Grats
Good on them; hopefully they're back in the game. Intel needs a nice kick in the nards for running roughshod over the market for the past 5+ years.
AMD wasn't able to turn a profit this quarter, but analysts are bullish on the chipmaker's solid Ryzen CPU sales. The world's other x86 chipmaker topped estimates with a 21 per cent jump in revenues, but still couldn't quite manage to get into the black. For the FY2017 Q2 period, ended July 1: Revenues of $1.22bn were up 19 …
I can't imagine Intel's share price going anywhere but down from here.
Mmmm. We all said that when the AMD64 destroyed the P4 in basically every conceivable metric, and then AMD released a true duel core chip than ran relatively cool while the P4's were getting so hot through increasing clock speeds that they could be used for heating the building.
Then Intel bribed OEM's to exclude AMD from the market, eventually caught up with AMD's advanced designs after a few years and basically wiped them out, allowing them only to survive as a niche market to maintain that Intel isin't a monopoly.
Never underestimate Intel. This time around, there is again a strange lack of cheap AMD machines from companies that were named in the anti trust thing last time around, which i'm sure is a total coincidence.
"Good on them; hopefully they're back in the game. Intel needs a nice kick in the nards for running roughshod over the market for the past 5+ years".
^ Absolutely this. Intel needs a serious competitor to keep it from lazy, monopolistic bad practice and unjustified high price rises. Everyone wins that way and I hope that AMD moves into profit by summer next year.
I bet you wish you'd spent the extra on the DX. But hey, who needs to do floating point arithmetic
Intel at first just had plain old 486 CPUs, no SX, DX. But then they had a problem in manufacturing, and some of the CPUs had problems with floating point. Actually, a lot of CPUs had problems with floating point. If Intel just scrapped the lot of them, they'd lose a lot of money... Hmm. Solution: sell 'em as 486SX CPUs, and the ones which actually worked properly as 486DXs. And put two sockets on the motherboards. Sell a 487 'math chip' which was really a 486DX with an extra pin; when you plugged the 487 into the motherboard it turned off the 486SX, so that the customer paid for two CPUs and got one which actually worked.
Marketing. I love it.
And if anyone thinks that AMD wouldn't pull that kind of crap if they thought they could get away with it, have a look at exactly why AMD sold triple-core CPUs. Hint: http://gizmodo.com/373185/amd-phenom-x3-triple-core-processors-are-crippled-quad-cores-in-disguise
Shocking, especially seeing that Vega is set to be a disappointment and Ryzen seems to have unresolved silicon issues!
http://www.guru3d.com/news-story/possible-radeon-rx-vega-3dmark-time-spy-benchmark-result.html
https://community.amd.com/thread/215773
Definitely not something this commentard would voluntarily buy!
The mystery GPU may not be VEGA and may not even be AMD at all (same device ID appeared to better 1080 benches back in December - only card that did that in recent times is the 1080 Ti)
The segfault problem in Linux builds seems to improve with SMT (Simultaneous Multi Threading - AMDs "hyper threading") turned off. So how you got silicon issue from that I don't know (could be silicon but more likely a problem with the SMT implementation during compiles) . Whilst this is a major issue for some it isn't as if Intel are free from their own issues (google i7 6700 over heating or try https://www.theregister.co.uk/2017/05/01/intel_amt_me_vulnerability/)
[disclaimer} My main machine is intel with nvidia (3770 + 970) so no I am not an AMD fanboi
...or even the SMT problems we talked about with Skylake last month: https://www.theregister.co.uk/2017/06/25/intel_skylake_kaby_lake_hyperthreading/
Chips have bugs, film at 11. Might be worth seeing if AMD can patch this bug in microcode (assuming it is their fault). You'll be a long time waiting for a faultless chip of this complexity though.
Sorry, the link about Vega was the best one I could be bothered locating on mobile, but there are plenty of other places saying similar things about it.
My concern WRT Ryzen is that it's based on the same tech as the upcoming Epyc server chips, where multithreaded workloads are the norm. Telling your customers to cripple their hardware to get reliability out of it is not going to go down terribly well in the land of Enterprise.
Over the years I've looked at AMD based machines many times. Not being an engineer I just don't have the knowledge, skills or, tbh the inclination to compare the cost/power of each type. It's difficult to know what I'm buying just within the one ( Intel) range. Let alone working out whether a given AMD alternative is better/poorer value for money. And the IT magazines don't really help. I've seldom, if ever, seen these things tabulated so that we can compare in the same way they would with different ranges of printers or routers (etc.). The manufacturers, both, come out with chip names and range codes that actually say very little,let alone about how well one chip performs against another. How many users could actually tell the difference between an Intel Cloudy Pussyinthewell chip and an Intel Muddy FishingLake chip let alone compare these to an AMD Athletesfoot CPU or an AMD Tampon CPU. So we end up just going for an Intel i3/i5/i7 because that at least sounds like it makes some kind of sense.