Do we still build, is this the right place?

[PlaneMan]

Some competition for Intel would be great news for the consumer, going back to the Socket A days would be nice, easy, cheap performance gains.

[Jason]

If it’s the dog’s swingers, I’ll be building a rig with it, if only to resell straight after. I’m excited.

[Ed P]

Drezha wrote:

Ed P wrote:

‘nice fire modelling machines on the cheap’ Not knocking your idea, but wouldn’t a Beowulf cluster of Pi Zeroes be better for Monte Carlo simulation work? Agree that CUDA etc give you ready access to parallel operations but there are cases where individual processors MAY be better. The example below is for disease, but an ‘explosive’ fire could result in similar propagation characteristics (i.e. one source rapidly becoming multi-source at geographically disparate locations.). …. However for creating a cluster running some thing like “condor” Which can be used for things like Disease out break modelling, where you are running the same mathematical model millions of times with varible starting points. (size of out break, wind direction, how infectious the disease is etc.. ) so thing like the Pi would be ideal. as you are general looking for a full blown CPU that can run standard code.http://research.cs.wisc.edu/condor/

It’s CFD modelling, not Monte Carlo simulations. We’re looking at smoke flow around a building normally (the software is here). My current firm do have a Monte Carlo simulation tool but it isn’t used much (I don’t think they market it enough). The modelling software doesn’t make use of CUDA either (a German research team tried to get the code running on CUDA but was unable to – lots of data I/O I believe the issue was which slows it down to make it not worth the bother).

Ah, interesting – I can now see why you would go for a ‘cellular’  approach. You may be interested in looking at combining the worlds of games to drive the simulation as they have done a lot of the graphics heavy lifting. I’m thinking in particular of perhaps using Voxels along with simple cell modeling to drive the voxel procedural generation. The attraction of going this way is that there are a number of readily available commercial game engines that have done the hard graphics graft.

[Jason]

Pre-launch Ryzen benchmarks leaked: http://wccftech.com/amd-ryzen-7-1800x-1700x-1700-benchmarks-leaked/.

Clock-for-clock, it’s as good as Intel in single-threaded workloads. Intel has the lead, as the Core i7-7700K is clocked 600 MHz higher than the Ryzen 7 1800X, so in cases where only single-thread matters, the Intel chip may be the best choice, at least until higher-clocked Ryzen models are released.

In anything else, only the 10-core i7-6950X can beat the 1800X, and that’s three times the price and needs an LGA 2011-3 motherboard.

For general computing, including gaming, the Ryzen chips are looking unbeatable. Even the 1700 punches hugely above its price, trouncing the i7-7700K for less money and at nearly a third fewer watts. A decent, full-featured AM4 board from Asus can be had for about £80, and cheaper prices and products are bound to follow.

Of course, these benchmarks may not be real, though they look it.

[The Duke]

Looks good, but let’s wait for he independent reviewers to get their hands on them. I’m due an upgrade so I’m all for competition.

[Dave Rice]

Sorry but this looks very much like the story touted in the Bulldozer article. A very narrow and focused set of “leaked” results.

[Jason]

If this is another post-fact scenario, I’ll probably never recover. Lord knows, there’s enough bull in the world already.

[Dave Rice]

Well the benchmarks are out and single threaded is the Achilles heel again.

If you have the right workloads to throw at it (unsurprisingly the same sorts as the leaked benchmarks) it is certainly a competitor. For mere mortals it’s still work in progress.

I think this headline sums it up pretty well “An i5 in Gaming, i7 in Production”.

But it’s a hell of a lot better than what’s gone before that’s for sure and that is something to rejoice.

[Anonymous]

Plus you have to remember the cost of the processor that it is being compared to.

[Dave Rice]

Yes you do.

Ryzen 7 1700 = $329, i5 7600K = $242, i3 7350K = $168

Average Gaming

[Jason]

It’s a little bizarre, as clock-for-clock it seems to beat the i7-6900K, even in games, yet clock-for-clock it doesn’t beat the i7-7700K. Yet the underlying cores of both those i7s are more-or-less the same. I’m not sure we’ve got the full picture yet. As things stand, the R7 1800X will eat into some of the very top end, but it won’t eat into the high-end mainstream, as the i7-7700K is cheaper (significantly).

All the action may well be in the i5 space and below, where if the price is right, the R5 and R3 chips could seriously hurt Intel. A 4 GHz 6+6 design at $199 is likely to walk all over an i5, as is a 4+4 design over the i3. I have yet to hear word of any Ryzens with less than 4+4, so there could be some real bargains in the Pentium space.

[Dave Rice]

Yes there could be, but it’s all still ifs, buts and maybe.

On the current showing it’s as before, cherry picked mutli threaded benchmarks and lacking in single thread. We all know there were some stupidly priced Intel chips that didn’t stack up against their own competition that no-one could work out who would buy them.

If the high end Ryzen struggles / holds it’s own against cheaper i3 and i5 in every day Joe type workloads it’s hard to see how the lower ones will fare better, but we’ll see. I really cannot see anything here that will have Intel quaking in their boots at the moment. Just look at what they’ve done with the Pentium G4620 and the i3-7350k Review: almost a i7-2600K and that’s a 60 watt.

Ryzen’s certainly a vast improvement and you can see where and how they can make Zen 2 more competitive again. But I’m afraid we’re in the reverse fanboi scenario from where we were in the days of the Opteron.

[Ed P]

Ars has an interesting review. Not a great cpu at the moment, (depends on price etc), but the next iteration may make an interesting  server. What I found interesting in the review was that AMD totally misjudged the inertia (and difficulty) of writing multi-threaded applications, and how few things need multi-threading rather than grunt.

[Dave Rice]

I think that’s a very fair review.

Ryzen isn’t a stinking pile of poo but neither is it what it was hyped up to be. How much of that hype was AMD PR and how much it was fanboi I wouldn’t like to say.

But hey ho, if it got us the P G4620 and i3-7350K that’s good and it looks like they have an architecture to build on. I’m sure Ryzen will do well in the niches it appeals to.

[The Duke]

Over hyping never ends well, it lets down more people than not. Leading to less future sales though damaged relations.

Whenever i recommend  a product I’ll alway highlight the items pit falls, I’ll always under sell it, while saying it’s great and will fit your needs.

Otherwise of i just sell the high points, leaving out the low, the buyer will always be underwhelmed. If they know going it, and you under sell the good bits, the buyer will be far happier with the same product.

It’s all about setting a realistic image of the product.

Part of my uni course tonight marketing, and everything I said above is the oppersite to the complete rubbish they was teaching. It was the worst subject I was ever forced to sit through.

I went the first two seminars, realised all the answers was common sense, never went another and aced the BS touting exam. Marketing, is BS on an industrial level. So one needs to teach these (all) companies it’s about setting expectations.

Under sell over deliver, and the customers will come via word of mouth. Apple with all the hate we give the do this, yes they have thier dog and pony shows, preaching to the choir, but their advert are very subtle,very  little info and pretty pictures. It’s up to the user to decide if they like it. The product speaks for itself. Not to me, but to 100s of millions.

But with apple being a religion, they do have an easier job than most at selling.

AMD have for 10 years under delivery, so this hype all but the old school fan boys just don’t believe it.

Nintendo are probably on their last chance with the switch, if this one doesn’t deliver, they are finished as a hardware firm, to much hype with little end product. The wii promised loads, and sold loads, but no one actually used them or bought games, as we all realised we had been duped. So no one bought the Wii U.

The switch is trading on the DS success and Mario and Zelda. However if Zelda and mario are the highlight (ie no other decent titles on a decent sale)  and if the hardware isn’t what the massive hype is hyping it as. Nintendo will be trading on creating mario games for android and ios within 2 year.

I actually think a Nintendo market place app for Android* would be a real winner and make alot more money than the switch, with relatively  small outlay.

They would need a different approach for ios as you can’t have your own market place on Apple and would need to cut apple in with 30% of all sales. They could avoid this on android. But 70% of something is better than 100% of nothing if they don’t launch on ios.

[Jason]

The Ars review is very biased and overlooks some important factors.

First, it’s clear that AMD’s version of Hyper-Threading isn’t yet working in many titles. Some games benefit from having it turned off. That’s not right, and quick patches to games — not hard to do — should see noticeably gains.

Second, clock-for-clock, AMD is in general only 10-15% behind Intel, and as above, some of that ground may be made up. But here’s the thing. Those extra 4+4 cores aren’t in general any help in games, so clock-for-clock it’s likely that the 4+4 versions of Ryzen will game pretty much the same as the 8+8 versions.

But AMD will be able to ramp up the clocks. The main reason the i7-7700K puts in such a strong hand against the 1800X is because it’s clocked 600 MHz, or 17%, higher. A 4+ GHz 4+4 Ryzen will close the gap significantly, and my guess is it’ll come in at around £200. Would you take a 10% drop in performance for a saving of £120 over an i7-7700K? I would.

We’ve not seen the whole picture yet. Ars and so on are rabbiting on that AMD’s shares dropped 7% on news that gaming performance wasn’t all that. This is true. What Ars failed to mention was that AMD shares were up about 700% on where they were this time last year, so a 7% drop is all but meaningless.

[Jason]

The Duke wrote:

Nintendo are probably on their last chance with the switch, if this one doesn’t deliver, they are finished as a hardware firm, to much hype with little end product. The wii promised loads, and sold loads, but no one actually used them or bought games, as we all realised we had been duped. So no one bought the Wii U. The switch is trading on the DS success and Mario and Zelda. However if Zelda and mario are the highlight (ie no other decent titles on a decent sale) and if the hardware isn’t what the massive hype is hyping it as. Nintendo will be trading on creating mario games for android and ios within 2 year.

You’ve obviously never visited our house. My son and I have sessions on the Wii U whenever he comes back from uni (and the Wii is still plugged in). We’ll be playing Zelda on the Switch today when he returns — launch day today, and I have both ready to go.

You’re forgetting that Nintendo is still posting profits, despite the “failure” of the Wii U. Folks have been talking about the company’s demise since about 1996, when the N64 launched with cartridges instead of CDs. The reality has always been somewhat different to the “alternative facts”.

I don’t expect the Switch to outsell the other two main consoles. Nowhere near. But I’m confident it’ll sell well enough to keep Nintendo in business and to give joy to those many games players who take pleasure from the key Nintendo franchises and appreciate games made with a passion and dedication to detail and excellence that no other publisher comes close to matching.

[The Duke]

Just cos your house is using the wii doesnt mean it’d a success, obviously some are using it, but it’s a known flop.

I use and like alot of things that are seen as flops.

The cube for me one one of the best consoles ever to be a flop, I loved the thing, but even then it was Mario and zelda that wad carrying it. Both was brilliant on that machine. Loved my cube. Our wii actually had more cube hours played than wii games.

Paper Mario on the wii I liked, but that too wasn’t a hit.

Nintendo is a one trick pony imo, they keep trying but ultimately they always have to lean on ip’s from 1980s.

Do you own a 3ds Jason? We never got one, we had a hand for of ds’s, the kids never asked for the 3ds, they had moved onto tablets and phones, however my nephew asked for a 3ds for Xmas, and they where almost £200, he never got one.

I was amazed somthing so old, wad selling for so much. Iirc correctly the DS was about £120 when new, and was in a league of one, we’ll 2 bit the psp wad much dearer and not not that decent for it. We had one psp, it’s UHD days, it got “shelved” quite quickly, the DS though live a fullfiled life, even had age cards and turned them into media players, the kids used to watch films on our early adventures on the continent.

I liked the DS alot. The previous I like more though, gameboy advanced I thing it was called, the first flip game boy. Bit again it was Mario that carried that, for its short life.

It’s amazing how much cash I’ve given Nintendo, I’m defo no hater, just not the fine I once was.

I’ll be getting at least one Switch that’s for sure, but that’s cos I’m a sucker for shiny crap lol. I’m now hopeful that it will live up to it’d hype though.

I have to disagree about game making with passion etc you said. I’d call it regurgitation of old ip’s,  relying on fans blind loyalty to defend it.

There are many great game developers out there making brilliant games,with new ideas.

[Jason]

I agree that my experiences don’t necessarily indicate a trend. I was just pointing out that there’s more than one way to measure success. If the company is turning a profit, I would count that as success. You also said that “no-one” bought the Wii U and that no-one used the Wii or bought games for it, as “we all realised we’d been duped”. Those statements are clearly false. That was my only point. As for relying on ’80s IP, that’s true enough, but when the games still sell in the millions and are eagerly awaited, does it matter?

Yes, we have a 3DS. A great machine it is, too. No idea how much they cost now. Got it at launch.

[The Duke]

I’d say the strength of sales, or lack of it, of the U, indicates the Wii high sales don’t represent its success.

Wii fit is what sold the Wii, it was a massive gimmick that was soon seen through. Though as I said, mario was good on it as it always is.

If you want a Mario player it’s a good by, but Mario is Mario is Mario. Since Mario 64 they all feel the same, just more shiny. As is Mario cart, of anything it doesn’t have the charm it once did.

[Author]

Posts