?
Feb 14, 2017 03:21
|
|
|
...so much for the AVX thing.
|
|
#
?
Feb 14, 2017 03:44
|
|
|
SwissArmyDruid posted:...so much for the AVX thing. Firestrike doesn't use AVX so you wouldn't expect to see any benefit from Intel's beefier SIMD units there. That's kind of the point though - it's hard to care about Intel's AVX advantage when next to nothing actually utilizes those instructions.
|
|
#
?
Feb 14, 2017 04:13
|
|
|
I just want to see leaks on a nice mATX board for Ryzen.
|
|
#
?
Feb 14, 2017 05:31
|
|
|
  I'm not really all that impressed with these tbh. Essentially the AMD octocore is only slightly outperforming the Intel hexacore at equivalent clocks - and the hexacores have always been the dirt-cheap part of the Intel HEDT lineup. It's going to be tough to sell 8-core Ryzen at $550 a pop when you can walk into Microcenter and pick up a 5820K for $330 literally today. I guess it all mostly comes down to clocks. If Ryzen is out-clocking Haswell-E (i.e. 5 GHz) then it will be viable, at equal clocks (4.5 GHz) it's meh, if the clocks don't go past 4 GHz it's a stinker. Same for the lower-tier chips, which are mostly let down by their poo poo-tier clocks as I see it. If those 6C and 4C chips can get the clocks up to 4 GHz they will be decent. But at 3.4 and 3.2 GHz they are going to be laughably outperformed by an unlocked i5.
|
|
#
?
Feb 14, 2017 06:24
|
|
|
So much for the Windows 7 rumor: http://www.pcworld.com/article/3167279/computers/amd-sorry-there-will-be-no-official-ryzen-drivers-for-windows-7.html
|
|
#
?
Feb 14, 2017 06:26
|
|
|
IIRC, AMD said they validated on W7 and W10, but never announced commercial support.
|
|
#
?
Feb 14, 2017 06:33
|
|
Paul MaudDib posted:
Isn't the 8C/16T Ryzen CPU starting at $389? And the 6C/12T top end version at like $260? If you can take the top end 6C/12T Ryzen up to 4GHz then it looks like it will be a good bit faster than a top end i5 for about the same price. Anyway, here's hoping AMD does well with this, Intel could use the competition.
|
|
|
#
?
Feb 14, 2017 06:36
|
|
|
AVeryLargeRadish posted:Isn't the 8C/16T Ryzen CPU starting at $389? $319, according to the averages of the fairly consistent leaks.
|
|
#
?
Feb 14, 2017 06:44
|
|
|
|
| # ? May 21, 2024 13:41 |
|
|
How is the quad-core at 3.2ghz doing better than the octa-core at 4ghz per core 
|
|
#
?
Feb 14, 2017 07:00
|
|