4770 vs 4770K?

[ehume]

If Intel is bringing back BCLK overclocking, do we need a K chip? For example, my i7 860 OC's better than my 875k, and it's older.

And I read that the K chips will be missing TSX and integrated graphics. True?

So, with a 4770, if it has BLCK OCing, TSX and IGP, should I look at that rather than a 4770k?

 

[ShintaiDK]

K chips seems to be missing VT-D, TXT, vPro and TSX.

They still got VT-X, IGP etc.

 

[lamedude]

First result on Google tells me its like SNB-E with the only reference clock ratio available are 1.25 and 1.66. If you're satisfied with a 25% overclock then you don't need it. 1.66 might need 2+GHZ RAM depending what memory multipliers are available and if your chip can't handle that big of a overclock you'll have to drop the CPU multi making you wish you had bought a cheaper CPU to begin with.

 

[Lonyo]

Best bet is to wait for reviews to find out what the CPUs are capable of.

 

[Wall Street]

Even without Bclk overclocking, I might prefer the 4770 non-K. If it can overclock 4 bins like current non-K chips it should be able to get within 10% of where I would clock a 4770k anyways (I try not to raise voltage when OCing).

I wonder on the K series chips how many overclockers will blow out the on-package VRMs. I would bet that the VRMs are going to be more similar to a 4 or 6 phase setup than an 8 or 12 phase setup.

 

[WhoBeDaPlaya]

Hate this God dang artificial dichotomy. On top of that, MicroCenter will probably only offer discounts and/or bundles with only the -K CPUs :|

 

[Borealis7]

Even without Bclk overclocking, I might prefer the 4770 non-K. If it can overclock 4 bins like current non-K chips it should be able to get within 10% of where I would clock a 4770k anyways (I try not to raise voltage when OCing).

I wonder on the K series chips how many overclockers will blow out the on-package VRMs. I would bet that the VRMs are going to be more similar to a 4 or 6 phase setup than an 8 or 12 phase setup.

wouldn't these exact same VRMs be offering protection from damaging overvoltage in the first place?

 

[Idontcare]

Hate this God dang artificial dichotomy. On top of that, MicroCenter will probably only offer discounts and/or bundles with only the -K CPUs :|

As Intel continues further down this road of being a virtual monopoly they will also continue further down the road of managing their business affairs more and more like that of a planned economy and less like that of free-market enterprise.

 

[VirtualLarry]

As Intel continues further down this road of being a virtual monopoly they will also continue further down the road of managing their business affairs more and more like that of a planned economy and less like that of free-market enterprise.

Sad days ahead. This is why I tell people to still consider AMD for their CPU needs. AMD doesn't generally play those games that Intel does, with controlling the market and forced market segmentation.

The only reason that we still even have PCI-E available to us for 3rd-party (non-Intel) GPUs, is because of NV's lawsuit settlement with Intel.

 

[Haserath]

The non-k does get pretty close to k on Ivy Bridge... I keep forgetting how close you can get them with a little bclk tweaking. So many people choose k and only take it to 4.5 anyway.

Hardly any limiting unless Haswell overclocks more.

 

[ShintaiDK]

The only reason that we still even have PCI-E available to us for 3rd-party (non-Intel) GPUs, is because of NV's lawsuit settlement with Intel.

What? PCIe is not owned by Intel.

 

[WhoBeDaPlaya]

The non-k does get pretty close to k on Ivy Bridge... I keep forgetting how close you can get them with a little bclk tweaking. So many people choose k and only take it to 4.5 anyway.

Hardly any limiting unless Haswell overclocks more.

Err... BCLK OCing on HW and non-HW parts are pretty different.

 

[TerryMathews]

Sad days ahead. This is why I tell people to still consider AMD for their CPU needs. AMD doesn't generally play those games that Intel does, with controlling the market and forced market segmentation.

The only reason that we still even have PCI-E available to us for 3rd-party (non-Intel) GPUs, is because of NV's lawsuit settlement with Intel.

This. Even when AMD had the performance crown (early mid K7 days) they were still very enthusiast friendly.

How many remember GFDs? Or changing multipliers on a socket A chip with a defroster repair kit?

THAT, my friends was hard core over clocking.

 

[BallaTheFeared]

Old man overclocking, what I can't understand is why we don't have an afterburner for CPU's yet, give me a slider please it's too hard.

Wasn't there something about real time clock speed changes with Intel without reboots?

 

[Idontcare]

What? PCIe is not owned by Intel.

As part of the Intel vs. Nvidia settlement, Intel agreed to not abandon PCIe for a lengthy period of time (5 or 10yrs IIRC) in its chipsets and mobo support.

The fear was that Intel was scheming to abandon PCIe in favor of a purely internally developed spec (along the lines of what they were thinking with thunderbolt versus USB) such that Nvidia and others would have to negotiate with Intel to secure the rights to develop products that would be compatible with the new bus interface standard.

Basically Nvidia was fearing that Intel was going to squeeze them (and AMD) out of the discreet GPU space (on Intel products at least, being 80% of the market) the same as Intel squeeze them out of the mobo chipset and integrated GPU spaces with the shift away from GTL+ FSB to QPI.

Old man overclocking, what I can't understand is why we don't have an afterburner for CPU's yet, give me a slider please it's too hard.

Wasn't there something about real time clock speed changes with Intel without reboots?

Haven't you OC'ed with an ASUS ROG board? They cost more than the CPU for a reason

 

[ehume]

K chips seems to be missing VT-D, TXT, vPro and TSX.

They still got VT-X, IGP etc.

This is what I wanted to know . . . except now I must look up what all those abbreviations stand for.

Best bet is to wait for reviews to find out what the CPUs are capable of.

Yeah, I thought that might be the bottom line. Except we need to know the kind of stuff ShintaiDK cited so we can make intelligent decisions.

 

[Haserath]

Err... BCLK OCing on HW and non-HW parts are pretty different.

You can still tweak it 5%. Just enough to get the top bin of 4.3 to 4.5.

 

[Wall Street]

wouldn't these exact same VRMs be offering protection from damaging overvoltage in the first place?

I don't know if you understand how VRMs work. The primary purpose of the VRMs isn't to protect the CPU, but rather to step the voltage down from the PSU voltage (12 V) to the CPU voltage (~1.2 V). If the VRMs blow, it is game over, computer doesn't boot.

 

[ehume]

According to this,

The overall concept behind VT-d is hardware support for isolating and restricting device accesses to the owner of the partition managing the device.

VT-d, at the time of this writing, includes four key capabilities

1. I/O device assignment. This feature allows an administrator to assign I/O devices to VMs in any desired configuration.

2. DMA remapping. Supports address translations for device DMA data transfers.

3. Interrupt remapping. Provides VM routing and isolation of device interrupts.

4. Reliability features. Reports and records system software DMA and interrupt erros that may otherwise corrupt memory of impact VM isolation.

Note that VT-d is not dependent on VT-x. That is, a VT-x enabled system can operate without VT-d, or without VT-d enabled or configured. You simply miss the benefits of the feature. Many people have asked about this point.

Why would they leave this out?

According to Wikipedia,

Intel Trusted Execution Technology (Intel TXT) is the name of a computer hardware technology whose primary goals are (a) Attestation – attest to the authenticity of a platform and its operating system (OS); (b) assure that an authentic OS starts in a trusted environment and thus can be considered a trusted OS; (c) provide the trusted OS with additional security capabilities not available to an unproven OS.

Intel TXT uses a Trusted Platform Module (TPM) and cryptographic techniques to provide measurements of software and platform components so that system software as well as local and remote management applications may use those measurements to make trust decisions.

I'm leery of TPM. Seems that a few years back the government wanted to hijack it . . .

Again according to Wikipedia (the Intel page is a morass of marketing hype):

Intel vPro technology is an umbrella marketing term used by Intel for a collection of computer hardware technologies to enable management features such as remote access to the PC (including monitoring, maintenance, and management) independent of the state of the operating system (OS) or power state of the PC, and security features.

TSX is Transactional Synchronization Extensions. According to this Intel blog, TSX is

At the lowest level with Intel TSX, programmer-specified code regions (also referred to as transactional regions) are executed transactionally. If the transactional execution completes successfully, then all memory operations performed within the transactional region will appear to have occurred instantaneously when viewed from other logical processors. A processor makes architectural updates performed within the region visible to other logical processors only on a successful commit, a process referred to as an atomic commit.

This is useful because

With transactional synchronization, the hardware can determine dynamically whether threads need to serialize through lock-protected critical sections, and perform serialization only when required. This lets the processor expose and exploit concurrency that would otherwise be hidden due to dynamically unnecessary synchronization.

So how much of this will be needed for enthusiast overclocking? As best I can tell, TSX might actually be helpful for me in getting my apps to run faster.

But maybe not. I don't pretend to understand all the ramifications. Thanks to ShintaiDK for specifying what will be absent.

 

[BallaTheFeared]

Haven't you OC'ed with an ASUS ROG board? They cost more than the CPU for a reason

My Pro board allowed it, my ROG board was sent back when I killed my i5 the second time so I never got to play with it, went with a WS Revo after that.

I was just being facetious.

 

[ShintaiDK]

This. Even when AMD had the performance crown (early mid K7 days) they were still very enthusiast friendly.

How many remember GFDs? Or changing multipliers on a socket A chip with a defroster repair kit?

THAT, my friends was hard core over clocking.

Why dont you just buy certain LGA2011 chips then? They cost the same as those chips you mention. And they got everything enabled, more cores and unlocked overclocking.

 

[DominionSeraph]

This. Even when AMD had the performance crown (early mid K7 days) they were still very enthusiast friendly.

How many remember GFDs? Or changing multipliers on a socket A chip with a defroster repair kit?

THAT, my friends was hard core over clocking.

If they were "enthusiast friendly", why were you modding your chip? It should have come with an unlocked multiplier.

 

[TerryMathews]

If they were "enthusiast friendly", why were you modding your chip? It should have come with an unlocked multiplier.

Because AMD chips in that generation weren't using electronic multiplier configuration, they were using physical?

Come on this isn't rocket science. They borrowed from the DEC Alpha which also had physical multipliers and not electronic.

Inteltards gonna tard? /DS

 

[ShintaiDK]

Because AMD chips in that generation weren't using electronic multiplier configuration, they were using physical?

Come on this isn't rocket science. They borrowed from the DEC Alpha which also had physical multipliers and not electronic.

Inteltards gonna tard? /DS

Palemino core K7 was electronic. Still locked.

http://www.cpuheat.wz.cz/html/AXP_multiplier/AXP_Multiplier.htm

Other positive effect of this change is that it allows for multiplier adjustments in range of 3x to 11x even on motherboards that do not support multiplier adjustments at all! And after L6 bridges change, it allows for complete range of 3x to 24x.

Maybe you suffer from the "In old days it was much better" syndrome

 

[TerryMathews]

Palemino core K7 was electronic. Still locked.

http://www.cpuheat.wz.cz/html/AXP_multiplier/AXP_Multiplier.htm

[/B]

Maybe you suffer from the "In old days it was much better" syndrome

In the old days it WAS much better.

Intel without competition is not good for the market. We have evidence, this isn't the first time it happened. The Pentium 1 era. The pre-K7 P2 era. And now.

Each time Intel tightened the reins, and the market stagnated..

I will say though, I don't think I saw that program at the time. Pretty neat.