AMD Makes 2nd Gen Ryzen Processors Official With Availability Starting Next Week (hothardware.com)

MojoKid writes: Today AMD announced official details regarding its new mainstream second-generation Ryzen family of processors. Pricing and detailed specs show some compelling new alternatives from AMD and a refined family of chips to give Intel even more competition, especially considering price point. These new AMD CPUs are all based on the 12nm Zen+ architecture and, at least initially, include four SKUs. The Ryzen 7 family features 8 cores and 16 threads along with 20MB of cache. Ryzen 7 2700 (65W) has a base clock of 3.2GHz and a turbo frequency of 4.1GHz. The top-of-the-line Ryzen 7 2700X (105W) ups the stakes with clocks of 3.7GHz and 4.3GHz respectively. The new Ryzen 5 family features six physical cores capable of executing 12 threads and 19MB of cache. The Ryzen 5 2600 (65W) has a base clock of 3.4GHz and a max boost frequency of 3.9GHz. The Ryzen 5 2600X (95W) ups those speeds to 3.6GHz and 4.2GHz respectively. AMD says that the Ryzen 5 2600, Ryzen 5 2600X, Ryzen 7 2700 and Ryzen 2700X will be available starting April 19th, priced at $199, $229, $299 and $329 respectively.

Comments

Re:Waiting for neural processors

[K. S. Kyosuke]

Yeah, we call those "GPUs".

Sounds the same

[iamhassi]

Sounds like what they have now, but the models that were X models are now base models. The 1600x is a 6 core that does 3.6 to 4.0. Now the 2600 is a 6 core that does 3.4 to 3.9. Hope I'm wrong and these new CPUs are amazing but from the looks of it this is just rebadging.

Re:Sounds the same

[Megol]

Close to what they have now but slightly improved and on a more advanced manufacturing process. Just as AMD said they would be.
No renaming.

If the leaked benchmarks are to be believed

[rsilvergun]

they've finally got within striking distance of Intel in single threaded performance (e.g. within 10%). And they blow Intel out of the water on Multi-threaded performance except for the highest end of Intel parts (e.g. the 7980XE outperforms a 1950x but it's 2x the price). Assuming this is right there won't be much point to buying Intel for gamers and (most) workstation users. The Ryzens we have today produce more stable frame rates (e.g. fewer 1% & .1% lows) thanks to the much better multi threading. Give them about 20-30% more single threaded performance with that advantage and it's going to be an AMD generation.

Re:If the leaked benchmarks are to be believed

[hcs_$reboot]

Plus the Ryzen is not subject to the Meltdown issue.

Re: If the leaked benchmarks are to be believed

[TeknoHog]

They [AMD] never hold up in real world tests

like Meltdown, for example.

Re:If the leaked benchmarks are to be believed

[SIGBUS]

It's a nice incremental bump, but I'm not so sure it's worth upgrading if you already have a Ryzen 7. At any rate, I've been quite happy with my 1700. If I hadn't already upgraded to that, though (and was going to upgrade from an old system) I wouldn't hesitate to pull the trigger, thoughI might save up a bit more and go for a Threadripper.

Re:If the leaked benchmarks are to be believed

[Anonymous Coward]

but all shipping parts are vulnerable to spectre and the followons. It's hard for me to justify buying a CPU, given that they remain worth feeding power for ~5 years and a spectre fix is expected in ~1yr. I'd hate to be a CPU company right now.

Re:If the leaked benchmarks are to be believed

[ravenshrike]

There is no SPECTRE fix coming. There is only hardware level SPECTRE mitigation. A full SPECTRE fix would require dropping out of order execution entirely. That ain't happening except for maybe the most secure of systems.

Re:If the leaked benchmarks are to be believed

[Megol]

Yes there are! Meltdown is due to Intel design decisions and can easily be fixed in hardware.

Current and older Intel systems can only be fixed by making the OS do more work - but that can be a fix too in that one can't abuse the Meltdown bug for anything. Not sure if current designs actually do enough for this to be true though, it would be easier to allow some information exposure in which case I'd agree that it's only mitigated but to a large degree.

Re:If the leaked benchmarks are to be believed

[F.Ultra]

A hw fix for Spectre2 is coming according to Intel.

Re:If the leaked benchmarks are to be believed

[F.Ultra]

Yes there is, or rather there will be since Meltdown (Spectre3) is a design flaw in the current Intel architecture that they can and will fix in new cpu:s.

Re:If the leaked benchmarks are to be believed

[Megol]

I have a Dell Precision m6300 and have periodically used it as the daily computer. Browsing, writing code, writing text, playing simple games, watching videos etc. The main limitation of the machine is the HDD which will probably be switched to an SDD soon.

That machine is 10 years old. The main limitations are the I/O hardware: the LCD screen isn't too good when used to modern bright IPS panels, the keyboard while very good for its time doesn't have the feel of a modern quality keyboard and the touchpad is just almost useable.

While it is nice to have a modern machine the power of even old computers are enough for almost everything.

Re:If the leaked benchmarks are to be believed

[Tsolias]

new benchmark from the el chupacabra tech review site and burrito surfaced yesterday with an x470 motherboard but with 3200MHz RAM.
2700x is ahead of the 8700k is all cases.
The "leaked" benchmarks were on an x370 motherboard with the minimum compatibility for the 2700x.

Re:If the leaked benchmarks are to be believed

[HiThere]

On current processors, I believe that you are correct. But there is nothing inherent in out of order execution that implies any SPECTRE mode capability. You just need *much* better isolation that is possible on current hardware.

In proof of this claim, imagine that you divide the chip into totally isolated processes that communicate via UDP. They don't share memory at all. When a fork in execution is seen coming, you send a message to two different processes to start working on one branch of the problem. This will be a bit verbose, because you need to send copies of all the resources that will be needed. After awhile you figure out which branch was the correct one, and listen to only that response. You don't know anything about what happened in the external processes, and the external processes only know what you sent them. This would be a bit slow, of course, but it's a thought experiment, not a proposal. The crucial step is the isolation of the separate processes.

Re:If the leaked benchmarks are to be believed

[Tough Love]

I'm not so sure it's worth upgrading if you already have a Ryzen 7.

True, but it's a no-brainer vs Intel for a new build.

Relevant to slashdot posters...

[AbRASiON]

But have they fixed this yet? Their second bug in Linux?

https://bugzilla.kernel.org/sh...

Re:Relevant to slashdot posters...

[Anonymous Coward]

But have they fixed this yet? Their second bug in Linux?

https://bugzilla.kernel.org/sh...

Yes.

Re:Do they include Meltdown fixes?

[ELCouz]

Hand over your geek card....it's revoked!

Re:Northbridge problems with Ryzen

[ELCouz]

I'm betting Intel is throwing money at researchers to discover vulnerabilities on competitors products.

Re:Northbridge problems with Ryzen

[Joe_Dragon]

stop shelling for intel

AVX512

[JBMcB]

One thing that might give Intel an edge is the upcoming AVX-512 extensions in the next cycle of processors. It'll allow two more registers for vector operations, along with a bunch more opcodes. It doesn't accelerate all operations, but what it does accelerate usually gets a pretty good speed boost. There's an HPC blogger that benchmarked the heck out of a couple of SSE/AVX/AVX2 chips, and each successive part increased some SPEC operations by 20-40%. Video encoding in particular got a good 30% boost from generation to generation - much more of a boost than the CPU optimizations alone.

Of course, AMD could clone these features, but they've been lagging in support for AVX. The Ryzen parts have half the AVX registers of the Intel chips. Sometimes they can make up for it through sheer parallelism, but not for every workload.

Re:AVX512

[Anonymous Coward]

Nope, leaving AVX at 128-bit is how they control power consumption (65W) vs a similar Intel with 256-bit AVX (95W). They made that choice consciously. Intel heavily downclocks CPU when 256-bit and 512-bit AVX is used, so the gain is not that big.

Re:AVX512

[Dwedit]

If you're doing something parallel enough that vectorization speeds it up, you might as well do it on a GPU instead. Even an integrated GPU is much faster than vector instructions.

Re:AVX512

[Agripa]

I wonder if Intel still cheats and uses a special compiler that favours their chips in secret, underhanded ways. You know, like they use to do.

Yes, Intel still does this both with their compiler and libraries. The only requirement from the court was that Intel say that they do this in their documentation.

Re:Waiting for neural processors

[AHuxley]

Something like Gigatron TTL microcomputer AC? https://gigatron.io/

Re:Northbridge problems with Ryzen

[andydread]

Didn't Intel produce processors in Israel at some point in the past? I'm sure there are some shared interests that can be dug up between this company and Intel

I saw some of the posts complaining

[rsilvergun]

about "just" 11 days uptime. As a Windows users I'd like to say, welcome to our world baby.

Re:I saw some of the posts complaining

[AbRASiON]

Some are getting less. It actually doesn't impact me but it does reduce my confidence using it on my next FreeNAS machine.

Re: Northbridge problems with Ryzen

[Anonymous Coward]

Nope.

If they did, they wouldn't have held back the information that Intel is just as susceptible to the ASMedia bugs as AMD is. That Israeli con-job was an Intel-sponsored hit-job from start to end, there is no doubt about it. Just look at the material, including the interview Arstechnica did with them. Look at how they twist all over the place to avoid saying "Intel" for instance, and add to that all the other weird stuff in the "story". It just doesn't add up, unless you add "Intel" to the mix. If you insert that in all the blank spaces, it suddenly starts making sense.

Spectre

[DrYak]

but all shipping parts are vulnerable to spectre and the followons.

Remember that they are 2 different vulnerabilities named "spectre".

Spectre variant1 affects AMD as it affects virtually any CPU under the sun that does speculative execution.
But relatively to the other vulnerability, it's much more moderate, it's the CPU speculatively access data to which the current process HAS access anyway. (e.g.: getting pass a size check and reading from another array of the same thread).
There are very few corner case where a thread should not read data to which is normally has access to (mostly in situations of JITed 3rd-party provided code - e.g. Javascript downloaded from the web - running in the same context as some sensitive data - e.g.: you password manager plugin) and proper process separation is the correct long term strategy anyway.

Spectre variant2 abuses the way speculative indirect branching is done (jumping to a location which is not known in advance : jump tables like C++ virtual methods, some possible C's "switch" implementations, etc.), and it's extremely CPU dependant as each CPU has a different way to speculate that.
It's much more scary than Spectre v1, because one process of the attacker (e.g.: a program that the attacker has uploaded into an Amazon EC2 VM - something that he should be able to do), can cause an entirely unrelated process to speculatively jump and execute arbitrary locations (e.g.: the *hypervisor* handling that VM could be forced to jump to selected pieces of code, and thus doing some return-oriented-programming. Again that's the hypervisor we're talking about, a completely different piece of code to which the attacker should never have had access in the first place).
- On Intel hardware (Xeons), spectre v2 exploit have been demonstrated successfully by Google's project zero
- AMD hardware can do speculative branching (so AMD has marked their hardware as potentially affected), but as of today nobody has manager so successfully demonstrate a usable exploit, and the jury is still out whether this actually exploitable (might be that the peculiar implementation AMD CPUs use to branch speculatively cannot be abused in any useful way to begin with) (so AMD is still indicating exploitability as very probably unlikely).

So currently, on AMD you're still safe regarding Meltdown (AMD hardware doesn't read data it doesn't have actually access to), and Spectre v2 (there are probably no way to abuse indirect branch prediction to achieve anything meaningful).